TURBINE STEAM PATH MAINTENANCE AND REPAIR Volume 2

TU RBIN E STEAM PATH M AINTENANCE AND REPAIR Volume 2 William P. Sanders, P. Eng. Library of Congress Cataloging-in-Publication Data Sanders, William P. Turbine Steam Path Maintenance and Repair Volume Two / William P. Sanders, P.E. p. cm. q.cm Includes index ISBN 0-87814-788-8 Copyright © 2002 by PennWell Corporation 1421 South Sheridan Road Tulsa, OK 74112 800-752-9764 sales@pennwell.com www.pennwell-store.com www.pennwell.com Cover and book design by Robin Remaley All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transcribed in any form or by any means, electronic or mechanical including photocopying or recording, without the prior written permission of the publisher. Printed in the United States of America 1 2 3 4 5 06 05 04 03 02 Turbine Steam Path Maintenance and Repair—Volume Two PREFACE The Turbine Steam Path, Damage, Deterioration, and Corrective Options This book has been prepared for those technical people responsible for the operation and maintenance of steam turbines. Steam turbines represent a complex technology for units commonly designed to operate hundreds of thousands of hours while being subjected to a severe environment and a variety of operating phenomena capable of degrading their condition. These units are required to continually operate in a reliable, safe, and cost-effective manner. Under such circumstances, these units cannot maintain their original design-specified level of performance indefinitely. All units will deteriorate with age. Owners anticipate this, and designers will normally leave an adequate margin, knowing that some level of such deterioration is tolerable. The technology of steam turbines—while mature—continues to evolve. More accurate and time-responsive diagnostic tools and techniques are becoming available to assist in predicting when a unit has deteriorated to the extent that corrective action is required. Similarly, tools are available to assist the operator in analyzing problems and determining the effective corrective action best suited to the condition causing deterioration. The improved understanding of unit condition and rates of deterioration now achieved, together with advances in materials, should allow units to be maintained in a manner that will help minimize maintenance concerns and costs. It is the premise of this book that units “as supplied” will fulfill two basic requirements: xii Preface • It is assumed the unit “as designed” represents an optimum selection of component sizing and arrangement • It is assumed the unit “as delivered” meets design specification within the range of tolerances provided by the design engineer, i.e., unit components have been manufactured, assembled, tested, and installed in such a way that they are in compliance with the original design specification The implication of this second assumption is that if nonconforming situations or conditions arose during the total manufacturing process (and exist within the unit), they have been evaluated by a competent design authority in the engineering organization of the manufacturing company and have been assessed as not having an adverse impact on the potential performance of the unit. In terms of turbine unit components, “design optimum” is a difficult term to define. The entire design process is one of compromise by the designer who wants a unit to be both efficient and reliable. These requirements often represent competing demands, forcing the designer to select from among various elements, possibly electing to downgrade one aspect of these requirements to meet the demands of the other. This is done consciously and with detailed evaluation to provide a balanced selection. Units delivered by a manufacturer represent the supply of elements that conform to the design principles established by his or her design function, and conform with the best technology available to that supplier at the time the design specification was prepared. However, the operator must recognize that the labor and material costs involved in building a steam turbine are high, and turbine suppliers must be able to produce units at competitive levels sufficient to allow them to achieve a profit margin enabling them to sustain business as well as finance further development. xiii Turbine Steam Path Maintenance and Repair—Volume Two Many power systems are currently experiencing significant changes in how they operate. Pressures from deregulation, environmental concerns and legislation, and an aging fleet of power generating equipment are shifting emphasis from the installation of new capacity to the maintenance and care of the old. There is a continuing increase in demand for electric powe,r but new capacity installation is not keeping up with it. Operators of turbine generators are therefore required to meet this demand with their existing fleets— aging units requiring greater care to reduce the possibility of forced outages. The prospect exists of units experiencing extended outages as damage is found at planned outages. Historically, as units have aged they have tended to be used less frequently. They are initially placed on spinning reserve and ultimately placed in reserve, mothballed, or retired—their capacity replaced with newer, more efficient units. An advantage of this dwindling reserve is that older units have continued to operate at high load factors and therefore become less susceptible to the rigors of start-up, shut-down, and the associated thermal transients. Unfortunately, there have also been fewer opportunities for plant maintenance to proceed with the maintenance outages required to maintain unit operational health. Maintenance problems associated with keeping aging units available are only going to increase. Operators who are expected to provide power on demand are going to experience even greater future challenges of damage and deterioration. They will be expected to identify not only the damage, but also the causative effects, and then find immediate solutions that will not jeopardize system security. This book examines the damage, deterioration, and failure mechanisms occurring with unfortunate consequences—on some units, with monotonous regularity—within the turbine steam path. These various forms of degradation can be the result of a number of factors related to conditions often beyond the control of operating and maintenance personnel. However, even if the steam turbine is operated xiv Preface precisely as intended by design, and suffers no external degrading effects for its entire operating life, the steam environment is one that can cause components to suffer various forms of distress. Under normal circumstances, the design process selects and defines individual components suitable for the design operating life of the unit (normally about 200,000 hours). At a mean load factor of about 75%, this represents a 30-year operating life. A number of unavoidable influences affect the operating life of the various components comprising the turbine. These include the steam environment itself, the stresses induced in the components by rotation, and stresses induced in various portions of the unit by expansion of the steam through the blade passages. There are also the effects of the high-pressure steam, causing high-pressure drops across some components that must be contained by the casings. External factors that can affect the reliability of components of the steam path and act to lower the expected operating life include the possible formation of corrosive elements at various locations within the steam cycle or impurities gaining access from in-leakage at sub-atmospheric pressures. There can be unit trips caused by a number of circumstances, from system trip electrical faults to lightning strikes on power lines. Many of these factors, while possibly occurring in a 30-year operating life, cannot be anticipated in terms of when, where, how many, or how severe their effects might be. The damage and deterioration that occur within the steam path can be of several forms. It can result in a gradual material loss—the growth of a crack—or an immediate failure causing a forced outage. Gradual deterioration can (depending upon type and location) be monitored and replacement parts made available, or corrective action taken to rectify the situation before it extends to an unacceptable degree. Immediate failure is most often the consequence of either mechanical rupture or the presence in the steam path of some foreign object, either generated within or having gained access from some external source (including “drop-ins”). xv Turbine Steam Path Maintenance and Repair—Volume Two In writing this book, I have tried to present information that plant personnel will be able to use to make value judgments on the type and severity of any damage, suggest possible causes, and then consider the most appropriate corrective actions that are available. To aid in the recognition and classifying of operational damage and deterioration, photographs are used to illustrate unacceptable or suspect conditions. Many of the damaging phenomena considered in these chapters do not occur in isolation. It is possible that several can and will occur simultaneously, demonstrating that components are subjected to more than one degrading influence. A condition may initiate due to one damaging mechanism introducing a condition of weakness, which then allows another mechanism to become predominant and drive a component to failure. This situation often occurs even though the driving mechanism would not have been capable of causing failure had not the weakness been introduced by the first, or initiating, mechanism. Before considering degradation and failure in any detail, it is important to define what constitutes failure and/or deterioration. An important consideration in any case of evaluation and condition assessment of a turbine is establishing what constitutes failure. The definition I find most acceptable is this: A condition exists within the unit that while it would not prevent the unit from returning to service and continuing to develop power, it could force it from service before the next planned outage. Various other definitions exist, and the definition of failure used in any situation—and therefore the responsibility for correction—can be controversial. This controversy is to some extent aggravated by possibilities; e.g., a crack that has been determined to exist may be predicted by the methods of fracture mechanics to be growing at a rate that would not cause complete rupture, forcing the unit from service before the next planned outage. xvi Preface As reserve power margins diminish, steam turbines—that currently have operating periods between major maintenance outages of three to eight years—could be forced to operate longer than intended when they were originally returned to service. Under these circumstances, it is difficult when making a prediction of a unit’s future operation to be certain there will not be some major change in its operating parameters. Parameters that can influence an acceptable definition of failure in any situation include the exact operating period, the unit load pattern, and the steam conditions the unit will experience over a number of years. A simple and conservative solution to this definition of failure would be to change any suspect component showing any crack or unacceptable damage-or-deformation indication. This may appear to be an expensive option, but is considerably less expensive than a forced outage requiring weeks or months to open, repair, await replacement parts, replace those parts, close the unit, and return it to service. Defining efficiency deterioration is somewhat easier. It is even possible to quantify such deterioration in terms of reducing steam path efficiency and unit output. What is not possible to determine is the extent of any mechanical deterioration that may occur and cause efficiency deterioration. This is an unknown situation not recognized until complete mechanical rupture occurs. There is normally no manner to predict such an occurrence—damage could be in the incubation phase—even when an examination of the steam path is made at maintenance outages. During operation, certain situations and phenomena are known to occur that have the potential to initiate damage or to cause deterioration in performance. These damaging and deteriorating phenomena can be of a continuous or intermittent nature, produced as a consequence of transient operating or steam conditions. Such phenomena can also be the result of sudden mechanical failures of components that cause more extensive consequential damage. The xvii Turbine Steam Path Maintenance and Repair—Volume Two most commonly occurring of these degrading effects are related to the formation of moisture in the steam path or solid foreign particles, possibly from the boiler or scale generated within the superheater and reheater tubes. Other sources include chemical contaminants that are introduced or gain access to the steam path on which they are deposited, and possibly act as corrosive elements. The other principal degrading conditions are the operational phenomena occurring during the operating life of the unit. The first two chapters of volume one provide general information. The first outlines what is considered necessary to define and constitute a maintenance strategy that represents management’s commitment to maintaining a healthy system. This chapter also outlines means of monitoring conditions indicative of damage. The second chapter deals with the spatial arrangement within the steam path and the factors that affect it. This is important because the performance (efficiency and reliability) of a turbine is influenced considerably by the alignment of the unit and the resulting axial and radial clearances and “laps” that are achieved in the hot operating condition. Chapters 3, 4, 5, and 6 discuss the various phenomena known to affect both the efficiency and structural integrity of the components. In the second volume, chapters 7, 8, and 9 consider repair and refurbishment options currently available. Fortunately, there are ever-present advances in these technologies, and as experience is gained, newer and improved methods develop to be applied to older units so they can continue to operate with high levels of availability—often with improved efficiency. Chapter 10 considers seal systems and gland rings, and provides means of estimating the financial penalties associated with excessive leakage. Seals are one area where operators and maintenance personnel can influence the cost of power generation and help reduce the cost of power to their customers. The final two chapters, 11 and 12, relate to quality and the inspection of elements being manufactured to replace damaged components. This is an area where many engineers feel the cost of xviii Preface undertaking such inspections is difficult to justify. However, what happens when components—manufactured when they are required in an emergency to return a unit to service—have any form of fault and force the unit from service prematurely? In such a case, the cost of inspection—ensuring that a supplier’s quality program is prepared and operating properly—is well justified. It is often said, “There isn’t time and money to do it right, but there is always time and money to correct it.” This statement is well applied to the manufacture or repair of components in an emergency, because the cost of a second outage is just as high as the first, and far more embarrassing. Because the steam turbine is a thermal machine designed to convert thermal energy to rotation kinetic energy, I have included an appendix that provides the basic thermal relationships required to understand the turbine and its operation. Situation evaluation The more susceptible areas in any turbine unit are a function of many complex factors—individual stress levels, stress concentration, mode of operation, and the operating environment. Individual components are also greatly influenced by the expertise with which the parts were designed, manufactured, and assembled, and the operating transients to which they have been subjected. The diversity of the factors that can contribute to damage precludes any generalization of cause or value. Steam path components are subjected to high stress, both direct and alternating. Many parts operate at high temperatures and are of complex forms interacting with one another in unpredictable ways. These factors, when combined with load and temperature transients that occur during operation, combine to make the steam path highly sensitive and a major source of concern to the designer and operator. While some concerns are common to most operators, the type of deterioration or damage to which any component or area is subjected xix Turbine Steam Path Maintenance and Repair—Volume Two normally varies from unit to unit. This accounts for the variety of concerns expressed by maintenance staff and the different dispositions of the various nonconforming conditions that will be developed in any situation. In many instances when corrective action is required, there is no optimum solution that can be followed without deviation. Operation and load demands will often negate the optimum. At other times, costs, special tools, skills, and the availability of replacement parts could require some form of compromise. These compromise solutions may have to be adopted from necessity, but the final disposition should provide the best balance between cost, risk, and the immediacy of returning the unit to service. The logical approach to maintenance and repair dispositions is: • Consider the available alternatives in terms of the original design requirements of the affected components • Evaluate possible solutions in terms of departure from the design-specified requirements Many “repair” or “accept-as-is” dispositions will have only a limited effect on unit performance and can be readily accepted. Other repairs can be proposed and accepted, representing a compromised condition. Such options should only be accepted on the basis that the unit will be operated with this compromised solution for as short a period as possible, and that the selected option does not represent a significant level of risk in the short term. If this is possible, plans should be put into effect immediately to develop an acceptable solution that can be undertaken within a reasonable time. The maintenance options The satisfactory performance of a steam turbine is influenced considerably by the manner and expertise with which it is maintained and xx Preface the load patterns it follows. While the plant operating engineer can control, to a large degree, the maintenance of the units for which he is responsible, he is unfortunately unable to exercise little influence on operating patterns. This is a responsibility of dispatchers who have a mandate to serve the demands of their clients rather than the turbine generators of their system. For maintenance to be cost-effective, it must be planned. When signs of distress, excessive wear, misalignment, or component deterioration are detected, the need for corrective action must be considered. These corrective actions should help ensure the situation does not deteriorate further, to the extent the unit is placed on a forced outage status, severely load limited, or suffers an unacceptably high degree of deterioration in efficiency. There are general maintenance requirements for any unit. Guidance for these is provided by the designer and should be followed for all routine matters. The designer will also provide recommendations for the operating time between opening sections of the unit for periodic maintenance and examination. During these maintenance outages, any findings that could affect unit performance must be reviewed in relation to their possible long-term effects. Maintenance actions Opening a unit for maintenance provides the opportunity to make repairs or to install replacement parts when the necessary skills and special-purpose tools are available. Such an opening also allows replacement parts to be ordered, which can be placed in the unit at the current or later outage, depending upon the delivery and required period of the outage. Replacement is made when an evaluation of any found operational nonconformance is judged to be placing the unit at risk if returned to service without correction. A detailed evaluation of each nonconformance should be made, and it should indicate if, and what, actions are required. xxi Turbine Steam Path Maintenance and Repair—Volume Two The principal purpose of a steam turbine maintenance inspection is to detect potential problems at an early stage. If this is not done, relatively minor situations could progress to the extent a forced outage or excessive loss in unit output and efficiency could occur. During such a maintenance inspection outage, parts can be examined visually for indications of failure, wear, or distortion. Also, non-destructive tests can be applied to critical components to determine if their ability to continue to perform satisfactorily has deteriorated and, if so, what remedial action should be taken or planned. A nonconformance in any part of the steam turbine unit is considered to have occurred when there are signs of mechanical failure, excessive wear, or any form of deterioration that has the potential to adversely affect the performance of the unit. Such nonconformances must be reviewed for short- and long-term effects. As soon as unit inspection indicates that a nonconforming condition has been found, it must be evaluated. The logic process of evaluation for both availability and efficiency is considered in chapter 1. This chapter outlines avenues the maintenance engineer should explore in deciding what corrective action needs to be taken. There are four decisions that can be reached. In some circumstances the decision is relatively simple, and is in fact obvious. In other situations, a decision is made based on the probability of failure, the possible cost of repair, and ultimately, the reparation of consequential damages that are the result of not taking corrective action. These four options can be considered: xxii • scrap and replace • repair • rework • accept-as-is Preface Of these decisions, possibly the most difficult and potentially most controversial is the latter—accept as is—a disposition that allows a component to return to service with no effort made to correct the nonconforming condition. There are two reasons for reaching and deciding upon this course of action: • There is little need to make any corrections. To make them will add no or marginal improvement to unit performance, and the condition will not place the unit at risk • The cost of replacing, repairing, or reworking cannot be justified. This is often a judgment call on the part of the engineer and can only be made if he or she is aware of any risks involved Such a decision should not be made as a desperation measure. The risks, if any, should be fully evaluated. The options and the probability of failure—from an extended outage to operation—must be fully considered. Therefore, the evaluation process can be a complex one. Occasionally, the solution is self-evident—such as when partial failure has occurred or when excessive damage exists. The most difficult decisions are those related to suspected damage or deterioration, and those for which it is difficult to determine the cause. In these instances of uncertainty, mature judgment is required, together with knowledge of the operating and maintenance history of the unit. This knowledge should help in the evaluation. The information in this book can also provide confidence in the selection of the final disposition. The availability of replacement parts, special skills, and tools will often influence which decision is reached. Care must be exercised to ensure that availability or non-availability of replacement parts does not force the owner/operator into a decision ultimately xxiii Turbine Steam Path Maintenance and Repair—Volume Two causing more expense and increasing the overall risk level to an unacceptable degree. Often, alternatives to these potential solutions are available. Some may degrade a unit’s rating or impose other restrictions in terms of maximum output or the time for which a unit can be operated. The compromise correction is ultimately more acceptable over the short-term, while the owner/operator arranges for a more palatable long-term solution. William P. Sanders Richmond Hill, Ontario, Canada August, 1999 xxiv TABLE OF CONTENTS List of Acronyms . . . . . . . . . . . . . . . . . . . . . .ix Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . .x Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . .xii Acknowledgements . . . . . . . . . . . . . . . . . .xxv Chapter 7—Operating Damage Mechanisms and Refurbishment Techniques for Stationary Components Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Stationary Blade Row Geometry . . . . . . . . . . . . . . . . . . . .15 Operating Phenomena Affecting the Stationary Blade System . . . . . . . . . . . . . . . . . . . . . . .36 Diaphragm Vane Repair Methods . . . . . . . . . . . . . . . . . . .59 Determination of Stage Discharge Area and Angle . . . . . . .88 The Computation of Adjustments . . . . . . . . . . . . . . . . . . .96 Diaphragm Thermal Distortion . . . . . . . . . . . . . . . . . . . .121 Repair Methods for the Diaphragm Sidewalls . . . . . . . . . .133 Correction of the Diaphragm Inner Web . . . . . . . . . . . . .141 Damage to the Outer Rings . . . . . . . . . . . . . . . . . . . . . . .148 Weld Repair of the Horizontal Joints . . . . . . . . . . . . . . .153 Stationary Blade Damage . . . . . . . . . . . . . . . . . . . . . . . .154 Components of the Casings . . . . . . . . . . . . . . . . . . . . . . .162 Casing Operating Problems and Repair Methods . . . . . . .174 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199 v Turbine Steam Path Maintenance and Repair—Volume Two Chapter 8—Refurbishment Techniques for Rotating Blades Introduction . . . . . . . . . . . . . . . . . . . . . Steam Path Cleaning . . . . . . . . . . . . . . . Blade Inlet Edge Erosion Damage . . . . . Moment Weighing of Refurbished Blades Erosion Shield Cracks . . . . . . . . . . . . . . Blade Trailing-Edge Erosion . . . . . . . . . . Solid-Particle Erosion by Oxide Scale . . Erosion Resistant Coatings . . . . . . . . . . . Solid-Particle Peening . . . . . . . . . . . . . . Massive Particle Damage . . . . . . . . . . . Corrosion Effects . . . . . . . . . . . . . . . . . Rotating-Blade Refurbishment . . . . . . . . Water Induction . . . . . . . . . . . . . . . . . . Fretting Corrosion . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201 . . . . . .204 . . . . . .206 . . . . . .242 . . . . . .254 . . . . . .255 . . . . . .257 . . . . . .264 . . . . . .271 . . . . . .273 . . . . . .281 . . . . . .291 . . . . . .303 . . . . . .304 . . . . . .305 Chapter 9—Damage Mechanisms Arising from Operation and Refurbishing Techniques for Rotating Components Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .309 The Rotating Components . . . . . . . . . . . . . . . . . . . . . . . .311 Coverband Damage, Repair, and Refurbishment Methods .362 Tie Wires Damage, Repair, and Refurbishment Methods . .410 Fusion Techniques for Rotating Blades and Stage Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . .427 Common Rotor Damage Mechanisms . . . . . . . . . . . . . . .433 Bends Induced in the Turbine Rotor . . . . . . . . . . . . . . . . .464 Blade Root Steeples and the Wheel Rim . . . . . . . . . . . . .480 Corrective Action for Rotor Rim Damage . . . . . . . . . . . . .498 Rotor Weld Repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . .508 Considerations of the Weld Repair Process . . . . . . . . . . .527 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544 vi Table of Contents Chapter 10—Seals, Glands, and Sealing Systems Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549 Steam Path Seals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .551 Functions of the Steam-Sealing System . . . . . . . . . . . . . .557 Steam Leakage Through Labyrinth Seals . . . . . . . . . . . . .559 Quantifying Labyrinth Leakage (Applying the Method of Martin) . . . . . . . . . . . . . . . .567 The Economics of Seal Maintenance . . . . . . . . . . . . . . . .595 Forms of the Seal Knife Edge Discharge Coefficients . . . . .604 Form of the Gland Rings . . . . . . . . . . . . . . . . . . . . . . . . .618 Forms of the Seal Strip and its Trimming . . . . . . . . . . . . .627 Insertion and Securing of Seal Strips . . . . . . . . . . . . . . . .630 Seal Strip and Gland Ring Materials . . . . . . . . . . . . . . . .641 Gland System Operating Problems . . . . . . . . . . . . . . . . .643 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .651 Chapter 11—Quality Assurance for Replacement and Refurbished Steam-Turbine Components Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .653 Responsibility for Quality . . . . . . . . . . . . . . . . . . . . . . . .657 Definition of Quality . . . . . . . . . . . . . . . . . . . . . . . . . . .658 Definitions of Performance . . . . . . . . . . . . . . . . . . . . . . .660 The Design Specification . . . . . . . . . . . . . . . . . . . . . . . .663 Reverse Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . .666 The Quality Assurance Program . . . . . . . . . . . . . . . . . . .677 The QA Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .679 The Engineering Review . . . . . . . . . . . . . . . . . . . . . . . . .680 The Responsibility and Administration of a QA Program . .683 The Inspection and Test Plan . . . . . . . . . . . . . . . . . . . . . .688 Purchaser Assurance of Quality . . . . . . . . . . . . . . . . . . .689 Product Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . .690 Nonconforming Situations . . . . . . . . . . . . . . . . . . . . . . . .699 Available QA Program . . . . . . . . . . . . . . . . . . . . . . . . . .704 The Machining of Turbine Components . . . . . . . . . . . . . .705 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .710 vii Turbine Steam Path Maintenance and Repair—Volume Two Chapter 12—The Manufacture and Inspection Requirements of Steam Turbine Blades Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .711 Radial Alignment of Rotating Blades . . . . . . . . . . . . . . . .712 Blade Manufacturing Techniques . . . . . . . . . . . . . . . . . . .723 The Blade Manufacturing Processes . . . . . . . . . . . . . . . . .726 Profile and Cascade Tolerances . . . . . . . . . . . . . . . . . . . .749 Profile and Placement Errors . . . . . . . . . . . . . . . . . . . . . .765 Passage Swallowing Capacity . . . . . . . . . . . . . . . . . . . . .770 Special Processes Applied to the Vane . . . . . . . . . . . . . . .774 Blade Root Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . .775 Factors Influencing Blade Pitch Errors . . . . . . . . . . . . . . .803 Requirements to Accommodate Stage Hardware . . . . . . .812 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .813 Appendix—Thermodynamics and the Mollier Enthalpy-Entropy Diagram for Water/Steam Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Physical Properties of Water/Steam . . . . . . . . . The Gas Equations . . . . . . . . . . . . . . . . . . . . . . . . The Heating and Expansion of Steam . . . . . . . . . . . The Entropy of Steam . . . . . . . . . . . . . . . . . . . . . . Reversibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Steam Properties and Diagrammatic Representation The Basic Power Cycles . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii . . . . . . . . . . . . . . . . . . . . .815 . . .818 . . .831 . . .833 . . .852 . . .857 . . .859 . . .873 . . .883 Chapter 7 Operating Damage Mechanisms and Refurbishment Techniques for Stationary Components INTRODUCTION The stationary components of the steam path are not subject to the same level of stress as a consequence of rotation and centrifugal loading. However, they can still be in a high temperature/high pressure environment and will therefore be subject to loads sufficient to affect their operating life. In addition, the alignment these components are able to maintain relative to the rotating components during their operating life can be affected by steam conditions and various operating characteristics. When some form of deterioration is found in stationary components (and possibly progressing to an unacceptable level), it is necessary to evaluate the situation and to take corrective action. To undertake such correction, procedures must be developed. 1 Turbine Steam Path Maintenance and Repair—Volume Two It is necessary to develop and assess suitable procedures for both monitoring and correction so that at each outage the various affected components can be examined, critical dimensions recorded, and a non-destructive examination (NDE) of critical regions undertaken. These condition reviews should be an integral part of all outages. Any nonconforming condition must be monitored so the condition can be corrected when deterioration has occurred to the extent the unit cannot be returned to service without continuing to operate at risk, or with a significant reduction in the operating efficiency. Two major stationary components possess conditions that must be examined because corrective action is most often necessary—the casing (including the gland seal housings) and the diaphragms, or stationary blades. (These stationary vanes are often referred to as partitions.) Both components can be subject to pressure, temperature differentials, and transients sufficient to cause distortion. In addition, these components will have steady stresses developed in them as a consequence of pressure differentials. The engineer responsible for turbine maintenance must establish programs for monitoring stationary components. With stationary components there is unlikely to be dramatic failures similar to those associated with rotating components. However, the consequence of stationary component deterioration can be just as damaging in terms of forcing the unit from service and the costs and delays of correcting the situation. In determining what should be monitored, the equipment manufacturer will define the basic requirements of the individual components. However, these can be summarized as follows. Diaphragms are designed to carry and locate the stationary blade rows within the steam path (see chapter 2 in vol. I). Diaphragms can have large pressure differentials developed across them. They also develop a torque on the individual vanes that will rotate them in their circumferential locating slot if not constrained by suitable keys and pins at the horizontal and vertical positions. Diaphragms and stationary blade rows operate at high temperatures for considerable 2 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components periods and are supported and located in the casing only at their outer diameters. Diaphragms and stationary blade rows also have a strength, or support, discontinuity at the horizontal joint. Note: Diaphragms and stationary blades perform the same function. By definition, diaphragms comprise an outer ring that locates in the casing, a stationary blade row, and an inner ring or web. The stationary blade row normally comprises individual blades that locate in the casing or blade carrier directly. The diaphragms are normally used throughout the steam path of an impulse design unit, whereas the stationary blade rows are normally inserted into the casings of the high and intermediate pressure sections of reaction designed units. The basic design of these two stages was considered in chapter 2. The most commonly encountered (and most damaging) mechanisms in diaphragms and stationary blades are the following: • Compounds deposited on surfaces of steam path components that are normally inert may be composed of materials that can become corrosive under the appropriate environmental circumstances • Material loss associated with solid-particle erosion caused by exfoliated boiler scale • Damage caused by impacts with solid particles, either carried into the turbine or (more likely) originating within the blade system, steam chests, and/or valves • Damage induced by water formed by condensation and then accumulated into larger droplets and deposited upon the surfaces of the steam path components. This water forms a surface layer, flows through the steam path, and will cause various forms of material loss 3 Turbine Steam Path Maintenance and Repair—Volume Two • Distortion that causes the diaphragm to modify from a circular to an elliptical form. This adjustment may cause the horizontal joint to either open or close. This effect, if observed, is monitored by measuring the diaphragm bore diameters at the horizontal and vertical centerlines • The high-pressure differential that exists across the vanes and inner web can cause an elastic deformation of the diaphragm. At high temperatures there can also be a plastic deformation, which will cause a dishing. This is normally checked by suitable “drop checks” The casing is the main structural component of the turbine and contains the rotating components. The casing also contains and provides alignment for the stationary steam path components. The following common deteriorating mechanisms can affect the turbine inner and outer casings as a consequence of the combined effects of temperature and pressure: 4 • The casing can distort from the true circular form. The casing must maintain its circular form along the length of the axis. Distortion in the horizontal or vertical direction can affect concentricity of the stationary blade rows. This distortion is more significant if the stationary blades are mounted directly into the casing. In the instance of a diaphragm type construction, casing distortion may not necessarily affect the steam path concentricity. However, it could affect the vertical and horizontal position of the diaphragm • There is often a tendency for a casing to “hump” and assume an “upward bow” in the cold condition. The casing must remain flat at its horizontal joint • The horizontal joint must remain closed and provide an effective seal against steam leakage that would allow the steam to bypass the stages or even complete sections Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components • As a consequence of large temperature swings, the casing will be subject to the effects of “low-cycle fatigue.” This will most often introduce cracks into the high temperature regions at small fillet radii, where considerable stress can concentrate as these temperature changes occur Lower stress levels in the stationary components lend themselves more readily to refurbishment. Many of the deteriorating situations that are encountered are readily correctable. Stationary blade definitions Chapter 2 provided definitions for the various portions of the rotating blades. In this chapter, repair techniques are discussed for the damage and repair of the stationary blade rows. It is therefore appropriate that similar definitions be provided for the stationary components. The names given to different portions of the steam path elements differ from manufacturer to manufacturer, which makes it difficult to be consistent, and can cause some level of confusion when describing various aspects of both damage and repair. In this text, the following definitions will be used: Diaphragms Diaphragms are manufactured by a number of processes, and have the primary function of expanding the steam and guiding it into the following row of rotating blades. In its simplest form, it comprises an outer ring, a row of stationary blades, and an inner ring designed to provide a pressure barrier between the stationary blades and the rotor. The major components [Fig. 7.1.1(a)] follow: An outer ring. This ring is located in the casing or blade carrier. Horizontal joint keys and crush pins are used to hold the diaphragm in both vertical and horizontal alignment to the rotor and achieve the alignment requirements outlined in chapter 2. 5 Turbine Steam Path Maintenance and Repair—Volume Two Root Attachment Outer Ring or Ledge Outer Ring Profile (b) Tip Seals Vane or Partition Tip Diameter 'Dt' Profile Inner Band or Ring Skirt Web or Inner Ring Root Diameter 'Dr' (c) (a) Shaft Seals Shaft Seals Fig. 7.1.1—Definitions of the stationary blade row components. K L w ψ2 Br Bs ψ1 D D Dp R Outer ring s s r r Fig. 7.1.2—A water catcher half produced as part of the outer ring on a diaphragm. 6 Vane or Partition (d) Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components This outer ring can also be extended to provide radial seals above the rotating blade row [Fig. 7.1.1(b)]. It is also possible (in the water region) that a portion of the water catcher can be built into the outer ring (Fig. 7.1.2). The vane. The vane expands the steam and directs it into the following rotating blade row. As in the rotating blades, the profile forms the pressure surface of one expansion passage and the suction surface of the adjacent passage. Vanes are subject to the same requirements of dimensional control and surface finish as the rotating elements. The definitions used to describe the vane profile are shown in Figure 7.1.3. In many stages of the high and intermediate (reheat) pressure expansions, the radial height of the vane is not sufficient to justify other than a cylindrical vane of constant profile. Other, longer stages have a profile that “varies” with radial height, although on many stages the profile remains the same but of reducing chord (Fig. 7.1.4). Inlet nose Chord Axial width Pressure face Suction face Discharge tail Fig. 7.1.3—Definitions of portions of the vane profile. 7 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.1.4—A stationary vane of constant profile, but with a reducing chord. The means of attachment of the vane to the outer and inner sidewalls is dependent upon the method of manufacture. This vane is often referred to as the partition. To remove as much kinetic energy as possible from the expanding steam, the previous rotating blade row would have been designed to discharge the steam as close to axial as possible. Therefore, the profile of the stationary blades is normally axial—i.e., the inlet angle “α0” is equal to 90 degrees. An inner ring or web. The inner web provides the pressure barrier across the stationary row. This web is also designed to carry radial seals at its inner diameter to minimize the quantity of steam that leaks past the stationary blade rows. Similarly, this web will often have a radial seal produced just below the root diameter “Dr” [Fig. 7.1.1(c)]. These strips are intended to limit the amount of steam that re-enters the main steam flow, which can introduce efficiency losses. This seal helps force the steam to flow through pressure balance holes produced in the wheels of the rotor. 8 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components The seal system produced between the inner surface or the inner web and the rotor are designed to minimize the leakage quantity that bypasses the stationary blade row, as this steam does not expand through the blade rows and therefore produces no power. Stationary blades. The stationary blades are mounted directly into the casing or blade carrier and perform the same function as the stationary blade row of the diaphragm. Figure 7.1.1(d) shows a single blade element that has the same radial height as the diaphragm vane of Figure 7.1.1(a). The tip diameter “Dt” and the root diameter “Dr” are the same on both steam path elements. This design is normally used on the high and intermediate pressure rows of a reaction machine in which there tends to be limited axial space and therefore the root block performs no function other than to secure the blade in its carrier. Caulking is also normally used to secure the blade. The inner band or ring can be integrally produced with the vane and root portion, or it can be attached by riveting using the same methods as on the rotating blade coverbands. This inner ring will also normally carry radial seals to minimize steam leakage. Nozzle plate. In addition to the diaphragm and stationary blades, the nozzle box is that portion of the main unit structure into which the steam enters the steam path via inlet pipes and valves. After entering the steam chamber, the steam flows around an inner belt (or a portion of a belt) and enters the first row of stationary blades. This row of stationary blades is contained in a structure called the nozzle plate. There are various methods of forming this steam chamber and nozzle plate. The methods used depend upon the inlet steam conditions and the method of controlling the unit (either throttle- or nozzle-controlled). In a throttle-controlled unit, the basic inlet is a continuous belt and steam flows around the complete 360 degrees. In 9 Turbine Steam Path Maintenance and Repair—Volume Two the nozzle-controlled unit, individual portions of the inlet are connected to individual stop and control valves. These inlet belts can form a portion of the casing (Fig. 7.1.5) or a self-contained chamber (Fig. 7.1.6), which are assembled separately, aligned, and welded into the casing. Fig. 7.1.5—Steam admission nozzle boxes. With this design the nozzle boxes are an integral part of the casing. The nozzle plate is the structure that carries the first row of stationary blades. This plate is designed to receive steam as it enters the high-pressure section delivered from the boiler. Steam enters from the main inlet pipes, passes through the main stop and control valves, and enters an inlet belt that passes around the rotor. This chamber is 10 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.1.6—A nozzle chamber with segmented steam-admission ports. The inlet pipes are connected to the nozzle block, which is a selfcontained forging. constructed in such a manner that no inner web is required on the nozzle plate. Therefore, the nozzle plate becomes a device attached to the steam chamber at both its inner and outer diameters. Depending upon its design, a 360-degree nozzle plate may span the upper and lower halves of the inlet and discontinue only at the horizontal joints. In smaller rated units, the nozzle plate may span 11 Turbine Steam Path Maintenance and Repair—Volume Two only the single (normally the lower) 180-degree half of the inlet. In some nozzle-controlled unit designs, it may also be produced as a number of individual segments, each admitting steam from a single valve and individual chest. When steam enters the main steam inlet belt and flows around the inlet perimeters (the annulus arc fed by an inlet pipe), it enters the first row of nozzles or stationary blades. This stage is normally designed for a high-pressure drop and there is, therefore, a considerable temperature and pressure gradient across it. Basic considerations biasing the designer to select this high-pressure drop for this stage include the following: 12 • This stage should not experience leakage past the blade row because of its method of attachment to the nozzle box. This means a large pressure drop can be accommodated without increasing the leakage potential • If a large pressure drop is designed into this first stage, and the steam is contained within a separate nozzle box, temperatures and pressures to which the casing is exposed are reduced to conditions at the first stationary row discharge. This reduces the pressure/temperature duty on the casing • These stages are designed with large diameters and generally possess a velocity ratio of blade velocity/steam velocity that results in a large pressure drop. If the first stage is a two-row wheel (a Curtis stage), the velocity ratio will be of the order 0.25-0.32. Therefore, the enthalpy drop and the pressure and temperature drop will be large, reducing the duty on the casing considerably. This will also require fewer stages (after the control stage) to complete steam expansion Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Casing definitions The turbine casing—essentially a cylindrical vessel—is the main stationary component of each turbine section. It encloses the rotating portions of the unit and locates the stationary blades either directly or through the location and support of an inner casing, which itself carries the stationary blades and/or diaphragms. The principle component of the casing are the shells, which provide the mechanical strength and carries and locates other elements such as packing heads, diaphragms, and the inner casing or blade carriers. The casing is normally split along its centerline at the horizontal joint. This is to facilitate assembly and provides access to the rotor and internal stationary portions of the unit. The shell halves are connected through a bolted flange at the horizontal joint. It acts to contain the steam and maintain its connection to the steam path blades. Casings may also provide locations for internal-gland packings or portions of the steam seal system. They could be equipped with internal moisture collection and drainage systems if moisture is present in the steam. In the case of minor failures, the high-pressure shells should also be capable of containing missiles generated from the rotor. Both the upper- and lower-half casings can be arranged to provide connections for welded stub pipes. External pipes are connected to these stubs, allowing steam to be extracted for regenerative feed heating or other cycle or process use. Such steam is extracted from the main steam flow. Other pipes, used to introduce or extract steam to other parts of the cycle, may also penetrate the casing. Usually, pipes connected to the upper-half casings are joined through flanges or other devices. This allows quick disassembly at outages and reconnection without the use of any form of heating or metalfusion techniques. 13 Turbine Steam Path Maintenance and Repair—Volume Two Special provisions are necessary in the casing to admit the high-pressure, high-temperature steam and to make provision for the differential expansion that occurs between the various portions of the shells. Such differential expansion occurs because of different temperatures (temperature gradient) along the axial length of the casing, and also because of the different rates at which the various parts of the turbine heat and cool with main steam temperature changes (see chapter 2, vol. I ). For doubleshell construction, it is necessary for the main inlet pipes to pass through the outer casing and introduce steam to the main steam inlet belt or nozzle box. High-pressure casings are normally supported at each end through arms that are produced integrally with and extend from the casing to pedestals located adjacent to and between the casings or sections. Transverse and axial keys are used to maintain alignment of the shells at these pedestals. Usually such keys have been hardened by nitriding and are located on the bottom vertical centerline to ensure correct alignment is maintained at all loads and during transient operating conditions. Low-pressure casings are designed to both contain the steam and also to minimize the “inward leakage” of air when exhaust pressure is sub-atmospheric. Low-pressure turbine exhaust creates large volumetric flow, so these low-pressure casings are usually produced by fabrication; because such fabrications are not structurally strong, it becomes necessary to support them for their entire perimeter at their horizontal joint or at a similar location below this joint. Turbine casings comprise a number of individual components which, when assembled, allow the unit to operate safely, achieving high levels of reliability and efficiency. Principal among these components are the following: 14 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components • Shells—the main structural components produced by casting or fabrication—in some designs, by a combination of both— dependent upon the experience and preference of the designer • Shaft end packing head—is attached to the shells and carries the gland rings, which are located where the rotor passes through the shells. Gland rings minimize the outward leakage of the steam or the inward leakage of air • Inlet section—the inlet to the steam path must be designed to allow free access of the inlet pipes as they transport steam to the nozzle box. They must also minimize steam leakage that will occur at those locations and must be designed to permit movement between the inlet pipes and the main body of the shells • Explosion diaphragm on low-pressure sections—in the lowpressure exhausts there is a need to provide for the rapid removal of steam from the internals of the casing in the event there is a sudden and high rate of pressure increase due to some transient condition • Diffuser at exhaust from the last stage—in an effort to maximize the energy extracted from the steam, the final rotating blade is arranged to exhaust into a diffuser, normally produced as part of the casing fabrication STATIONARY BLADE ROW GEOMETRY The primary function of the stationary blade row is to provide controlled expansion of the steam from a high energy level to a lower one; convert thermal energy to kinetic energy, and direct the 15 Turbine Steam Path Maintenance and Repair—Volume Two resulting high-velocity steam jet into a row of rotating blade elements. These represent the primary functions. However, there are other requirements that must be considered. The stationary blades must also be able to accept steam from the previous rotating blade row without incurring high incidence losses and then redirect this steam from the inlet direction through the turning angle “θ” to discharge it without incurring excessive profile losses in the process. Note: The turbine is designed to contain a number of stages with a specific ratio of “U/Co” (blade tangential velocity to steam adiabatic velocity). The enthalpy drops in each stage establish the stage inlet and discharge pressures, and the stationary blades are designed to pass the required quantity of steam within those inlet and discharge pressures. Stationary blade two-dimensional considerations Consider the basic convergent nozzle shown as Figure 7.2.1. Figure 7.2.1(a) shows development of the passage shape through which the steam must expand, from an initial width of “Wi” to a final or discharge width “Wd.” Figure 7.2.1(b) shows the developed width sensed by the steam as it flows through and between the passage walls. If the nozzle shape formed by these passage walls is maintained as a constant at various radial heights, the discharge area at any position “g-g” can be found as the product of width “g-g” and the radial height “H.” This nozzle form would possibly be suitable if the steam were to enter and discharge without a need to be deflected from an incoming angle to a different angle at discharge. However, it is also necessary for the nozzle to deflect the steam from an incoming angle (most often an angle that is close to the axial) to an angle which, to achieve maximum efficiency, should be as small as possible (consis- 16 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.2.1—Development of the passage shape in a convergent nozzle, with no turning angle. tent with providing adequate nozzle discharge area and enabling the steam to enter the rotating blade row). Consider the mean of the passage width line “k-m” shown as Figure 7.2.2. Steam enters the nozzle at an angle that is in a substantially axial direction “A” and discharges at an angle “α1”, which is small and inclined to the tangential direction “T.” It is also required that this nozzle have a discharge width “Wd” to provide the required design discharge area. The details of the profile can be established in the following manner—an axial width “Y” is set consistent with strength requirements, and an inlet flow width “Wi” is calculated. It is then necessary to establish a profile shape that will provide both the turning angle “θ” and the discharge width “Wd” that are required. 17 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.2.2—The steam turning angle of a nozzle passage. The actual determination of a profile shape is often undertaken by trial and error methods. Engineers in this task gather considerable experience. Computerized systems of detailed design and evaluation have somewhat simplified the assignment, and the engineer is able to achieve greater optimization in establishing suitable geometries for the vane profiles. In examining the requirements shown in Figure 7.2.2, several factors are worth attention: 18 • The discharge width “Wd” will, as drawn, be formed as the gap between two adjacent profiles. Figure 7.2.3 shows the discharge tail between two such surfaces • Because the vane discharge tail cannot extend beyond the discharge edge “a-a,” the actual throat is formed as shown at the discharge point Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.2.3—Details of the vane discharge tail. Fig. 7.2.4—The developed steam expansion passage, forming a convergent passage between two stationary profiles. 19 Turbine Steam Path Maintenance and Repair—Volume Two • The actual discharge angle of the vane will be somewhat different from the steam angle, and there is some small deviation (angle “Γ” as shown). The extent of this deviation depends upon the curvature of the discharge tail suction surface. The discharge tail has some small thickness (“b”) Figure 7.2.4 shows two profiles and details of the passage shape formed between them, including line “k-m” and the passage width variation from “Wi” at inlet to “Wd” at discharge. Stationary blade three-dimensional considerations The row discharge area is established as the product of the mean (or effective) throat and the radial height of each stationary blade row opening. Consider several forms of vane and the variation of throat “Wd” along the radial height. Vanes of constant section. Such a vane is shown in Figure 7.2.5. The profile is established and used along the entire radial height “H.” This type of vane is used for those stages with small radial height-to-mean Fig. 7.2.5—The radial variation of discharge opening along a vane of constant profile. 20 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components diameter ratios (H/Dm) of less than about 0.20. In such stages there is a relatively small increase in pitch from the root “Pr” to the tip “Pt.” An examination of the steam passages shows that the throat (now given the more familiar symbol “O”) is formed on a curved tail of the suction face. Because this is the smallest flow section it is clear that the throat, and therefore the discharge angle, changes along the radial height. This is an acceptable condition and can be allowed for by the designer, who determines the discharge area “Ad” by integration along the radial height. (This is discussed in a later section.) Vanes of constant but reducing profile. A similar profile used for vanes of larger radial height-to-mean-diameter ratios employs a vane of constant profile, but one that changes its chord “C” and other principal dimensions along the radial height. Such a vane-stacking diagram is shown in Figure 7.2.6. Five sections are shown at equally spaced radial locations from the tip “t” to the root “r.” In this figure the root section is the narrowest at “Wr” and the tip section “Wt” is the widest where “W” is the axial width. Fig. 7.2.6—Vanes of constant profile, but reducing chord “C”. 21 Turbine Steam Path Maintenance and Repair—Volume Two In this design the throat “O” is formed on approximately the same, but scaled position of the suction surface, maintaining the same ratio of “O/P” and, therefore, a constant discharge angle along the total radial height. The discharge area “Ad” can be determined using the same method as in Figure 7.2.1. At each radial location the vane has the same setting angle (“ ξ ”). Vanes of twisted profile. Modern machining techniques allow stationary vanes to be produced to a true vortex form, at a cost not significantly higher than many other types. Fig. 7.2.7—The vortex vane, with a profile of varying section. 22 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Figure 7.2.7 shows a typical vortex or twisted-type vane, where the inlet nose has been adjusted to accommodate the inflowing steam from the previous rotating blade row at different angles (“Vt,” “Vm,” and “Vr”). The total discharge area is found in the same manner as the previous forms, and the throat characteristic can vary along the length of the vane to accommodate design requirements. Vane tilt The radial disposition of the vane is selected to improve efficiency, both by maximizing the energy conversion process and also by minimizing secondary losses that could occur by introducing vortices and turbulence, both of which induce losses. The vane can be tilted in both the tangential and axial directions to improve the performance of the stage. Tangential tilt. In many longer stages the stationary vanes are given a tilt in the tangential direction of rotation. This is done to suppress the radial component of flow and to guide the steam flow into the following rotating blade row in a more axial direction. This effect is shown in Figure 7.2.8, where a tilt angle “ Ψ ” has been used. This tilt “ Ψ ” often requires some special considerations at the horizontal joint, and while causing some small increase in manufacturing costs, it is more than offset by the savings available in fuel costs. In the diaphragm half shown in Figure 7.2.8, the true horizontal joint “H-H” is shown along with the actual joints (“h-h”), which enable the vanes to be carried in the outer and inner rings. Since both halves are symmetrical, the joints close. Axial tilt. The stationary vanes will often increase in chord (and therefore, axial width) as their position from the inner root diameter increases towards the outer root diameter. To minimize the axial gap between the stationary and rotating blade rows consistent with maintaining adequate clearance at all radial positions, the stationary 23 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.2.8—Tangential tilt ‘ ψ’ in the stationary blade row. blade is given an axial tilt. If a minimum axial clearance “Ca” is required, the tilt shown in Figure 7.2.9 can be used. This tilt is shown as a forward tilt “ ζ” from the true radial position “R-R” towards the root of the blade. This tilt allows an axial gap (“Aa”) to be maintained between the discharge edge of the stationary blade row and the inlet to the rotating blade. Because there are no centrifugal effects on the stationary blades, this tilt will not induce any additional stresses in the vane. This tilt and the proximity of the edges will also allow minimum laps “Lo” and “Li” to be used. Compound radial tilt. For large radial-height blades with high discharge mach numbers and a large radial flow component within the axial flow, the vane can be given a forward curvature to suppress this radial flow. This suppression causes an advantageous mass flow distribution. The outer flow portion of the vane is given an advanced forward curve. This blade then promotes flow through the center region of the blade path, which is more efficient. 24 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.2.9—The axial tilt in a diaphragm. Stationary vane profile details The profile is required to deflect the steam through a suitable angle from inlet to discharge and to provide a discharge area sufficient to ensure energy conversion in accordance with the requirements of the designer. The profile’s form is established by several methods, including hand and computer calculation and detailed design layout to ensure that the form of the passage is acceptable. There are certain details of the profile that the designer considers in order to maximize energy conversion efficiency. These include the following: 25 Turbine Steam Path Maintenance and Repair—Volume Two The inlet nose. The steam entering the stationary blade row is moving with a relatively low velocity because kinetic energy has been extracted from it in the previous rotating blade row. In the majority of stages it is also entering in a principally axial direction. This means that the inlet nose is at about the 90-degree position. This is not always so; the stationary vane form must be selected from the requirements defined by the velocity triangle that is produced in the stage design process. Various forms of nose can be used. The intent of the nose is to accept the steam and cause it to divide and flow down either the pressure or suction faces of the profile. The radius and immediate form of the nose is important, as it should introduce minimal flow disturbance likely to cause the boundary layer to separate. Figure 7.2.10 shows a typical nose form. It is produced from an inlet radius “n” blending to a radius “a,” and then to “c” on the suction face and a radius “b” on the pressure face. This type of inlet is used in many existing stages in current operation. A major consideration with this form of inlet is that at the points of blending from one radius to another—or at any other location—there should be no sudden change in the radius of curvature, as this can promote boundary layer separation. Fig. 7.2.10—Details of the inlet nose. 26 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Discharge tail. The discharge tail thickness “b” (Fig. 7.2.11) should be as thin as possible, recognizing that there can be high loads induced there, and that it is necessary to maintain a nominal thickness to avoid crack development. On some stages—where the possibility of material loss by erosion occurs—it is possible to increase the tail thickness to extend the life of the vane. An aero/thermal design requirement is to maintain the discharge tail relatively straight. This helps prevent premature separation of the boundary layer. Excessive curvature on the tail suction surface can create a large “wake” region downstream of the row (Fig. 7.2.11), thereby inducing losses in the stage. The separation angle “Γ” must be maintained at a minimum. It is probable the flow will be able to sustain the boundary layer for a short distance beyond the throat (“O”), but it will not be able to continue to turn with the curvature of the tail suction face. Fig. 7.2.11—Flow separation of a curved discharge tail. 27 Turbine Steam Path Maintenance and Repair—Volume Two Surface curvature and finish. Surface curvature should be such that there is no sudden change from one radius to another. This is difficult with a profile defined from blending radii, but at the position of change, the blending should be as smooth as practical. When machining and hand polishing produce vanes, it is possible to make this transition; and by finishing the surface in the major direction of the chord, the condition of the vane can be improved considerably. The design requirement of surface finish must be observed. Stationary blade row dimensional requirements A number of different manufacturing processes are used in the production of the impulse and reaction stationary blade rows. For this reason, it is not possible to consider the requirements of each, or the details of each step that should be monitored during manufacture. This section will discuss the row geometric characteristics and consider the more important aspects of each. In the impulse stage design, about 90% of the stage heat drop is converted in the stationary blade row to velocity energy. For this reason there is every incentive for the manufacturer to control the quality of the stationary blades and ensure engineering requirements are achieved in terms of dimensional conformance, surface finish, and structural integrity. Attaining design requirements to a large extent depends upon the means of manufacture. Design requirements are normally defined with the manufacturing process to be used and practical limitations considered in the specification of acceptability. While specified tolerances on each of the major parameters may vary, the following (as shown in Fig. 7.2.12) are considered the most essential in achieving an acceptable product. 28 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.2.12—Stationary vanes and the principal control dimensions. The blade opening (throat) “O.” The throat and its variation along the radial height of the vane are fundamental in establishing the efficiency of the blade row. Consistency from opening to opening means an effective opening (“Oe”) is necessary to establish the individual aperture areas and, in total around the row, to achieve the stage discharge area. When a blade row is manufactured for a diaphragm, or when individual blades are assembled in a casing, the opening will vary from outer to inner section. It is therefore necessary to monitor and, if required, to adjust these individual values to achieve the correct area. The vane pitch, “P.” The pitch between blades is established in terms of the diameter being considered and the number of blades in the row. The actual pitch achieved in any row and its variation from passage to passage depends upon the method of manufacture. For individual body blades, there should be little variation since the pitch will depend upon the machined root block, which can be held within very tight tolerances. There is of course the possibility of vane lean, which will cause a cumulative error along the radial 29 Turbine Steam Path Maintenance and Repair—Volume Two height. This can often be corrected by means of minor adjustment or positioning with an attached coverband. For cast, welded, and fabricated pieces, locating the individual stationary vanes from some marked position and attaching them to inner and outer rings usually achieves the pitch. The number of pitches is selected to accord with the design requirements, achieve discharge area and angle, and complete a half circle (180 degrees). The final and most important measure is the cumulative pitch around the half diaphragm. This is measured and controlled to ensure joint blades are in their correct position so pitch tolerances at the horizontal joint are achieved. The ratio “O/P.” The second most important function of the stationary blade rows is to assure the steam is discharged at the correct angle into the rotating blade row. For the rotating blade the discharge angle was defined as “β2” and determined as the ratio of opening to pitch. Similarly, for the stationary row, the discharge angle “α1” can be defined as: Therefore, this ratio “O/P” is critical to the total stage performance. It is possible for the area to be correct, but if the variation of throat is outside the design specification, this will modify the discharge angle variation along the radial height and will also modify the radial steam flow distribution. The vane-setting angle, “ ξ.” The blade as adjusted to its correct setting angle “ ξ ” is designed to provide an expansion passage sufficient to produce the correct throat—the desired ratio of “O/P” and the width “W” (Fig. 7.2.13). However, any change in the setting 30 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components angle will modify both the throat and the width “W.” The modification of the throat will modify the ratio “O/P.” While some level of error can be tolerated, this must be controlled within design tolerances. Normally, variation in the width either positive (“+dW”) or negative (“-dW”) can be accepted in a diaphragm, but must be considered more carefully in a stationary blade assembled to the casing in a reaction unit. These blades may have a carefully controlled axial clearance that may not be able to be compromised. Fig. 7.2.13—Showing the effect of varying the vane setting angle ‘ ξ’. Vane tilt angle, “ Ψ.” It is also possible with some forms of construction to adjust the vane tilt angle “Ψ” (Fig. 7.2.8) from the true radial line. This tilt causes a radial inward pressure on the steam and reduces the radial flow component. It also minimizes the nozzle effect on the downstream side of the vane. It is further possible that the tilt angle will vary along the radial height for those vanes with a compound radial tilt. Inlet and trailing edge setback, “dai ” and “dao.” The blade lattice is normally built up by whatever form of construction is used so 31 Turbine Steam Path Maintenance and Repair—Volume Two that the inlet and discharge nose lie in a true tangential plane. In fact, there is some degree of tolerance in the axial shift “dai” and “dao,” allowing the vanes to be either “proud” or “recessed” on both edges. This effect is shown in Figure 7.2.12. Sidewall discharge diameters, “Dot ” and “Dor.” The sidewall diameters “Dot” and “Dor” [shown in chapter 2 as Fig. 2.12.1(a) for an impulse design and Fig. 2.12.1(b) for a reaction stage] show the importance of maintaining these diameters so the correct lap can be achieved in these stages. The requirements discussed above define the optimum form and arrangement of the stationary blade row elements. However, for various considerations (such as mechanical strength and assembly) it is not always possible to achieve these in the final design, and it becomes necessary to make compromise selections of various parameters and requirements to achieve an acceptable stationary row. These changes include the following: Vane cross-section irregularities. This section has discussed the essential nature of the vane sections—the need to ensure that both are the same from element to element, and mounted so as to ensure the expansion passages generated between them are as identical as possible within manufacturing limitations. There are however, three situations where this identical form of passage is not achieved. The design engineer is aware of this and in fact has specified the steam path so that this does not occur. These situations include the following: • 32 Nozzle box—the nozzle box is located at the inlet end of the high-pressure section, and accepts steam entering the unit from the boiler or steam source. In nozzle-controlled units this box is designed to admit steam to a selected quadrant of the entire 360-degree inlet. By design, each of these steam Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components admission quadrants is isolated from its neighbors so that the steam entering it is discharged through the nozzles into the rotating blade row over a very defined arc Therefore, at the inlet to this stationary blade row, there are walls produced at the end of each inlet passage (Fig. 7.2.14) where there are differences in the expansion passage shape. These differences are not significant in terms of passage flow characteristics, but represent a departure from the complete similarity from one passage to the next. Pe Pz Fig. 7.2.14—Showing the endwall shapes of a nozzle segment. If these normal nozzle vanes have an axial width “W,” and the nozzle plate is made marginally thicker at “Wu,” then there is a small difference (“Wu-W”) that is a short guide for the steam entering from the inlet plenum. The nozzle plate is attached to the inlet plenum; the inlet pitch to the plate is shown as “Pe.” At discharge the total pitch is “Pz,” which is equal to the sum of the individual nozzle discharge pitches. Normally “Pe” is greater than “Pz” because of the sidewall taper at the inlet ends • Extended vanes (for axial strength)—the vanes used in certain stages that have a high-pressure drop across them must 33 Turbine Steam Path Maintenance and Repair—Volume Two be evaluated for their axial deflection under the effects of pressure and steam momentum loads. If these loads become excessive, it is common to use extended axial width vanes (Fig. 7.2.15). Such vanes are used on a portion of the total in each stage and extend the axial depth of the steam path from “W” by an amount “E” to “Wu” Fig. 7.2.15—Extended section vanes, used for axial strength, within the blade annulus. The extended portion of these vanes is selected to preserve (to the greatest extent possible) the aerodynamic form of the expansion passage; it has little or no effect on the flow distribution of the steam. If, however, the incoming steam has an inlet angle significantly different from that of the extended vane metal angle, there could be some effect on the “swallowing capacity” of the individual passages • 34 Throttle controlled units at their horizontal joint—units designed for throttle control allow steam admission over the complete inlet arc on the first stage in the high-pressure section of the unit. This mode of admission removes the highimpact forces that the first stage experiences in the nozzlecontrolled units. This is why the first-stage rotating blades are normally designed for lower stress levels Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components The first-stage stationary row will still receive steam from the inlet plenum, which because of design-required discontinuities at the horizontal joint will introduce a small “dead band” at this location. In Figure 7.2.16 the dead band is shown as the tangential length “Db,” which occurs at the horizontal joint “h-h.” No steam is admitted over this small arc and therefore there is no steam load on the rotating blades. This change in steam load is sufficient that the rotating blades experience a 2/rev stimulus. Under certain circumstances this high-frequency load change can result in cyclic damage and rotating blade failure. A solution some manufacturers have adopted is to remove material from this dead band region (Fig. 7.2.16), producing only a small dead band of tangential width “w.” The removal process must be controlled so as not to introduce too much curvature onto the outlet surface, thereby producing an excessive wake. However, the net effect should be to lower the influence of these impact loads on blade life. Fig. 7.2.16—The method of reducing the effect of ‘dead band’ in a throttle controlled unit. 35 Turbine Steam Path Maintenance and Repair—Volume Two OPERATING PHENOMENA AFFECTING THE STATIONARY BLADE SYSTEM Phenomena that cause stationary blades and diaphragms to deteriorate are, in many instances, the same as, or related to, those affecting the rotating components of the steam path, i.e., many of the problems encountered in rotating blades are present in the stationary blades, as well. However, these phenomena may manifest themselves in a different form or there may be a different appearance or form of the damage. Because the stationary blade components are not normally as highly stressed, repair methods and materials are often available allowing them to be refurbished, and repairs can be undertaken on components that appear to have suffered irreparable damage. The principal factors or influences causing deterioration of the stationary blade system follow. Chemical deposition Common among problems affecting stationary blade passages is chemical deposition on the vanes and sidewalls. This deposition results in the buildup of layers of various compounds within the expansion passage. Such deposition will cause a surface roughening and therefore a frictional loss in the steam path. Although the deposits will accumulate on the stationary blading, the patterns are different from those seen on the rotating components. The probable reason for this difference is that during the passage of the steam through the rotating blades, the contaminants have a radial flow component, causing greater quantities to be deposited on the outer sections of the steam path. This is not the case in the stationary components. 36 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components The radial flow component in the steam transporting the chemical contaminants causes the main chemical deposit to occur on the underside of the coverband and outer regions of the rotating blade. This radial flow phenomenon also accounts for the higher deposits on the outer sidewalls of the stationary blades, although the stationary components have a more even flow distribution. Figure 7.3.1 shows deposits on the stationary vane portion of a high-pressure stage diaphragm as seen from the horizontal joint. Figure 7.3.2 shows deposits on a lower-pressure stage—definite accumulation patterns at the opening—and Figure 7.3.3 presents deposits that occur around the strengthening stub attaching a blade portion at the horizontal joint. Fig. 7.3.1—Showing similar deposits on a lower pressure stage. There is significant deposit in the throat region. 37 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.3.2—Concentration of deposits in the inner flow portion of the expansion passage. Fig. 7.3.3—Deposits on the strengthening stub of a horizontal joint blade. 38 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components The even deposition pattern in the stationary blade rows will result in a somewhat larger frictional energy loss, because all areas of the vane and expansion passage can be affected. There is, therefore, a greater likelihood that when a stationary blade passage is examined, it will be noted that the deposit is more evenly distributed over the entire radial length of the vane and inner and outer sidewalls. There is also a tendency for solids to precipitate from the steam near the opening of the passage, mainly onto the convex or suction surface of the stationary blade vane. When a unit is opened, the opportunity exists to blast clean the stationary blades and the inner and outer sidewalls to remove as much of the chemical deposit as possible. It is suggested that when deposits are found, samples be collected and analyzed to determine their chemical composition and establish if aggressive compounds are present. If gaps are present at the junction of the vane and sidewalls (produced by casting or welding cast-type diaphragms), no special effort should be made to remove deposits from the gap. That’s because this deposition is of more benefit than harm—it removes a flow-disrupting device, i.e., deposits in this region can help minimize the turbulence caused by steam flow into and out of the gaps. The exception to this is when chemically aggressive compounds are found in the deposits. Controlling the cleaning procedure is important, as this can damage the discharge tails of the vanes (see chapter 6, Fig. 6.4.1). This damage can result from the use of an air pressure that is too high, too hard, or too large a grit, by exposing the surfaces for too long to the effects of the grit, or from holding the nozzle too close to the vane. The procedure for blasting should be defined, calibrated, and applied under careful conditions not causing severe distortion of the vanes. The requirements for blast cleaning the stationary elements are the same as outlined in chapter 6 for the rotating blades. 39 Turbine Steam Path Maintenance and Repair—Volume Two Solid-particle impact damage Stationary components (as rotating components) are subject to solid-particle impact damage. Generally, the consequential harm associated with impact damage suffered by stationary blades is not as severe as that on the rotating blades because there is normally a larger impact velocity between rotating blades and any debris than there is between debris and the stationary blade rows. Also, the rotating blades tend to have craters formed on the inlet edge, whereas stationary blades and diaphragm vanes have a greater tendency for the damage to occur predominantly on the outlet or discharge tail. This latter area is less robust and damage can occur on either the pressure or suction face. Despite the lower impact velocities (compared to rotating blades), impacts can result in craters of various size and form, produced on the surface of the vanes, or massive damage, including significant material rupture, which requires replacement of a portion of the vane or stationary blade row. These impact craters will be a source of continuing energy loss and efficiency degradation. If large enough, such areas of damage can also be a possible source for the development of flow disturbances in the steam, which are capable of inducing vibration damage in following blade rows. Such a level of damage is not a common situation though its influence should be considered in the event of unaccountable failures after units are returned to service with uncorrected damage. Damage that occurs on the stationary blade rows requiring remedial action is most often associated with the discharge tail. The cross section of the stationary blade profile—which is thinner than the rotating blade—tends to be deformed to a greater extent than other portions of the profile. In addition, damage at the discharge tail can influence both the outlet area and angle at which the steam is discharging from the stage at that point. 40 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Diaphragms and stationary blades are often damaged by particles generated within the steam path from some mechanical rupture upstream of the stage. These detached pieces break loose and transport through the steam path, and though these are broken or chopped into smaller pieces by the action of the rotating blades, they can often rebound between the stationary and rotating parts of the unit during their passage through the stage. Such multi-impacts can cause bending and cracking of the vane and discharge-opening distortion. When it occurs, this type of damage occurs in a relatively short period of time and can often be detected, or inferred, by the change of pressure measured at various locations within the unit. If performance monitoring is carried out, this type of damage can be more readily detected because such damage will have an effect on the unit efficiency level. It is difficult to quantify this damage because there tends to be little consistency of measurable damage from one incident to another. Possibly the most suitable method of quantification is the level of repair required to return the components to a fully serviceable condition. Damage is termed light, medium, or heavy, according to the definitions. • Light—this damage can normally be corrected to an acceptable degree by hand dressing, using a file and/or emery cloth. Such correction can even include bending the discharge tail to achieve an acceptable condition to a degree. Figure 7.3.4 shows a blade row that has suffered impact damage at the outer regions for steam flow and also shows evidence of some larger impact that has deformed the discharge tail. Figure 7.3.5 shows similar damage; in this case, damage has progressed to a slightly more severe condition but can probably be returned to acceptable parameters by dressing and minor bending 41 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.3.4—Relatively light impact damage in the outer regions of a vane. There is also evidence of mechanical tears in the vane. Fig. 7.3.5—Light peening damage of greater severity in the outer regions of steam flow. 42 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components • Medium—this damage is similar to that shown in Figures 7.3.4 and 7.3.5 but has increased to the extent it cannot be repaired by hand dressing and bending. Figure 7.3.6 shows the outer section of a vane in which damage has ruptured the vane at the outer sidewall. Similarly, Figure 7.3.7 shows a slightly different form of this same level of damage; here the vane does not exhibit such severe crater damage but it has been torn from the sidewall and the opening has closed as a consequence. This form of damage (with the closed opening) is normally caused by a piece of free debris in the axial gap between the stationary and rotating blade rows. There are medium pits and dents that must be repaired. It is normal with this type of damage to remove the discharge tail and rebuild Fig. 7.3.6—Medium damage, being most severe in the outer regions of steam flow, which also exhibits a radial crack starting at the sidewall. 43 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 7.3.7—Localized medium type damage at the outer regions of steam flow. • Heavy—this damage is severe and normally requires the removal of the complete discharge tail and weld refurbishment. There are situations, particularly with large, low-pressure stages, where the complete discharge tail can be removed and replaced by weld attaching an insert. Such damage is most often repaired by total weld deposit, certainly for small radial-height vanes Figure 7.3.8(a) shows a control-stage nozzle plate in which a piece of by-pass valve seat (Stellite 6B) has detached and lodged in an opening. Another piece had passed into the axial gap between the stationary and rotating blade row and caused extensive damage to the stationary row vanes [Fig. 7.3.8(b)]. 44 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.3.8(a)—Mechanical damage on a nozzle block. A piece of pilot valve seat. Fig. 7.3.8(b)—Damage caused by a piece of pilot valve seat that has passed through the control stage stationary vanes. 45 Turbine Steam Path Maintenance and Repair—Volume Two It is not always possible to be sure that such damage conditions exist from condition monitoring alone. Any measured change in state-line efficiency from “enthalpy drop tests” could be due to a number of different conditions related to conditions such as surface texture seal clearances, and will provide no evidence of mechanical damage. Unlike rotating elements, such damage causes no change in vibration levels and is therefore not easy to detect. Solid-particle erosion After the boiler superheater delivers steam or returns it from the reheat section, the first stages of a turbine can be subjected to relatively heavy damage as the result of the gouging action caused by exfoliated scale particles carried over from the boiler. (This is discussed in chapter 4.) This erosion of the stationary blade row can be a slow process, and although its effects and presence can be inferred from both pressure and state-line efficiency measurements, because it is a gradual deterioration, monitoring must be undertaken with test-quality instrumentation. It is often difficult to differentiate between the effects of this damage and damage due to chemical deposition. Figure 7.3.9 shows a portion of a stage from a high-pressure stationary blade row. This row has suffered material loss from the vanes discharge tail. This loss affects both the discharge area and angle and, if not corrected, it will cause a significant operating fuel cost penalty. The scale that enters the row collides with the pressure face of the vanes, and because of its hard abrasive nature, it removes metal from this surface (see chapter 4, Fig. 4.8.5). There is a tendency for the scale entering the stationary and rotating blade rows to migrate to the outer steam flow diameters. Such scale will, therefore, cause more severe damage in that region. This effect can be seen in Figure 7.3.10, where material 46 Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.3.9—Severe material loss due to solid particle erosion on the discharge edge of the vanes of a high pressure stage. Fig. 7.3.10—Extensive damage in the outer flow section of a high pressure stationary element. This damage is caused by a combination of SPE and solid particle impact. The damage has been severe on the thinned portions of the vane. 47 Turbine Steam Path Maintenance and Repair—Volume Two has been lost primarily in the outer diameter positions. In this stage the eroded vane sections have been further damaged by solid-particle impact. The material loss that occurs on the vane discharge tail is of two basic forms (chapter 4): • A gouging or grinding action that removes material at a relatively linear rate. Such a material loss is shown in Figure 7.3.11 • A brittle or chipping type fracture common in austenitic materials. Figure 7.3.12 illustrates partial loss by the chipping condition These types of failure and base materials—i.e., martensitic or austenitic steels—can be repaired. However, the weld filler rod and stress relief requirements will be different for the two steels. Fig. 7.3.11—Severe material loss as a result of a gouging removal from the discharge edge. This stage shows evidence of an earlier repair, some of which material appears to have been lost at a faster rate than the base material. 48 Next Page Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components Fig. 7.3.12—Brittle or ‘chipping’ type failure of a nozzle block discharge tail. Fig. 7.3.13—Erosion damage causing material loss from the outer section of a stationary vane inlet nose. This loss is causing an ‘undercut’ of the vane, and could weaken it against creep type damage. 49 Chapter 8 Refurbishment Techniques for Rotating Blades INTRODUCTION The rotating blade is the steam path element that seems most susceptible to damage in normal operation. This is partially because both water and any mechanical debris impacting the blades will, because of the high tangential velocities of the rotating blades, develop sufficient force to possibly cause severe mechanical deformation. In addition, the stress levels in the rotating blades are so great that when combined with irregularities of steam flow, they create the possibility of causing dangerous levels of vibration. There are also discontinuities in the blade shape, creating a potential problem of stress concentration in many regions. These factors, together with the possible presence of aggressive chemical compounds and the probability of externally imposed shock loads, indicate the continual potential for failure or serious damage to these components, resulting in an 201 Turbine Steam Path Maintenance and Repair—Volume Two interruption of service. Many millions of rows of blades are in service, and their failure is relatively rare. However, they do remain the component causing more outages than any other in the unit. Some of the performance-deteriorating mechanisms do not require repair or replacement. However, they may require the turbine be cleaned. Even this upgrading action must be performed to accepted standards to prevent deterioration of the unit. This chapter considers the damaging mechanisms and deteriorating performance (discussed in chapters 4 to 7), and any other phenomena, in terms of their ability to be repaired. In some instances these procedures are described in general terms. However, many of these repair and refurbishment techniques represent relatively new techniques, which continue to develop as newer welding techniques and improved materials become available. None of the procedures described in this chapter should be considered absolute. It is normally necessary to review each failure or incidence of damage separately, and determine the local condition, which must be considered within the review. Refurbishment/repair technology continues to develop, making new procedures and techniques available. These newer techniques may offer a superior or more cost-effective repair. If welding techniques are employed in the procedure, these must be selected with care relative to the materials involved, and then their application must be calibrated, specified and controlled, because the blade materials (and possibly the stage hardware) can be affected by the preheat temperature; the rates of temperature change for both heating and cooling for the repair and stress relief. The deteriorating mechanisms affecting the blade vane will be examined relative to their effect on the blades. Consideration will be given as to how these damaged or deteriorated blades can be rectified to achieve an acceptable condition to the extent the component can be returned to service. Other rotating components of the turbine, 202 Refurbishment Techniques for Rotating Blades including the blade root are considered in chapter 9. The root is conveniently left to chapter 9, because any repair or corrective action taken on it must be considered in relation to the transfer of load between the blade root and rotor; any change must be reviewed in terms of the effect on the entire blade. When corrective action is required, there is often more than one procedure available. If alternative refurbishment techniques are available, they are considered and described. It must be recognized at the outset however, that because of the diversity of form of the components of the steam path, and from manufacturer to manufacturer, it is unlikely any repair procedures can be recommended as a remedy for all maintenance problems. In practice, it is necessary to examine each turbine, and make an assessment of what repairs can be achieved for a reliable and economic solution. Then undertake these in relation to the requirements of the design. Many of the observed deterioration effects of operation will require components to be replaced. Other situations, either because of the extent of damage or their location, will suggest certain repair procedures can be used without jeopardizing the integrity of the components that have suffered damage. This condition review normally requires mature judgment of the situation, and an assessment of any risk involved. This is particularly the case when repair procedures require large amounts of heat to be used, because if not controlled, they can modify the metallographic structure of the materials. The possibility of modification to the blade or stage hardware materials is especially important in components subjected to high temperatures, pressures, and operational stresses. In many instances, using the existing repair procedures with advanced materials and methods returns the component to a condition making it superior in quality to that which was originally supplied. The basis for the superior nature of these repairs, is that since the component was new, both materials and welding procedures have advanced to the extent product quality can be improved. This is not an adverse reflection of 203 Turbine Steam Path Maintenance and Repair—Volume Two the original design or designer, but a demonstration of advances in state-of-the-art modifications for product improvement. Also for a minimum cost and delay, the owner/operator can return his/her unit to service confident it will perform at a level reflecting the adequacy and acceptability of the repair procedure. There are occasions when, although repairs to existing components may be possible, it is faster and more cost effective for the operator to purchase new components. The factors influencing this depend considerably upon the delivery period for the replacement components, the repair time, the cost of any forced or extended outage, and the availability of materials that are needed to construct the components. No general rules for such a decision exist; each case must be judged individually. The owner should be aware however, that when parts are replaced, the damaged components can often be refurbished and carried as inventory spares for possible future use. For this reason, components should always be removed without destruction, if possible. STEAM PATH CLEANING An essential part of turbine maintenance is the cleaning of the steam path. The steam path surfaces will become coated with various compounds carried over into the turbine unit as a consequence of operation. Although this cannot be avoided, it can be reduced. These deposits have potential degrading aspects to the performance of the turbine unit. Chemical deposits on the steam path components, both the vanes and sidewalls, cause a deterioration of the stage efficiency. This efficiency loss occurs as a consequence of inducing flow separation and turbulence into the steam flow. When a turbine is opened 204 Refurbishment Techniques for Rotating Blades for inspection, and the steam path becomes accessible, it is a normal procedure to clean the components, normally by blast cleaning, although hand techniques can also be used effectively. However, the latter tends to be slower and more costly. Despite the amount of time the manufacturing department of a turbine supplier devotes to preparing the steam path elements to achieve design requirements, these surface conditions will not be maintained for extensive periods after the unit is placed in service. The requirements of steam path cleaning were discussed in chapter 6. It is necessary to consider the methods available for cleaning and how they may be applied to the rotating blades. The available cleaning methods include: • blasting with a suitable medium • water washing • hand cleaning using a suitable solvent Blast cleaning. Blast cleaning is the most commonly used method of removing deposits from the rotating portions of the unit, once it is available for cleaning. This is essential as it allows nondestructive examination and removes even the most persistent deposits. The considerations listed in this chapter must be observed. Modern usage prefers aluminum oxide as the most suitable blasting media. These procedures result in both rotor and blades being suitable for NDE, and return to service. Water washing. Although this procedure is not as commonly used as blasting, it has been used with varying degrees of success, when a unit is open with the rotor and diaphragm removed. Water washing will dissolve soluble deposits in the unit, but the insoluble deposits must be removed by mechanical force, which requires a high-pressure jet. This slower method is made faster with higher 205 Turbine Steam Path Maintenance and Repair—Volume Two pressures, which again introduces some level of risk associated with bending the edges of the blades (if they are thinned). The use of “on-line” water washing is found to be effective on geothermal units that operate on steam, which is initially in the wet region. It has in general been of little use on large utility units with high initial steam conditions, and any efficiency gain has not been sustained for a period sufficient to justify the use of such a procedure. One possible advantage to water washing is that it can remove some of the chemically aggressive compounds from hideouts, if they are water-soluble. However, it should not be concluded that all compounds will be removed. This water washing should only be considered a diluter action. It is not a complete remedy. Hand cleaning steam path elements. Hand cleaning methods tend to be a slow process, but can be justified in situations where certain surfaces are inaccessible, or the use of blasting for extended periods would be required, and such extended exposure to the effects of blasting could possibly damage sections of the blade system. BLADE INLET EDGE EROSION DAMAGE Stages operating with a two-phase flow of water/steam can be subject to moisture impact, and under certain conditions will suffer erosion damage. Blades can operate successfully for many years and the erosion severity should not be such that they need to be repaired or replaced during a normal operating life, providing no unanticipated conditions occur. Manufacturers normally predict erosion life, and expect the blades to be satisfactory for a 30-year operating life. However, a level of erosion penetration can be reached when this 206 Refurbishment Techniques for Rotating Blades impact type damage has proceeded to the extent remedial action must be taken, or the blade would suffer material loss beyond the point at which repairs could be effective. Under such conditions the blades could be operating at risk. Certainly there exists the possibility of missiles being generated from the blade because of high stress concentration and the rate at which erosion is proceeding. It is not possible to make any general statement regarding the need for remedial action or when it is necessary. Rather, each stage and blade within a stage must be judged on its own merit. When, from previous observations, it is deemed necessary to take corrective action and a turbine is to be repaired, then such repairs should be undertaken before the erosion pits penetrate the shield and progress into the base or parent material to any significant degree. Most often the need for blade or shield replacement occurs when a shield has been lost, or secondary erosion or leading edge and/or cracking is present. Figure 3.8.9, in chapter 3 shows heavy (secondary) erosion at a blade inlet edge. This erosion may often require repairs to the blade before the entire shield life is expended by normal erosion. Such replacement is normally required only when localized penetration of the erosion, or a crack propagates too far into the base material. In the case of secondary erosion, if it becomes necessary to repair or take remedial action, efforts should be made to trace the source of secondary collection or concentration and see if corrective action is possible. In the case of crack propagation, it is necessary to determine if the crack has extended into the blade material. This will determine the form of repairs that must be undertaken. Note: Secondary erosion is defined as the local erosion where moisture appears to have been concentrated, causing excessive penetration at one radial location. This is normally seen on each blade in the row. 207 Turbine Steam Path Maintenance and Repair—Volume Two Moisture-impact erosion removes most material from the rotating blade inlet edge in its outer sections. This damage, while causing some small reduction in stage efficiency, normally poses no danger in terms of blade failure, unless some form of secondary damage, resulting from other phenomenon is also present. When the efficiency loss is small, it is difficult to justify (with any degree of certainty) the need for corrective measures. Therefore, the need to make repairs or refurbish the blade is a consequence of the secondary erosion damage, which can become more severe. The secondary damaging mechanisms or phenomenon that require corrective action can take several forms, but can generally be considered to be the following: Detachment of a braze attached erosion shield. If a braze attached shield detaches, there is consequent exposure of the softer material of the blade vane. After a relatively short period of operation, the blade material will erode at what can be an accelerated rate. With some longer shields, it is possible they will have been attached as a number of individual segments on each blade. It is possible, with these designs, for individual segments of the shield to detach, leaving the blades with localized areas exposed, and protection intact in others. Figure 3.8.12 of chapter 3 shows the consequences of such an incident where the outer segment of a shield has detached allowing heavy erosion to occur in the blade material in the outer sections of the blade. Local erosion penetration. It is possible for moisture collection points to exist in the stationary portion of the steam path. These locations tend to concentrate the moisture detachment points, and therefore the radial location at which the moisture makes contact with the rotating blade inlet edge. This will result in severe secondary erosion penetrating the erosion shield. Such damage is shown in Figure 3.8.9 in chapter 3. 208 Refurbishment Techniques for Rotating Blades Cracking originating in a weld-attached shield. With weldattached shields, cracks can initiate at the blade inlet edge. Such cracks normally originate in the HAZ, and propagate towards the inlet edge. However, a small proportion of these cracks will run either radially along the heat-affected zone (HAZ), or across the vane towards the discharge edge. Figure 8.3.1 shows such a crack, and Figure 8.3.2 the form of the weld-attached shield, at the position “Z” (the point at which these cracks originate), and the degree of protection required to prevent erosion in the softer blade material. Figure 8.3.3 shows an exposed crack, of the type shown in Figure 8.3.1 in the shield of a last stage blade element. Fig. 8.3.1—A crack initiating at the leading (inlet) edge of a last stage blade with a weld attached erosion shield. 209 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.2—The crack initiation point in the HAZ of the weld attached erosion shield. Fig. 8.3.3—The exposed section through a leading edge crack showing the failure surface. 210 Refurbishment Techniques for Rotating Blades TAPPS Cracks initiating in the braze material. Figure 8.3.4(a) shows a crack originating at the inlet nose of a blade with a braze-attached erosion shield. This figure shows the blade cross-section, and in Figure 8.3.4(b) a magnification of the inlet nose. At this location there had been a fault in the braze, a porosity hole, which filled with water and collected corrodents that helped initiate a corrosion fatigue failure. TAPPS Fig. 8.3.4(a)—A high cycle fatigue crack initiating at the inlet nose of a last stage blade. Fig. 8.3.4(b)—The initiation sites on a blade inlet nose as shown in figure 8.3.4(a), the cracks initiating in the attaching braze material, where a porosity hole existed, and collected corrosive ions. 211 Turbine Steam Path Maintenance and Repair—Volume Two A number of initiation sites can be seen on the inlet edge, but the primary failure appears to have been at the inlet nose tip. Thermally hardened inlet edges. Some manufacturers provide erosion protection in the form of thermally hardened, vane inlet edges. This method is known to provide good protection. However, should there be a failure to maintain steam reheat temperature, an increased moisture content will result, or a crack developing in the transition zone between the hardened and unhardened material, then corrective action must be taken. In this situation, because there are no externally attached shields, some drastic procedure may need to be adopted. A previous solution has been to “crop” the blade below the crack, and also the blade that is diametrically opposite; this is necessary to preserve dynamic balance. When a situation of damage, or need for repair or refurbishment exists, engineering decisions are required concerning the condition, its severity, and the most appropriate corrective action. There are various procedures that should be evaluated, and where appropriate corrective action is required. These available corrective measures include the following: 212 • Braze attach new shielding to blade material. This is the most common solution, and can be undertaken in the field with the rotor in the unit, if necessary • Braze attach shielding, with weld metal build up. If the blade material has suffered erosion, it could be possible to deposit small amounts of a suitable filler material such as Inconel 82 and hand dress. However, this could require a short period of localized stress relief for the blade material • Weld attach an erosion resistant bar nose to the blade inlet edge. If the shield and blade material have been penetrated extensively, there are considerable advantages to replacing Refurbishment Techniques for Rotating Blades the entire inlet nose with a material better able to resist erosion. This is considered later in this chapter • Weld attach nose, as above, and build up base material to provide a suitable attachment seat or area. This is considered later in this chapter • Raw weld deposit, using a Stellite 6B® stick. This deposit is made directly onto the blade material Weld attaching a solid bar inlet nose This is the procedure used for the weld attachment of a preformed erosion shield, or solid bar nose, to the inlet edge of a blade vane, the vane having suffered damage to the extent the existing shield needs to be replaced to ensure the blade can be returned to service and continue to operate at an acceptable level of performance. Preparation of the blade vane. Upon receipt of the blades, portions of them are used to establish both the vane geometry, and the position of the vane placement on the blade root platform. This includes establishing the vane setting angle “ ξ” at various radial positions. The information gathered must be statistically significant, and sufficient to allow a shutter gauge to be constructed. This gauge must contain sufficient planes of measurement so that each blade can be placed in it and measured for geometric conformance at completion of the repair procedure. Gauge planes should not be more than about 6.0" apart in the radial direction. A typical shutter gauge is shown as Figure 8.3.5. The blade shown here has its inlet nose removed. The inlet nose will often have been partially destroyed by erosion, or removed or deformed by some other phenomenon. This portion of the vane will therefore need to be reconstructed from the existing material. It is normally possible to gauge the outer portion of the blade form from the existing vane material, so that shutter gauges in the outer regions can be reproduced. 213 Turbine Blading (UK) Ltd. Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.5—A blade having had its inlet edge removed, is located in a shutter gauge for dimensional checking prior to repair. After construction of the shutter gauge, the remaining blades are placed in it, and gauged for conformance. Any significant variation in the main vane characteristics, or the parameters shown in Figure 8.3.6, will be noted as this is indicative that the original blade has either been deformed during operation, or was produced outside manufacturing tolerances. Significant deviation from the mean values as measured in the shutter gauge should be noted by blade number. Each blade must be completely gauged. When this has been completed, and details recorded, the inlet edge will be prepared for the attachment of the bar nose. This nose is traditionally Stellite 6B®. However, it is also possible to employ a steel tool, which is hard and can possibly provide superior protection. If the blades being refurbished are from a unit operating on steam produced in a BWR (a 214 Refurbishment Techniques for Rotating Blades direct nuclear cycle) application, Stellite 6B® is not recommended because Stellite 6B® has a high cobalt content and can become radioactive when working on steam delivered directly from the reactor. This makes their rework more difficult. β1 T C -dT β2 b Fig. 8.3.6—The major characteristics of the vane profile. The nose material removal process requires the inlet edge be cut away, and the cutting process must not generate any excessive amount of heat. For this reason a normal cold cutting procedure is required. The material removed must include the entire inlet edge and any remaining erosion shield. If the shield was attached by welding, the material removed should be sufficient to eliminate the HAZ. It is therefore necessary to ensure the following: • Any HAZ material from the original attachment should be removed. This is important if the shield is a “wrap-around” type, as shown in Figure 3.8.7(d) in chapter 3. The back face must be examined by hardness testing to ensure all hard material is eliminated 215 Turbine Steam Path Maintenance and Repair—Volume Two • The form of the bar to be attached should account for any discontinuities in the blade vane. Such discontinuities can include tie wire bosses or holes, together with local thickening or thinning at the blade tip The inlet edge preparation can take several forms depending upon the extent of damage, the form of the blade, the rate of vane twist, blade material, and the amount of inlet removed. The form to be used in any application will be selected based upon the requirements of the blade geometry, and previous experience. In the run-out region at the base of the shield, the radius should be smooth, and large to the extent it will avoid any significant discontinuities. The final shape will depend upon the form of the shield and vane at the run-out position. If the blade vane contains holes (intended to allow tie wires to pass through) special design considerations are required to allow sufficient material to remain after the weld attachment. Such a preparation is shown in Figure 8.3.7(a), with details of the “cut-out” in Figure 8.3.7(b). Here the selection of radius “r” and thicknesses “d” must be made to minimize the effect of stress concentration, which could occur in this region. A blade, in a shutter gauge, with the inlet edge removed is shown in Figure 8.3.5. Here the removed material can be seen to be discontinuous above the tie wire hole. This is to allow a hole to be replaced in the 12% Chrome material (AISI 403) coupon that is attached in the outer region, and at the same time remove the HAZ away from the hole that is to be produced. Inlet edge material of the same form will be removed from each blade so each will have the same geometric form in the region where the shield is to be attached. The shield can be of the same form and, depending upon the vane twist, have the same turning angle. If some form of weld preparation chamfer is to be used, each blade must have the same geometry in this area. 216 Refurbishment Techniques for Rotating Blades ’r’ (b) Gap ’g’ ’d’ Gap 'g' Stellite bar nose Contour around tie wire hole Tie wire hole Blade vane (a) Fig. 8.3.7—A blade with its inlet nose removed, and contoured to avoid the tie wire hole. Preparation for the welding process. Before any attempt is made to attach a new shield by welding, the region on the vane adjacent to the “cut-out” inlet edge must be cleaned of any deposits or grease. This is to be done using a suitable degreaser, wire brush, and solvents that do not contain any corrosion-producing compounds. After production of a weld preparation on the blade vane and shield, and before the main attachment weld is made, the pre-twisted Stellite 6B® nose is attached to the vane at two or more radial positions as shown in Figure 8.3.8(a), by a “tack weld.” This attachment and positioning is required to achieve the following objectives: 217 Turbine Steam Path Maintenance and Repair—Volume Two • The gap “g” between the shield and vane material is established at the desired value for the entire length. This gap is critical, and must be checked at the radiused run-out region at the base of the shield • The weld-prep angles “α” on both the Stellite 6B® nose and the vane must be correct, within a specified tolerance. This is normally checked by some form of gauge. A typical weld prep is shown in Figure 8.3.8(b) • The shield to be attached must be positioned so that sufficient material is available, that the nose form, with the correct radius and blending to the pressure and suction faces can be reformed by machining or hand work at completion of the weld attachment The tack welds must be spaced and be of an adequate length, otherwise they will not be able to restrain the shield position during the full welding process. These welds should be attached first at the tip section, then on alternate pressure and suction sides of the vane/shield interface, maintaining gap “g” to the end of the shield. Tack welds should not be made at the lower run-out radius at the base of the shield, or opposite any tie wire holes. To undertake weld attachment, the blade is placed in a welding fixture, which is produced from compatible material to that of the vane. This fixture is to provide a precision fit, locating the blade at several positions on its contour and sufficient to retain both the vane setting angle “ ξ” and correct radial alignment during the welding process. The fixture clamping must be designed so as not to interfere with the welding access during shield attachment. Prior to welding, both the fixture and blade should be slowly preheated to a temperature of 400-600°F. (The exact temperature depends upon the blade material.) This heating is to be undertaken using an electric furnace. The temperature ramp rate is not to exceed 218 Refurbishment Techniques for Rotating Blades 200°F/hr. The furnace heating should be such that the entire blade fixture assembly is at the correct preheat temperature. This preheat temperature should be monitored and maintained by suitable methods, such as tempil sticks or contact thermometers. If the temperature falls too much, the entire assembly should be reheated to the original temperature. Gap "g" Positioning tack welds Shield of erosion resistant material Fig. 8.3.8(a)—The shield with a “tack weld” holding the erosion shield at the correct location prior to full weld attachment. Gap α1 α2 Blade vane material Shield material Fig. 8.3.8(b)—The weld preparation on the vane and shield for the attachment of a full nose erosion shield. 219 Turbine Steam Path Maintenance and Repair—Volume Two The welding process. The welding process should be completed using “gas tungsten arc welding” (GTAW), with a high frequency start, foot pedal control, and “direct current straight polarity” (DCSP) capability. The power supply should be capable of supplying an output up to 250 amps at 20 volts. The argon shielding gas should be of adequate purity and of welding grade. The tip size and gas flow rate should be adequate to fully protect the molten puddle. A thoriated tungsten electrode of suitable size and tip shape should be used. The weld attachment of the shield should be completed by the use of continuous pass weld deposits, using Inconel 82 filler material, deposited alternately on the pressure and suction surfaces. Each pass is to be started at the tip section, and terminated by a gradual arc break, which is achieved by manipulation of the foot pedal current decay control. The crater on the tack ends of each weld pass should be ground out. Only uncontaminated stainless steel wire brushes and alumina/silicon carbide grinding wheels should be used. The weld deposit is completed by alternate passes and dressing on the pressure and suction faces of the vane/shield interface. On each pass the weld deposit should begin at the tip section, progressing to the base of the shield run-out region. Attention should be paid to the base of the Stellite 6B® inlet edge insert run-out. At this position, the weld should run into a block of steel of the same chemical composition as the blade material, and the weld should terminate in this material rather than on the blade vane. A blade with an attached nose is shown in Figure 8.3.9. Here the inlet edge is to be attached by welding and will be stress relieved and dressed to the final form by hand controlled grinding. Post weld heat treatment. At completion of the weld deposit process, the refurbished blade should be placed into an oven for post 220 Turbine Blading (UK) Ltd. Refurbishment Techniques for Rotating Blades Fig. 8.3.9—A blade with the inlet edge removed, and the bar nose shield ready for attachment. weld heat treatment. If the post weld heat treatment is undertaken only when an oven load is ready, the oven temperature should be maintained above 400ºF while the blades are loaded, i.e., the welded blades should not lose their preheat prior to stress relief. In the stress relief oven the blades should be suspended or stood in a vertical position in their fixture (Fig. 8.3.10). This minimizes the probability and extent of distortion during heating. The stress relief process should be undertaken in a vacuum furnace, to prevent oxidation of the blade vane and root surfaces. When the oven is full, the temperature should be raised at a ramp rate not to exceed 150°F/hr. to a temperature suited to the stress relief process (1,200 to 1,275°F). The temperature should be 221 Turbine Steam Path Maintenance and Repair—Volume Two maintained at this value within +/-25°F for the full stress relief cycle. This temperature is to be continuously monitored, and held for not less than one hour for each 1.0" of maximum section thickness, or four hours minimum. Turbine Blading (UK) Ltd. When this initial heat soak is complete, the blades should be cooled to 1,100°F, at a rate of 15 to 25°F/hr. The blade can then be cooled to 400°F at a cooling rate not in excess of 125-150°F/hr. At that time the blades can be cooled in still air to room temperature, being careful to exclude all external drafts. Fig. 8.3.10—A blade after nose attachment ready for stress relief. Inlet edge profile finishing. When the welding and stress relief operations are complete, the weld deposit area and inlet edge must be dressed to achieve the final profile requirements. This shaping should be accomplished by using uncontaminated alumina/silicon 222 Refurbishment Techniques for Rotating Blades carbide wheels. The surface finish should be consistent with the surface on the original blade, and have a main direction that is radial. The blades are to be mounted to the profile fixture and checked for conformance to their original geometry. The inlet edge refurbished portion of the vane can be checked during the final shaping process by means of specially prepared profile gauges. Weld inspection. Each blade that has been refurbished should be 100% inspected by liquid penetrant testing techniques and radiography. This testing should be undertaken in accordance with the requirements of American Society of Mechanical Engineers (ASME) standard. Turbine Blading (UK) Ltd. Figure 8.3.11(a) shows a blade before weld refurbishment, and Figure 8.3.11(b) shows the same blade after attachment of a complete nose of Stellite 6B®. Here the nose inlet form has been completely restored. Fig. 8.3.11—In (a) is shown a last stage blade with evidence of moisture impact erosion on the outer flow sections. In (b) the damaged portion has been replaced with a solid bar nose of Stellite 6B®. 223 Turbine Steam Path Maintenance and Repair—Volume Two Final moment weighing. When this entire procedure is complete, the blades must be moment weighed, and depending upon the change in mass, a new disposition on the rotor may be required. This is influenced to some extent by the root or fastening form, and the need to return the blades to their original rotor position. Weld attached inlet edge, and build-up of base material A special case of the welded inlet edge shield occurs where there has been extensive damage caused by erosive penetration into the blade material below, or under the shield. This blade material, which had been removed, provides the shield-connecting surface, on which a weld preparation might be required. In this case the blade material must be restored prior to attachment of the Stellite 6B® nose shield. Figure 3.8.12 (chapter 2) shows a blade that has suffered extensive damage as the result of erosive material loss from the outer section of the blade, after losing the braze-attached outer portion of the erosion shield. In this instance weld refurbishment was undertaken in two separate steps. First, a coupon of material that contained the chemical and mechanical properties consistent with those of the blade material was attached. The second step attached a solid inlet edge shield of Stellite 6B®. The steps in this attachment are shown in Figure 8.3.12. The requirements for vane geometry, weld preparation, welding, the post weld heat treatment, and inspection apply to both phases of the process. It is also possible, if relatively small amounts of vane material have been lost, to re-establish the blade by weld deposit using a filler rod consistent with the material properties of the blade material. However, there are possibly other options, such as removing slightly larger portions of the vane until an acceptable base on which to attach a shield is obtained. 224 Turbine Blading (UK) Ltd. Refurbishment Techniques for Rotating Blades Fig. 8.3.12—The three steps in the repair of the blade shown as Figure 3.8.12. If blades are joined into groups by weld attachment or some other process, this can make the refurbishment more complex. The groups can be broken into individual blades, which must then be reconnected after refurbishment, or the refurbishment can be completed by weld repairing the groups. Figure 8.3.13(a) shows a welded group of four blades that are damaged locally by secondary erosion at their inlet edge. In Figure 8.3.13(b) this same group has been restored by the weld attachment of solid Stellite 6B® inlet edges to each blade, without breaking down the groups. After weld attachment the requirements of stress relief and geometric adjustment must be observed. 225 Turbine Blading (UK) Ltd. Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.13—Shown in (a) is a four blade group with eroded inlet edges. In (b) these same type of groups after weld repair with a solid stellite nose. Similarly, in further developments, it is now possible to weld repair the blades without their removal from the rotor. This has significant advantages in terms of time and cost. However, the quality control process requires strict observance, as does the delicate control of technical processes, in terms of shaping and stress relief. Figure 8.3.14(a) shows a blade that has lost a considerable portion of its tip section inlet edge by moisture-impact erosion. Figure 8.3.14(b) shows this same blade after the damaged material is removed, and the blade is rebuilt. Figure 8.3.14(c) shows the coupon that is to be weld attached to the blade material. Figure 8.3.14(d) shows the results of coupons being weld attached, and in Figure 8.3.14(e) is the rotor at completion of the repair process. 226 Refurbishment Techniques for Rotating Blades Fig. 8.3.14(a)—Exhaust blades which have suffered a heavy material loss in the outer flow sections. 227 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.14(b)—The blades of Figure 8.3.14(a) after removal of the damaged tip material. 228 Refurbishment Techniques for Rotating Blades Fig. 8.3.14(c)—The partially rebuilt blades, and the coupon that is to be attached. Fig. 8.3.14(d)—Stress relief of the blades after coupon attachment and vane shaping. 229 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.14(e)—The rotor after completion of the repair. Braze attached shield In both the power and the industrial sectors of steam turbine repair of erosion damage, the most commonly encountered shield replacement technique is removing the old, damaged shields, and replacing them by brazing on new shields. This is an operation that can be undertaken in the field, even with the rotor sitting in the lower half of the casing. However, if the rotor is still in the casing, ensure that excess braze material and flux do not drop into the casing. Also ensure that all old braze material is removed from the blade before brazing new shield material. There are various requirements for achieving an acceptable braze jointing of an erosion shield to the inlet edge of a blade. Regardless of the procedure or braze materials used, it is essential to conform to the following basic requirements, which are common to all procedures: 230 Refurbishment Techniques for Rotating Blades • The surfaces to be joined must be clean—these surfaces must be prepared by removing any oxide surface scale, grease, oil, or dirt. Cover the surface in such a manner that the formation of oxides is prevented during the brazing process. An oxide scale would prevent the formation of a strong joint, and the oxidization would be accelerated by the application of higher temperatures There are various methods of cleaning and removing oxide scale. The most common method in blade application includes solvent cleaning, using such products as chlorinated hydrocarbons and petroleum-based solvents. The use of alkaline and acid cleaning is also common. Alkaline cleaning is acceptable, but is often a slower process, and is more suited to producing a smaller number of joints. Acid or mechanical cleaning is a successful process for the range of stainless steels used for most blades. However, it is necessary to control the concentration of any acid used and to be sure all traces of the acid are removed from the surface when cleaning is complete, and before brazing. • The surface can also be cleaned by blasting—however, the blasting medium used should be one that does not leave a deposit on the surface because this could impede the brazing process. Nonmetallic blasting media should not be used to clean the surface, unless special care is taken, and some other cleaning or washing is used to remove any surface deposits In addition, any surface that cannot be cleaned should be considered as non-bonded, and the holding strength will be limited to those surfaces that are cleaned. • Fluxing of the joint surfaces must be undertaken to promote an acceptable bond—the primary function of the flux is to combine with, and even stop the formation of products that would prevent an acceptable joint from being formed during the brazing process. Secondary requirements or advantages 231 Turbine Steam Path Maintenance and Repair—Volume Two to the use of flux include the formation of a barrier, which helps prevent an oxide scale from forming on the braze surface, and also the formation of a sealer cap on the joint to keep it clean and lower the heat loss immediately after completion of the brazing process • Preheating of the joint to the correct brazing temperature must be achieved—this preheat temperature is selected so the heat in the blade metal is sufficient to melt the braze material, but not so high as to heat the blade material to the extent its mechanical properties are influenced by a change in its metallographic structure. The steels used for long low-pressure stages, are those most likely to be affected by brazing; they tend to be the fine-grained, low carbon martensitic, or precipitation hardened types. The primary precaution is to avoid heating the blade material too rapidly, and to a temperature that will modify its material structure causing a hardening, which could eventually result in stress corrosion cracking The braze filler material must receive heat from the blade material to melt it. If the filler is melted by heat from the brazing flame, then it will cool immediately after the flame is removed, because heat will soak into the blade material from the molten filler. This will prevent a flow of the filler and a suitable bond will not be formed between the joint faces. • 232 Joint geometry is of critical importance as its strength is dependent upon the thickness of the braze material connecting the surfaces—in turbine blades the normal direction of the load is to produce a shear stress in the braze joint material. The approximate relative strength of the joint is shown in Figure 8.3.15. In this case the optimum gap between surfaces would be between 0.002" and 0.004". This is a difficult clearance to maintain, and fixtures and clamps are often employed Refurbishment Techniques for Rotating Blades Fig. 8.3.15—Braze joint strength as a function of braze gap. It is common for the braze attachment of erosion shields to use a “wafer” of filler material that is about 0.003" thick. However, clamping may still be required, but this must be applied carefully, because as soon as the braze material wafer melts, there is a tendency for the shield that is being attached to move to an equilibrium position that may not be optimum in terms of gap clearance. This is particularly so because the shield is thin, and required bending to match the form of the vane inlet edge. • Post brazing activities for martensitic steels are minimal— however, they are important and should be followed, as they will normally improve the total quality of the joint and therefore blade row It is important to carefully cool the blade after the brazing process. Ensure the blade is brazed in a facility that is not subject to excessive drafts where sudden chilling of the joint could occur. It is often better to clad or cover the joint with some form of insulating material to slow the cooling process. It is also necessary to clean and remove any excess braze material from the surfaces of the joints. Cleaning of the joint at completion of brazing is important to remove 233 Turbine Steam Path Maintenance and Repair—Volume Two both excess flux and excess braze material, particularly the material that has deposited as beads in the region of the joint. Cleaning should be completed as soon as possible after the brazing and cooling process. Clean by washing the surfaces with a jet of wet steam or hot water, or by mechanical means, using a wire brush or fine emery cloth. • Post braze inspection is important to ensure the joint is strong, and meets engineering requirement—in the case of brazing erosion shields, an ultrasonic testing (UT) inspection will give an indication of the percentage of attachment that has been achieved. Such UT inspection is not possible however, if the blades are still mounted on the rotor, and the process must be controlled to ensure acceptable quality. Should the braze surface contain holes, as seen in Figure 8.3.16, they should be filled by hand puddling, as they represent locations where corrosive elements can collect Braze attached shield, with base metal build-up If an existing shield is attached by brazing and the shield has been penetrated locally, or has been lost, and the operator wishes to retain the brazed shield form, it is possible to re-attach a new shield in the field. Depending upon the amount of blade material lost (in terms of the remaining braze attachment area), it is possible to rebuild the vane material in-situ, by making small weld deposits of a suitable filler material. The filler material must be consistent with the parent material of the blade, or an Inconel filler can be used. At completion of this weld deposit and surface dressing, a shield is braze attached to the restored vane. If this form of repair is to be undertaken, having lost only one portion of a segmented shield, it is necessary to remove the entire shield, 234 Refurbishment Techniques for Rotating Blades Fig. 8.3.16—Holes at the shield/vane interface of a brazed on erosion shield. re-profile and clean the entire inlet edge, and attach a complete new shield. If the new shield is to be reattached by brazing, it may be necessary to weld rebuild and stress relieve the section locally before attaching the new shield. If the new shield is to be attached by welding, then the repaired vane and shield can be stress relieved together. 235 Turbine Steam Path Maintenance and Repair—Volume Two There are two forms of geometry that are common in braze attached shields; these are shown in Figure 8.3.17. The “build-up” and “replacement” procedures used depend upon the form of shield used on the original blade. There may be situations where blades with an initially recessed shield will be repaired in-situ, and the final shield form results in the proud nose being used. This is because it is difficult to construct the shield recess on a blade vane with it mounted on the rotor. The disadvantages of the aerodynamic losses associated with this form of shield could be more than offset by the savings in time and cost of in-situ repair. Fig. 8.3.17—Forms of the braze attached erosion shield. Raw weld deposit As a temporary measure, a “raw weld” stick can be deposited on the inlet edge of blades to provide protection until more appropriate 236 Refurbishment Techniques for Rotating Blades measures can be taken. This repair does not represent a total solution. The material that is deposited can cause “carbon depletion” at the blade material/deposit interface, making the weld brittle and susceptible to cracking. This form of repair can be effected by the use of a Stellite 6B® weld rod, and with the blade “in-situ.” It is necessary to preheat the blade inlet edge, bringing it to a temperature suited to the parent material on which the weld is to be deposited. The Stellite 6B® rod is then deposited using an oxyacetylene technique. A stress relief operation is necessary upon the blade/weld deposit. An important feature of this repair is control of the cooling cycle after the stress relief heat soak is complete. With this form of protection there is a high possibility of microcracking during operation. The bar nose inlet edge repair is preferable, as it is more durable and it lessens the risk of cracking. The oxyacetylene repair is not recommended for more than about one year of service, but may be useful in extending the life of a blade row until replacement blades are available. Figure 8.3.18 shows 44" 1,800 rpm exhaust stage blades that were temporarily weld repaired to prevent serious erosion penetration until new blades could be made available. After removal, the blades would be available for permanent repairs. Dependent upon the amount of raw weld material that has been deposited, the inlet edge can be re-profiled. This procedure is most suited to making repairs for heavy secondary erosion, where local penetration of the shield has occurred, and the operator is not able to replace the entire blade, or even replace the entire shield at that outage. A major advantage of this form of repair is that it can be undertaken within a relatively short time, say during a normal maintenance outage, and can provide protection of the blade so it can be permanently repaired at a later date. It will also allow the unit to continue to produce full power. (The exhaust stage can account for up to 10% of the unit output.) 237 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.3.18—Raw weld deposit on long exhaust blades. Dressing an eroded surface There is always a temptation for the operating engineer to dress smooth an eroded surface if penetration damage has occurred, but the shields do not require any refurbishment, or replacement. This should not be attempted. By dressing the material, the rough surface, 238 Refurbishment Techniques for Rotating Blades which captures and retains a portion of the impacting water, is destroyed. It is these “retained pools” of water held at the surface by the irregularities that cushion subsequent impacts, lowering the material loss rate. Repair of off-shield erosion (on the vane) As described in chapter 3, there are situations where the erosion damage extends beyond, behind, or below the erosion shield. This type of damage at serious levels is not common, but for a welldesigned blade and turbine, operating at rated steam conditions with a suitable shield, such damage should not occur. A shield replacement program, as described in this chapter, often extends coverage because the eroded portion of the shield and the blade HAZ is cut away. However, the refurbisher of the blades does not remove more material than is necessary to achieve an acceptable weld joint. For this reason, it is not always certain that protective coverage will be extended. If an examination of the blades indicates that the potential for “off-shield” erosion exists, and the new inlet edge form will not correct this condition, even with its extended coverage, then alternative shield forms can be considered. A full inlet edge repair can more effectively cover the following forms of “off-shield” erosion. Beyond-shield erosion. This form of erosion can be protected against by the use of an extended width shield. However, blade mechanical characteristics should be investigated for changes. It is normal when a refurbisher attaches full inlet edge shields to perform vibration tests to determine characteristics before attachment of the shield, and then test it again after attachment to ensure there has been no significant change in the fundamental, or lower order harmonics. Below-shield erosion. If there is significant erosion below the shield, there could be advantages to extending the length of the 239 Turbine Steam Path Maintenance and Repair—Volume Two shield down the radial length of the vane. The required extension can be judged from the level and pattern of erosion on the blades that are to be refurbished. Considerations of changes in vibratory characteristics are the same as for the “beyond-shield” correction described above. Before making a decision to extend a shield, the damage should be assessed in terms of the hours of operation on the blade, and the probable consequences of not making an extension. Between shield segment erosion. This type of damage occurs on blades with a brazed multi segment shield. This will normally be corrected by the use of a full inlet edge, full-length shield. Pressure surface erosion. Should this form of erosion occur, a full inlet edge (bar nose) shield would possibly rectify this condition, depending upon the extent of the erosion on the pressure surface. Such damage is unlikely to be a cause for major concern. Tenon erosion. The tenon erodes as a consequence of moisture centrifuged to the outer diameters of the blade row, which rebounds in the radial space between the coverband and the casing inner surface. Such erosion is shown in Figure 3.8.14 in chapter 3. This type of damage cannot be tolerated if the moisture removes tenon material to the extent that the joint between the coverband and tenon becomes visible for an extensive length, say 10% of the total periphery of the tenon. In this case refurbishment must be undertaken. There are several options that should be considered, including the weld rebuild of the existing tenon material to provide more shear section, the removal of the coverband and a weld rebuild to form a new tenon, and undercover brazing or welding to provide greater attachment strength of the coverband to the blade outer platform. Figure 8.3.19 shows the result of undercover brazing on a last stage blade, where tenon material had been lost. If brazing attachment is attempted, it must be ensured the brazed joint is 0.002" to 0.004" in thickness, and has a generous radius to ensure a strong 240 Refurbishment Techniques for Rotating Blades Fig. 8.3.19—Under cover brazing to attach a coverband where tenon material loss has been experienced. joint. It is also possible to undertake further peening of the existing material. However, it is necessary to ensure this does not “work harden” the tenon material, making it more brittle. The most appropriate action in any situation will normally depend upon the amount of joint exposure, and the operating temperature of the stage. The repair and rebuild of tenons is covered in greater detail in chapter 9. 241 Turbine Steam Path Maintenance and Repair—Volume Two MOMENT WEIGHING OF REFURBISHED BLADES For longer blades it is important the “moment weight” of the individual blades is measured, and then a determination made of the most appropriate distribution of the blades around the rotor rim so as to minimize the total “out-of-balance” forces. These forces need to be balanced by the addition or removal of weight from the rotor itself. This achievement of balance is necessary to minimize the field balance adjustment that is required on the rotor. For the longer blades each element is individually moment weighed, and then an optimum disposition around the rotor is determined. This moment weight data must be provided to the blade purchaser in the form of a drawing or other method, and is delivered as part of the blade supply. If this information is not provided, the blades are of no use, and as such the delivery is not complete. After determination of the most appropriate location of the refurbished blades, it must be considered whether these long blades are, depending upon the form of the root fixture, best returned to their original position. In this case the “out-of-balance” forces in the original position should be determined, and if these forces are beyond those that can conveniently be balanced on the rotor, then some minor trimming should be considered. Note: If the refurbished blades have a root form that requires tangential loading, their assembly may be a little more complex, as it may be necessary to make a tangential shift so that the closing blades are in the optimum position. This should not however, make the mounting order different. 242 Refurbishment Techniques for Rotating Blades The moment weighing process There are calibrating/measuring devices available that quickly and accurately determine the moment weight of individual blades. These devices have a moment arm that can be adjusted to accept the individual blades, and read directly the moment weight. However, the basic principle of moment weighing is relatively simple and can be undertaken by anyone with a single accurate scale, (calibrated to measure to at least 0.5 oz.), and two knife edges on which the blades can be supported in a repeatable manner. This determination of moment weight, the optimum position in blading sequence on the rotor, and the residual unbalance can be undertaken in the following manner. Blade numbering Each blade should be assigned a discrete number from “one” to the total number of blades “Z” in the row. (Note: if extra blades are supplied as inventory spares, they should be identified and moment weighed also.) Blade weighing. Each blade should be accurately weighed, supported on an accurate scale. It is preferable to support the blade vertically, to ensure the weight is applied correctly on the scale pan, as shown in Figure 8.4.1, and determine the weight “Wb.” Blade moment. The blades should then be supported on two knife-edges, as shown in Figure 8.4.2, and the tip reaction “Tt” recorded for each. In setting up to make these measurements, it is necessary to ensure the blades are supported at a position corresponding to the discharge root diameter, on plane “X-X.” Data analysis. After recording these two pieces of data, the following analysis can be made, using the total information: 243 Turbine Steam Path Maintenance and Repair—Volume Two • The measured blade weight “Wb.” This must be accurate to the nearest 0.5 oz • The recorded tip reaction “Tt,” with the setup as shown in Figure 8.4.2. This reaction should be accurate to the nearest 0.01# It is necessary to locate the measuring device at the tip of the blade as close as possible to the tip section. If the blade has tenons for the attachment of a coverband, the knife-edge should be located on the tenons as shown in Figure 8.4.2. h Wb Fig. 8.4.1—Blade weight determination. 244 Refurbishment Techniques for Rotating Blades Fig. 8.4.2—Moment weighing the blade. The fixed knife-edge should be located at the root, at diameter position “Dr,” i.e., the knife-edge must be located at the same position on all blade elements. • The known root diameter “Dr” of the blade at discharge • The measured blade height “h” at discharge The measured weight and tip reaction are as shown in Figure 8.4.2. This diagram shows the distance “Kr” above the root diameter, which represents the position of the center of gravity. From this diagram, and taking moments about plane “X-X” at the root diameter “Dr,” the value of “Kr” can be found. Taking moments about “X-X” gives: Wb . Kr = Tt . h Therefore: Kr = Tt . h Wb 245 Turbine Steam Path Maintenance and Repair—Volume Two Then taking moments about the shaft center “O,” Figure 8.4.2, the following equation is obtained: Mw = WB (Dr/2 + K) Where “Mw” is the moment weight of the blade. Substituting for “Wb” from the previous equation gives: Therefore, from knowledge of the measured weight and tip reaction when supported from the root diameter, the effective moment weight of each blade can be determined. Example 8.4.1 Consider a blade with a vane height of 18.00 ", mounted to the rotor at a mean diameter of 60". One element in a group of 60 blades is weighed and found to have a weight of 10.091#. When supported on a knife-edge the tip reaction is 4.190#. These forces, weights, and distances are shown in Figure 8.4.3. Dm = 60.00" x 18.00 Dr = Dm - h = 42.00" Kr 0 Tt = 4.190# x Wb = 10.091# Fig. 8.4.3—Moment weight of theFigure blade in8.4.3 Example 8.4.1. Moment weight of the blade in Example 8.4.1. 246 Refurbishment Techniques for Rotating Blades Apply this data to find “Kr” using the previous equation. Therefore: .h Kr = TtWb x 1800 Kr = 4.190 10.091 Kr = 8.474 inches The moment weight of this blade can be found from the application of the equation for Mw giving: Determination of tangential position To select a suitable tangential position of the individual blades, and minimize the “out-of-balance” force the following procedure is suggested: Blade listing. A listing of the blades is prepared, ranking them from heaviest (moment weight) to lightest. This ranking should include any available spare blades. The available spares are removed from the group of blades. (It is suggested blades representing averages for the various ranges of weight available are removed.) First blade selection. Select the heaviest blade #1 and place it at the “top dead center” (TDC) position, which is position (1) of Figure 8.4.4. 247 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.4.4—The determined blade disposition around the wheel. Blade placement. Place the next heaviest blade #2 opposite position (1), at the “bottom dead center” (BDC) in position (31). The next blade #3 is placed next to #2, one pitch “l” counterclockwise. The next blade #4 is placed adjacent to blade #1 and one pitch counterclockwise. The next blade #5 is placed adjacent to the #4 blade in a counterclockwise direction. This process is completed placing blades adjacent to the next blade in a counterclockwise direction until all positions have been filled. 248 Refurbishment Techniques for Rotating Blades Example 8.4.2 Table 8.4.1 lists 60 blades arranged in descending order of moment weight, determined by the methods outlined in example 8.4.1. These moment weights vary from 293.474 #-inches to 280.791 #-inches. Using the method outlined above, the following disposition of the blades, as shown in Figure 8.4.4 was determined as suitable. The “out-of-balance” moment resulting from the individual moment weights from the 60 blades can be resolved into vertical and horizontal components. This is done as shown in Table 8.4.2, and as shown, results in a total out-of-balance force “Fh” of 0.8449 #-inches. Units lb-inches 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 293.474 293.118 292.841 291.887 291.756 291.415 291.340 290.906 290.730 290.464 290.370 289.936 289.873 289.562 289.562 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 289.536 289.431 289.101 289.060 288.663 288.615 288.480 288.469 288.252 288.154 288.079 288.069 287.503 287.503 287.349 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 287.331 287.076 286.994 286.627 286.563 286.239 286.080 285.616 285.605 285.070 285.002 284.471 284.415 283.992 283.988 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 283.550 283.535 283.303 283.276 282.999 282.984 282.606 282.210 281.700 281.408 280.888 280.870 280.765 280.742 280.719 Table 8.4.1—60 Blade Moment Weights 249 Turbine Steam Path Maintenance and Repair—Volume Two The sum of the individual horizontal forces is found from: Similarly the sum of the vertical forces can be found from: Using this information the out-of-balance force “Fh,” its phase displacement “ω” can be found. The resultant force “Fh” can be found from the resolution of the vertical and horizontal components of the out-of-balance forces. The angle of inclination “ω” of the resultant out-of-balance forces “Fh” lies in the second quadrant, and can be found from: The direction of the out-of-balance force “ω” can be found from resolution of the vertical component of -0.667 #-inches, and the hor- 250 Refurbishment Techniques for Rotating Blades izontal component of +0.519 #-inches. For example 8.4.2 this angle “ω,” in the second quadrant can therefore be found from: This is shown in Figure 8.4.5. In the matter of determining the most suitable position of a number of blades, programs exist in the facilities of both original equipment manufacturers (OEMs) and other manufacturers, which are superior to this simple hand method. However, owners with replacement blades available, but no balance data can yield good results with this method. + Vert. Resultant = 0.845 #-ins Hor = 0.519 #-ins. - Hor. + Hor. ω° Vert = -0.667 #-ins. ω = 37.9 ° - Vert. Fig. 8.4.5—The resultant “out-of-balance” force and phase angle ‘ω’. 251 Turbine Steam Path Maintenance and Repair—Volume Two Posn. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 Bld 1 59 58 55 54 51 50 47 46 43 42 39 38 35 34 31 30 27 26 23 22 19 18 15 14 11 10 7 6 3 2 60 57 56 53 52 49 48 45 44 41 40 37 36 33 32 29 28 25 24 21 20 17 16 13 12 9 8 5 4 Mom Wt. Angle a 293.4740 0 280.7420 6 280.7650 12 281.4080 18 281.7000 24 282.9840 30 282.9990 36 283.5350 42 283.5500 48 284.4150 54 284.4710 60 285.6050 66 285.6160 72 286.5630 78 286.6270 84 287.3310 90 287.3490 96 288.0690 102 288.0790 108 288.4690 114 288.4800 120 289.0600 126 289.1010 132 289.5620 138 289.5620 144 290.3700 150 290.4640 156 291.3400 162 291.4150 168 292.8410 174 293.1180 180 280.7190 186 280.8700 192 280.8880 198 282.2100 204 282.6060 210 283.2760 216 283.3030 222 283.9880 228 283.9920 234 285.0020 240 285.0700 246 286.0800 252 286.2390 258 286.9940 264 287.0760 270 287.5030 276 287.5030 282 288.1540 288 288.2520 294 288.6150 300 288.6630 306 289.4310 312 289.5360 318 289.8730 324 289.9360 330 290.7300 336 290.9060 342 291.7560 348 291.8870 354 Sin a 0.0000 0.1045 0.2079 0.3090 0.4067 0.5000 0.5878 0.6691 0.7431 0.8090 0.8660 0.9135 0.9511 0.9781 0.9945 1.0000 0.9945 0.9781 0.9511 0.9135 0.8660 0.8090 0.7431 0.6691 0.5878 0.5000 0.4067 0.3090 0.2079 0.1045 0.0000 -0.1045 -0.2079 -0.3090 -0.4067 -0.5000 -0.5878 -0.6691 -0.7431 -0.8090 -0.8660 -0.9135 -0.9511 -0.9781 -0.9945 -1.0000 -0.9945 -0.9781 -0.9511 -0.9135 -0.8660 -0.8090 -0.7431 -0.6691 -0.5878 -0.5000 -0.4067 -0.3090 -0.2079 -0.1045 Res. Hor. 0.0000 29.3455 58.3743 86.9599 114.5777 141.4920 166.3426 189.7219 210.7187 230.0966 246.3591 260.9132 271.6370 280.3009 285.0568 287.3310 285.7749 281.7740 273.9794 263.5295 249.8310 233.8545 214.8439 193.7548 170.2003 145.1850 118.1424 90.0290 60.5886 30.6102 0.0000 -29.3431 -58.3962 -86.7992 -114.7851 -141.3030 -166.5055 -189.5667 -211.0442 -229.7544 -246.8190 -260.4244 -272.0782 -279.9840 -285.4218 -287.0760 -285.9280 -281.2204 -274.0507 -263.3313 -249.9479 -233.5333 -215.0891 -193.7374 -170.3831 -144.9680 -118.2505 -89.8949 -60.6595 -30.5105 Σ 0.5192 Table 8.4.2—Blade Disposition and Resolved Moments. 252 Cos. a 1.0000 0.9945 0.9781 0.9511 0.9135 0.8660 0.8090 0.7431 0.6691 0.5878 0.5000 0.4067 0.3090 0.2079 0.1045 0.0000 -0.1045 -0.2079 -0.3090 -0.4067 -0.5000 -0.5878 -0.6691 -0.7431 -0.8090 -0.8660 -0.9135 -0.9511 -0.9781 -0.9945 -1.0000 -0.9945 -0.9781 -0.9511 -0.9135 -0.8660 -0.8090 -0.7431 -0.6691 -0.5878 -0.5000 -0.4067 -0.3090 -0.2079 -0.1045 0.0000 0.1045 0.2079 0.3090 0.4067 0.5000 0.5878 0.6691 0.7431 0.8090 0.8660 0.9135 0.9511 0.9781 0.9945 Res. Vert. 293.4740 279.2041 274.6296 267.6349 257.3458 245.0713 228.9510 210.7076 189.7320 167.1749 142.2355 116.1660 88.2602 59.5798 29.9607 0.0000 -30.0361 -59.8929 -89.0213 -117.3309 -144.2400 -169.9052 -193.4463 -215.1865 -234.2606 -251.4678 -265.3521 -277.0808 -285.0469 -291.2368 -293.1180 -279.1812 -274.7323 -267.1404 -257.8117 -244.7440 -229.1751 -210.5352 -190.0251 -166.9263 -142.5010 -115.9484 -88.4036 -59.5124 -29.9990 0.0000 30.0522 59.7752 89.0445 117.2427 144.3075 169.6719 193.6671 215.1672 234.5122 251.0919 265.5951 276.6680 285.3804 290.2880 Σ -0.6665 Refurbishment Techniques for Rotating Blades An alternate means of moment weighing Other means of determining the moment weight of the blades, using one or two accurate balances are also available. Figure 8.4.6 shows the required setup for recording the reaction with one or two balances “Bt” and “Br,” which record the readings of weight “Tt” and/or “Tr.” X Y1 Y2 h Dr/2 Kr O Bt Tt Wb X Br Tr The 'B's are the two balance scales on which the blade is placed, the measured weights being 'Tt' and 'Tr'. Fig. 8.4.6—The moment weighing of a blade using two balances ‘Bt’ and “br’. With only one balance “Bt” available, and taking moments about the lower balance: With two balances “Bt” and “Br” available, the value of “Kr” and “Wb” can be found from: The sum of “Tt” and “Tb” is the weight of the blade “Wb.” 253 Turbine Steam Path Maintenance and Repair—Volume Two EROSION SHIELD CRACKS If erosion shields are attached by welding, there is a possibility that cracks could initiate in the HAZ, particularly when the weld filler material that is used for the shield attachment is not sufficiently ductile. These cracks result from both residual stresses and those induced by rotation. The cracks can also result from possible differences in the expansion coefficients between the blade material, filler material, and the erosion shield. One of the major advantages of using an Inconel as weld material is that Inconel is ductile and therefore provides a ductile barrier between the shield and blade material. This weld filler ductility allows for differences in thermal expansion, and the internal stresses associated with temperature changes. The Inconel can therefore accommodate the consequential plastic deformation, which otherwise would lead to crack initiation. Note: However, whenever a weld is made to hard alloys (e.g., an AISI 410 or Stellite 6B®), a suitable stress relief operation should be performed. If AISI 410 or a similar material has been used as the weld filler material, and a crack initiates in the HAZ, corrective action is required. This condition cannot be allowed to persist when the unit is returned to service. The most appropriate action is to weld a full inlet edge attachment in place of the erosion shield. This is however, a time consuming process, often requiring the blade be removed from the rotor. An alternative solution is to cut out the cracked shield, and prepare to replace or otherwise correct the problem at the next outage. It is possible to operate blades with complete inlet edges missing for extensive periods. However, there must be a limit to the number of blades within a row that can be modified by having shields removed. It must be considered that the removal of a shield can modify the natural frequency of the blade, and also affect any blades 254 Refurbishment Techniques for Rotating Blades with which it is batched. Removing the inlet nose is not considered ideal; therefore it is prudent to make arrangements for repair at the first outage opportunity. BLADE TRAILING-EDGE EROSION Trailing-edge erosion is becoming relatively more common, occurring on the last stage blades of units that have operated at light loads with operative water-cooling sprays (see chapter 3). This erosive damage that occurs as a consequence of the root recirculation of water is caused most often from the hood cooling spray water. In the region of this damage, towards the root section, the blade vane stresses are tending to a maximum. Therefore, such erosive damage occurs at a radial location on the vane, where stresses, due to both direct centrifugal loading and dynamic loading due to steam and centrifugal bending effects, are high and can be close to their limiting values in terms of achieving a suitable “factor of safety.” When trailing-edge erosion is observed, a first action should be to examine the spray nozzles at the outer edge of the exhaust blades to ensure they are orientated in their design direction, and don’t spray directly towards the blades. If this form of damage is found, the depth of penetration should be measured, and the condition monitored with depth readings at repeatable locations as often as possible. Review the extent of penetration to determine the depth to which the gouging (erosion cutting) has penetrated into the blade material. A critical consideration of this damage is that the root section discharge edge is, by design, relatively thin. Therefore if grooving exists, it will lead to stress concentration, and failures can be induced. 255 Turbine Steam Path Maintenance and Repair—Volume Two No efforts should be made to weld repair these blades, since the gouging exists in a high stress region, and the elimination of residual stresses cannot be guaranteed. While a crack, should one exist, may appear to be successfully repaired, further cracks will probably form and propagate from the welded region soon after the unit is returned to service. Similarly no efforts should be made to hand dress any damage that is found. If cracks do exist, the blades should be considered in need of replacement. While stages have continued to operate with extensive trailing-edge erosion, this is not considered an acceptable practice, and eventually these elements will fail and the unit will be forced from service. Blades should be purchased as soon as it is determined that the grooves are progressing at a rate allowing the groove to penetrate the complete thickness of the discharge edge replacement. Note: As a general rule, if penetration has proceeded to a depth of 1/16" (or half the trailing edge thickness if the blade trailing edge thickness is less than 1/8") then the blade should be monitored, and preferably replaced at the first opportunity. This, although an expensive option, is probably far less onerous to the operator than the possibility of massive blade failures in the exhaust stage and the consequential damage to both the turbine and condenser. There is no firm technical basis for this recommendation. However, at this time material rupture has not been found prior to these conditions having been reached. It cannot be predicted at what radial position on the trailing-edge erosion that cutting and cracks will initiate. Knowledge of the natural frequencies the position of “nodal lines” and regions of maximum stress is probably valuable. However, it is difficult (if not impossible) to establish all of this data with the blades mounted to the rotor, as individual differences on the blade vanes will modify these positions from blade to blade. 256 Refurbishment Techniques for Rotating Blades When this condition of erosion is found, there is every reason to monitor its progress, and possibly purchase replacement blades for installation at a suitable opportunity. SOLID-PARTICLE EROSION BY OXIDE SCALE This form of damage is normally most severe in terms of material loss in the stationary blade rows (see chapter 4). However, the rotating blades do suffer material loss, and such loss needs to be monitored to estimate the rate of deterioration, and the point in the maintenance cycle at which replacement blades should be made available. The damage that occurs on the stationary vanes and sidewalls tends to cause a deterioration in expansion efficiency, whereas the damage sustained to the rotating blades will cause a deterioration in efficiency, and can also expose the rotating blade to mechanical deterioration, possibly leading to failure. There are three distinct areas where rotating rows should be examined for damage, and some effort made to quantify and record the conditions. On the rotating blade inlet edge. At this location there is material loss due to the “gouging action” of the scale particles as they flow through the stage. A portion of these particles pass between the blades, cutting and removing material from the vane inlet edge suction surface as they do. Another portion of the particles rebound between the stationary and rotating blades again, removing material from the inlet edge and nose. In general this damage is apparent, and is more severe towards the tip, or outer flow sections of the blade. 257 Turbine Steam Path Maintenance and Repair—Volume Two On the rotating blade tip section suction face. When solid-particle erosion damage is found on a stage, it is necessary to examine the pressure face just under the coverband. The design of coverbands used on control stages is often an integral coverband, forming an inner band with an outer coverband and secured by tenons formed on the outer surface of the inner coverband. This form of design is shown in Figure 8.7.1. With this form of damage the inner integral coverband is “undercut” at the tip section just below the integral portion of the inner coverband, as shown in Figures 4.8.9 and 4.8.10 in chapter 4. This damaging effect weakens the attachment of the coverband system, increasing both stress and stress concentration in this region, and modifying the natural frequencies of the blades. The dimensional modifications will also have an adverse influence on stage efficiency. Casing Diaphragm Outer Ring Cr Rotating Blade Vane x Stationary Vane L1 Ca L2 Fig. 8.7.1—The double cover, with the inner band forming an axial seal, and radial seals on the outer band. Damage to the tenons attaching the outer coverband. The scale that is centrifuged out from the stationary blade row, or reaches the casing sidewalls after repeated impacts between the stationary and rotating blade vanes, will pass over the blade coverband, causing erosion damage on the rivet heads formed from the tenons. If fox-holed tenons are used, as on many control stages, this becomes significantly less of a problem as the coverband provides protection for the tenon heads. 258 Refurbishment Techniques for Rotating Blades If the coverbands do not include a fox-holed attachment, this should be considered as a corrective action. However, to make this modification it is necessary for the coverband to have adequate radial depth to allow attachment. Or the coverband thickness can be increased to allow a foxhole to be used. In either case an analysis of the stresses induced should be considered. If the material loss at the tenons is occurring on a row that does not have an integral inner coverband, the covers can be secured by the use of undercover brazing, or welding the underside of the coverband to the blade tip. This can, if applied correctly, strengthen the attachment, and allow the rotor to be returned to service. These forms of solid-particle erosion deterioration on rotating blade vanes are a form of damage that cannot at present be repaired, but weld deposit can be applied to rebuild tenons. Preventive and corrective actions These forms of solid-particle erosion deterioration of stationary and rotating blades are a type of damage that cannot, at present, be repaired except to apply welding to rebuild tenons. However, there are methods available to make the rotating blades more resistant to this type of damage, and when damage is present to lower its rate of progression. These include: • The use of erosion-resistant coatings on the blade vanes and sidewalls. This layer provides a hard, more erosion-resistant surface. This is achieved by coating the affected surfaces with a thin layer of material that is much harder than the blade material, and is therefore able to resist the erosive effect of the scale. These coatings modify the blade material substrate by diffusing elements from the coating into the blades and sidewalls 259 Turbine Steam Path Maintenance and Repair—Volume Two • To operate the unit in a full arc admission mode, at all loads. This reduces nozzle and blade erosion by distributing the particles to all passages at a lower velocity. It also reduces the high impact forces developed on the weakened control stage blades. However, this does introduce significant efficiency losses into the unit, particularly those that operate at partial loads for extended periods • Changes in inlet stage nozzle geometry have been found to accelerate the scale most effectively, and help minimize the material losses. These changes include increasing the axial gap from stationary row discharge to rotating row inlet. Also, modifying the nozzle passage outer sidewall contour has been found to allow the scale to accelerate closer to the steam velocity By increasing the gap between the nozzle discharge and the rotating blade inlet, there is greater axial space for the scale particles to accelerate; therefore they have a less severe impact effect on the following rotating blades. • The introduction of blades designed to modify the distribution of the steam flow that carries the particles into the blade rows, and therefore modifies the effect of impact These various and other methods of reducing this damage will be considered. However, at this time this is an evolving science and rapid changes in technique are occurring as experience is gathered. It is unlikely that any one solution to this problem will be found in the immediate future. However, both the operator and designer can implement certain actions and changes to limit the extent of this damage within any time frame. The actions to be used are based on palliative measures in both the boiler and turbines. These include the following: 260 Refurbishment Techniques for Rotating Blades Acid washing of the boiler. This is the periodic chemical cleaning of the boiler tubes. This cleaning is expensive, but experience has indicated that in their new condition it will normally require about five years of operation to produce sufficient scale on the tubes, which will detach and cause the damage noted on the stationary and rotating blades. If chemical cleaning is employed, it will need to be repeated at about five year intervals. The initial cleaning is the most expensive for the owner, as the initial wash requires the engineering and installation of cleaning lines. This piping system, once installed, can be reused on subsequent washes. Improved boiler tube materials. This is the use of austenitic (or similar) steel that provides a resistance to the formation of thick oxide scale on the boiler tube surfaces. This steel would be used for the superheater and reheater tubes. Unfortunately, there are disadvantages associated with the use of austenitic steels, in that their heat transfer capabilities are inferior to those of the low carbon steels, and therefore a larger heat transfer surface is required in the boiler. There are systems now available for applying a high chromium content deposit on the surface of the boiler tubes. These treated tubes are better able to resist scale formation for considerable periods, and can be retained for the life of the normal unit. Processes exist for “chromizing” new units and a “chromate” treatment of existing units. While these processes can be an expensive investment, the probable cost recovery in terms of reduced maintenance and efficiency losses could soon allow the investment to be recovered. Use of full arc admission. The use of full arc admission at unit start-up ensures scale admission is more evenly distributed around the entire inlet annulus of the first stage nozzles. It also tends to lower the steam velocities at entry to the rotating blades. The principle time for scale exfoliation is at start-up, particularly at cold starts. 261 Turbine Steam Path Maintenance and Repair—Volume Two Therefore, if during the start-up period the scale can be distributed around the complete inlet annulus, damage levels will be reduced. Many examples exist showing that erosive damage is most severe in the nozzle or stationary blades that are the first to admit steam to the turbine section. This is certainly because these stationary blade elements pass the greatest amount of scale that is more abundant at start-up. This severe damage can be seen in chapter 4, Figure 4.8.1, for an austenitic material, and in Figure 4.8.5 for martensitic vanes. Many manufacturers now recommend and are designing bypass systems into their units. This system allows the initial steam to bypass the steam path at start-up, and flow directly to the condenser. This “bypass operation” is continued until the danger of scale exfoliation and carryover has been reduced. Most large, advanced condition units are now built to operate at full arc admission at start-up, thereby admitting steam to the entire ring of the first stage stationary blades. This action allows the total damage to be spread around the entire inlet ring, rather than be concentrated in one area, or segment, of the inlet annulus. Monitoring boiler tube temperature. It is important to carefully monitor boiler tube temperature, and limit it to a level that will prevent or minimize oxide scale formation. This requires rigorous control of temperature in the superheater and reheater walls to provide assurance that scale formation is minimized. With modern computer control methods on new units, this is an easier objective to achieve. Vane material improvements. Attempts have been made to manufacture the stationary blades and nozzles of a more erosion-resistant material. Thus far, no steel has been found with superior resistance to the extent it can be considered a complete solution. However, some success has been achieved by the use of Stellite 6B® inserts on some vanes that have been weld repaired. However, unless the filler material has a comparable hardness, the HAZ will be attacked and lose a disproportionate amount of material. 262 Refurbishment Techniques for Rotating Blades Nozzle geometry changes. Changes in inlet stage nozzle geometry have been found to accelerate the steam most effectively and help minimize material losses. Axial gap increase between blade rows. By increasing the axial gap between the nozzle discharge edge and the rotating blade inlet, there is greater axial space for the scale particles to be accelerated to a velocity nearer that of the steam. The increased velocity of these particles has a less severe impact effect on the following stage rotating blades. The use of protective coatings. These coatings are used to coat or impregnate the surface material of the vanes, providing a harder more resistant surface. There has been considerable activity in identifying and providing a coating that can be applied to both the stationary and rotating blade elements, which will provide a surface better able to resist the erosive damage. Various coatings capable of providing an acceptable level of protection are available, from work undertaken by Electric Power Research Institute (EPRI), and other companies. The rate of erosion is considered to be influenced by several factors, which are beyond the control of operators and possibly even engineering design: Scale particle size. The rate of material removal from surfaces is influenced by the size of the particles impacting them. Size does affect metal removal rates, erosion being almost insignificant with particles smaller than 5 microns, increasing to a maximum with particles of a size 50 to 100 microns. Particle hardness. Particle hardness also influences erosion loss rate, and varies at about the 2.3 power of scale hardness. The temperature of the blade also has some influence on erosion damage. It has also been suggested that on impact, there is a localized heating and annealing, or even melting at the point of impact. If this were so, it would tend to negate considerations of other environmental temperature effects. 263 Turbine Steam Path Maintenance and Repair—Volume Two These considerations do little to cause turbine designers and operators to feel they can place any great confidence in designing away from the solid-particle erosion phenomenon. Design considerations are such that small vane angles are sought to reduce thermodynamic and aerodynamic losses, and maximize stage efficiency. This requirement aggravates the material loss by solid-particle loss impact. Present knowledge suggests that careful monitoring of the level of deterioration of both the stationary and rotating blades is important to help plan blade replacements and repairs (by welding). Planning for a convenient outage reduces the negative effect on system security. It is necessary to ensure deterioration does not occur to the extent a forced outage condition is induced. EROSION RESISTANT COATINGS There are a number of materials and processes available for coating stationary and rotating blades, which are intended to extend their life in an erosive environment. These coatings can improve the material’s surface resistance to solid-particle erosion by producing a protective layer of material, which is able to resist the damaging effects of the scale. These various coatings can be applied by different processes. It is the combination of the various materials and processes that introduce the different levels of protection available to the industry. The maintenance engineer should consider these combinations when the use of such a coating is being evaluated. There has been considerable research to identify and provide a coating that can be applied to both the stationary and rotating blades, while providing a surface better able to resist the erosive damage. Various coatings are available, from work undertaken by EPRI and other companies, capable of providing protection suitable for application to both stationary and rotating blade elements. 264 Refurbishment Techniques for Rotating Blades Coating technology applies protection through the modification of the blade material surface by application of substrate chemistry. It does this by two alternative methods: • Diffusion coatings (diffusion alloying)—This is the diffusion of a resistant material into the substratum of the area to be protected. This is a surface conversion process in which the surface substratum reacts with the diffusion material, normally a boride compound, to form a more protective layer. While there are a limited number of such materials that can be applied at this time, these processes offer a cost effective means of covering complex forms. The diffused coatings tend to be life limited in many turbine applications because the coatings tend to be thin (0.001-0.004" ), and for long cycle times between outages they may be considered to have limitations. However, the increased durability of the nozzle greatly reduces blade damage Research continues on these diffusion processes and materials, and there is the probability that suitable long life materials and processes will eventually become available. • Overlay coatings—This is coating the surface to be protected with a layer of material that adheres to it. The deposited material does not rely upon reaction with the blade material substrate, although normally there is some. These overlays make available complex coating materials, which can offer high resistance to erosion The basic elements used for coatings are, at higher temperatures aluminum compounds, which form an aluminum oxide, Al2O3, and at lower temperatures chromium compounds, which form a chromium compound, Cr2O3. For solid-particle erosion protection, chromium carbide is currently the most popular and effective choice. While all the technologies discussed below may not be currently applied to steam turbine parts, they are mature processes and are 265 Turbine Steam Path Maintenance and Repair—Volume Two available if and when needed. The protection technologies that are currently available include: Electron beam physical vapor deposition. The basic elements used for coatings are, at higher temperatures aluminum compounds, which form an aluminum oxide, Al2O3, and at lower temperatures chromium compounds that form a chromium compound, Cr2O3. For solid-particle erosion protection, chromium carbide is currently the most poplar and effective choice. By robotic control it is also possible to achieve a uniform surface, or to vary thickness to increase coverage in the more susceptible areas. Plasma coating process. Like the electron beam vapor deposition process, the plasma coating process is an overlay “line of sight” treatment, in which the material to be deposited (a powder form) is heated above its melting point and accelerated towards the surface to be covered. As the powder particles impact the surface to be covered, they form a series of “splats,” which overlap and fuse to each other, and are attached to the surface by mechanical bonding. While the temperature of the plasma may reach 50,000°F, the coated part remains relatively cool during the coating process and therefore the mechanical properties of the blade material are not affected. This process can also be undertaken in a vacuum, which improves the coverage and purity of the material deposited. Figure 8.8.1 shows the plasma coating process, and Figure 8.8.2 a 200x etched surface of CoCrAlYSi coating. Chemical vapor deposition. This process allows metallic, intermetallic, and refractory compounds to be deposited. At this time this process has little application to steam turbine components for solidparticle erosion protection, but the ability to cover complex shapes at a relatively low cost may make it applicable at some future time. Pack cementation. Pack cementation is a method in which the parts to be protected are packed into a mixture of powders contain- 266 Refurbishment Techniques for Rotating Blades Fig. 8.8.1—The plasma coating process. ing aluminum or chromium, and then heated to a temperature for a period of time sufficient to form the protective coating required. This method is employed for diffusion alloying. 267 Turbine Steam Path Maintenance and Repair—Volume Two Pack cementation allows complete coverage of all surfaces in which it is required to achieve coverage. This method is economical and areas that are not to be covered can be masked. Figure 8.8.3 shows a packed retort containing the parts to be coated, with the retort and parts being loaded into a furnace for heating. Fig. 8.8.2—A 200x section of a plasma coat. Gas phase coatings. This process is similar to the pack cementation process, except the component to be coated is not surrounded by the powdered mixture. In this process the powder mixture is converted to the gaseous phase and made to surround the component at an elevated temperature. This process is limited to aluminum and chromium coating, but it does allow control of coating thickness and microstructure by eliminating any reliance on the contact areas, as is required with the pack cementation process. 268 Chromalloy Refurbishment Techniques for Rotating Blades Fig. 8.8.3—The pack cementation process; the samples to be coated are loaded into a furnace. Considerable research is being undertaken on coatings suitable for steam path components. Progress has been made in terms of establishing acceptable coatings based on the cost, suitability of coverage, and life. One material has been reported to offer superior properties, and its wear resistance is superior to many other product coatings. It is about 35 times more resistant than an uncoated AISI 422 material. Figure 8.8.4 shows a curve of comparative resistance for various coating, and the base material AISI 422. The coating material used in the EPRI material SPE-8515-HT, is a complex chromium-containing compound (FeCrAlY), which is plasma sprayed onto the component, and is 0.008-0.015" thick, and is suitable for protecting both stationary and rotating blades. 269 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.8.4—Comparative solid particle erosion resistance rates for various coatings. This EPRI coating (designated SPE-8515-HT) is produced by a plasma spraying process, and after deposition it is heat treated at 1,000°F. This heat treatment increases the surface hardness as shown in Figure 8.8.5. Fig. 8.8.5—Effect of heat treatment exposure time on erosion resistance. 270 Refurbishment Techniques for Rotating Blades SOLID-PARTICLE PEENING Peening, or small particle impact damage, occurs on the rotating blades as a consequence of small hard particles, or debris that are free to rebound within the blade path system (see chapter 4). This type of damage causes a deformation of the surface layer of the vanes, producing small craters, or mechanical deformation. These deformations disrupt steady flow, and cause losses due both to the separation of the boundary layer, and the turbulence this creates. There are various sources of these small particles, and these are discussed in chapter 4. In summary the following are the most likely: • The result of mechanical failure of some portion of the steam path, upstream, being “chopped” into smaller particles, which are then free to move with the steam • Debris carried into the unit from the boiler. This is normally weld bead, of a size that can pass through the steam strainer • Debris entering the unit from some external source • Small parts left in the unit during an outage The normal appearance of this damage is a small indentation, possibly with a small rim or lip at some portion of the crater edge, as shown in Figure 8.9.1. This type of damage causes a loss in stage efficiency and therefore output. The most appropriate corrective action is to dress the crater rim, removing the deformed material that is above the edge of the crater. The dressing effect is shown in Figure 8.9.2. No effort should be made to remove any original surface material, and as much of the original surface contour as possible should be preserved. The lips of the craters are best removed using fine emery cloth wrapped around a file end. 271 Turbine Steam Path Maintenance and Repair—Volume Two Direction of Particle Impact Dress lip to remove high spots Original surface Figure 8.9.1impact crater. Fig. 8.9.1—Typical section through a small Typical section through a small impact crater. Fig. 8.9.2—A blade vane after dressing the high spots to remove proud lips. 272 Refurbishment Techniques for Rotating Blades After dressing, the blade vane will not have been completely restored, and there will still be energy losses associated with the remaining craters presence. However, dressing of the crater lip will have helped to reduce the losses. In some instances these craters have been “filled-in” by the deposit of an Inconel filler material. This will however, require a stress relief of the blade material, and the maintenance engineer should evaluate if the potential gain is justified. MASSIVE PARTICLE DAMAGE Massive particle damage is normally the consequence of mechanical rupture of steam path components upstream of the rotating blade row, failure within the same row and the consequential damage caused by large particles trapped within the stage, or failure of components within the valve system, downstream of the steam strainer. If inspection indicates massive damage has occurred to the blade vane due to the passage through, partial passage through, or lodgment within the steam path of a large particle, immediate corrective measures are normally necessary. Failures producing large particles that can pass into the rotating blades have a high probability of causing severe damage to the rotating blade row; this damage is discussed in chapter 4. Figure 4.3.6 shows a row of control stage rotating blade, which has suffered severe damage to the extent the blades must be replaced. Often this type of damage is aggravated because the particles are too large to pass through the blade path, and have not been (or are too hard to be) broken into pieces small enough to migrate down the steam path. The fact that such articles cannot move downstream may in fact prevent further damage from occurring to other rows. 273 Turbine Steam Path Maintenance and Repair—Volume Two Unfortunately, often a sufficient number of pieces of debris are small enough to migrate, while the larger particles are contained. Therefore the steam path will suffer damage on several rows. The massive particles that are contained between stages rebound in the axial gap between the stationary blade outlet and the rotating blade inlet, and the level of damage shown in Figure 8.10.1 occurs, reducing in severity further down stream. When massive particle damage has occurred, it is necessary to review the blade condition before any decision is made concerning the remedial action that should be taken. Alternatives exist. If it is not possible to restore the condition of the existing blades to the extent that there is confidence they can continue to operate until the next planned outage, then it becomes necessary to replace them. It may not be possible to weld repair, because the probability of successful welding is remote, and at this time the capability to rebuild does not exist. If however, there is a need to return the unit to service, certain options are available. These include the following: Fig. 8.10.1—Massive particle damage. 274 Refurbishment Techniques for Rotating Blades To continue to operate the unit in its damaged condition. Depending upon the nature of the damage and the possibility of any material fracture at the impact points, this will be an operation with some level of risk. Stage efficiency will certainly be reduced. It is normally prudent to specify, after an engineering evaluation, the maximum period for which the blades can continue to operate in this damaged condition, and at the same time order replacement blades so repairs can be affected when the unit is next removed from service. This option represents a management (rather than engineering) decision. All an engineer can expect to contribute is an assessment of the condition and the probability of and estimated time to failure. Note: The setting of a maximum period of operation relies entirely on the experience of plant engineers. While the blades can be dressed and straightened, the need to undertake NDE at completion of adjustment exists. Unfortunately, even the act of straightening can cause further damage, which may not be apparent from NDE results. By bending and dressing. Depending upon the extent and location of the damage, efforts can be made to correct the situation by bending, dressing, and NDE. However, this can be risky, and although it can assist in returning the unit to service for a short period, it should only be done after very careful examination of the condition. In this instance, new blades should be ordered to allow permanent repairs to be undertaken as soon as possible. With these two options an evaluation should also include a consideration of the consequential damage that would be induced if failure does occur during the proposed operating period. Removal of the blades. This can be done, and the rotor can continue to operate for some period, normally at reduced output, and possibly with reduced initial and reheat steam conditions. Normally replacement blades should be ordered. However, if due to predicted load factors, or the age of the unit, it is intended to continue 275 Turbine Steam Path Maintenance and Repair—Volume Two to operate in this condition, it will be necessary to provide a pressure reducing baffle in place of the blades. Since the pressure drop associated with high reaction stages are relatively small, it is possible to remove a rotating blade row and continue to operate without further modification of the steam path. However, with the impulse type stage, if rotating blades are removed it will normally be necessary to incorporate a pressure-reducing baffle in place of the blades to ensure the energy level of the steam entering the following stage is not too high. Note: The pressure-reducing baffle is used to replace the stationary blade row or diaphragm to ensure the energy is removed from the steam. This baffle consists of a perforated plate and is designed to make the steam flow more uniform. This baffle allows the unit to continue to operate with the design steam conditions and therefore minimizes the loss of output. In the case of last stage blades, it is often possible to remove an outer portion of the blade or blades and operate in this manner for a period. However, replacement blades should be ordered for installation as soon as possible. If the unit is returned to service without a row of blades, it is advisable that the root portion of the wheel be covered by either the blade root or dummy roots to protect it against attack from chemical corrosion, water, debris within the steam path, or other phenomena that could degrade its condition. There is normally a considerable reduction in efficiency of this stage, which is particularly evident at part load. If the outer portion of only one or two blades is damaged, these can be removed along with other blades diametrically opposite their rotating position, in order to retain balance. Note: If rotating blades are damaged it is normally best, particularly in high temperature stages, to machine off the vanes, leaving the root portion present to cover the wheel portion of the attachment 276 Refurbishment Techniques for Rotating Blades and minimize the possibility of rotor damage by oxidation or other phenomena that might affect it. In the event of massive particle damage, it is usually necessary to replace the blades. Not to do so will require the unit be operated at risk, and there will certainly be a reduction in unit efficiency while the damaged blades are used. The possibility of failure can be reduced by returning the unit to service and operating in a mode that will not induce higher than necessary stresses in the row. The operating parameters that must be controlled include: For control stages Control stages in partial admission turbines are subject to highlevel impulse loading. This loading results from the blades passing alternately into arcs of steam admission and dead bands. To minimize the stresses produced on these blades, the following operating limitations are suggested: • Operate at full arc admission at all loads. This reduces the impact loading on the control stage rotor blades. To do this the unit will be modified to throttle control, and some load limit may be applied • If the unit has sliding pressure control, this should be used at all loads, but again a load limit will probable be required For last stage blades Last stage blades are subject to high stress levels, and an assessment of the damage level must be made to determine the best temporary corrective action possible. These blades are tuned elements, and any corrective action must consider the effect of mechanical dimensional changes on blade frequency, and the possible consequences this will have in returning them to service. There are certain factors in modifying long blades that should be considered: 277 Turbine Steam Path Maintenance and Repair—Volume Two • The effect on coverband and tie wire batching patterns. Consider if these can be retained if some blade elements are removed or cropped • The position of erosion shields relative to the damage. Cracks can initiate in any specially prepared recess made to accommodate the erosion shield • The effect of any tie wire admission holes present in the vane. If a blade is to be cropped how will this affect the stress in this area Note: If a last stage blade is damaged and is removed, another blade diametrically opposite must also be removed to preserve dynamic balance. Also, the removal of these two blades introduces two large “holes” in the row outlet. This can cause large variations in the pressure at inlet to the row, and the blades on either side of the “holes” will have large and turbulent steam bending forces developed on them. In time progressive failures will occur. For this reason, the situation should be corrected as quickly as possible. It will also be necessary with blade removal to adjust any tie wires and coverbands to eliminate the possibility of these elements spanning two pitches, or alternately making batches too short. For all stages In addition to the special considerations for the control and last stages blades, all blade rows with damage will operate with a reduced load producing capability, and probably with increased stress and stress concentration. To minimize the stresses induced in these elements, the following actions should be considered: • 278 Maintain unit inlet and reheat temperatures at their design or lower delivery conditions. Do not exceed design temperature for any reason Refurbishment Techniques for Rotating Blades • Maintain initial and reheat temperatures at constant values, and prevent temperature excursions as much as possible • Maintain unit load at a sensibly constant value (not exceeding an agreed level, set as a consequence of the damage). System load swings should be picked up by other units within the system • Minimize overspeed transients. The requirements for periodic valve testing should still be observed, but such transient conditions should be avoided as far as possible By exercising this level of control over the unit steam conditions and operation, the possibility of failure will be reduced, but not eliminated. It is necessary to minimize transient operation, particularly temperature swings, which can induce high thermal stresses in the blades. In the event of damage to the rotating blades of the types shown in Figures 8.10.2 and Figures 4.3.6, 4.4.5, and 4.4.10 in chapter 4, if the owner/operators makes a decision to dress these blades and return the unit to service, there are several actions that should be taken and considered. These actions are in addition to the ordering of replacement blades, and include: • An initial nondestructive examination. This should be undertaken using dye penetrant or magnetic particle techniques. Any cracks found should be dressed out. If a crack becomes too deep, a detailed analysis should be made, and the risks assessed • The inlet edge should be dressed removing any excess material that would block or obstruct flow of the steam into the blade passage. A minimum of material should be removed. However, some small amount of rounding and blending of the inlet edges should be undertaken 279 Turbine Steam Path Maintenance and Repair—Volume Two • After dressing, a final dye or magnetic particle check should be made of the remaining material, and any cracks should be removed • If the coverbands have sustained damage, they should be checked by hardness testing. It may be necessary to replace these or at least trim them to remove hard spots. The extent of permissible dressing can only be established in terms of its effects on coverband stresses. No general rules can be provided for coverband trimming, and each case must be evaluated separately for the level of damage and risk involved Note: The decision to make temporary repairs to blades having suffered some level of damage and then return them to service is one that is made only when considerable pressures exist to generate power for the system. This is rarely a justified decision, because there are considerable efficiency losses associated with this, in addition to the expense associated with re-opening the unit after a relatively short service period. In each instance the engineer and management should review carefully the “return-to-service” decision for its overall costs, and possible consequences. Fig. 8.10.2—Massive particle damage to a rotating blade row. 280 Refurbishment Techniques for Rotating Blades CORROSION EFFECTS When corrosion damage is discovered on a blade, it often occurs in areas that are not washed or influenced to some extent by the flowing steam. The occurrence of corrosion damage is normally a reason for concern among operators, and the condition found to be present must be evaluated. Such damage has the potential to force a unit from service, and depending upon the nature of the damage can be a cause for expensive repairs and/or corrective actions. The regions most susceptible to damage While the vane normally presents the most visible signs of damage and deterioration, other regions of the blade suffer, and will usually be more significant in terms of their potential to force the unit from service. These include the following: The root fastening area. The root fastening is an area where corrosive ions can wash and accumulate in sufficient strength that as conditions change, the ions become active and initiate corrosion damage. A major concern with this type of accumulation is that the presence of corrosion is difficult to detect after the unit is removed from service. If it is suspected that corrosive action is occurring, there is no easy selection of the correct action. To remove the blades will require an expensive rebuild of the tenons, and not to remove them could place the unit in a condition where it is operating at risk. This becomes a decision and judgment call on the part of plant staff. Figure 8.11.1 shows the condition of a rotating blade row, as removed from service, showing the corrosive attack that is present on the lower portion of the rotating blades, and will almost certainly have penetrated into the root portion of the root attachment. In chapter 6 Figure 6.2.1 shows a rotating blade row with heavy deposits on the blade root area. This deposition occurred on a unit 281 Turbine Steam Path Maintenance and Repair—Volume Two with seawater cooling, and had suffered a condenser tube leak. When such a condition is found, determine the constituents of the deposits, and assess if they are of a corrosive nature or could form corrosive compounds. The composition will assist in the decision. Fig. 8.11.1—Corrosive pits on the lower portion of a rotating blade row, where corrosion may also exist in the load transfer portions of the root attachment. The tie wire hole in the vane. The tie wire hole represents a region where corrosive ions can accumulate. At these holes the wire can be loose [see Fig. 8.11.2(a)], or have a braze connection between the wire and vane as shown in Figure 8.11.2(b). Both arrangements are subject to corrosive attack. The braze connection can be weakened by cracking at the interface between the wire and vane, often caused by poor wire/hole alignment, and the introduction of high bending stresses, which 282 Refurbishment Techniques for Rotating Blades cause separation between the wire and vane as the unit goes into operation. These small separations can be sufficient to allow the collection of corrosive products and then the initiation of cracks at the point where residual stress exists. Fig. 8.11.2(a) and (b)—Tie wire attachment to the blade vane, provides regions where corrosive ions can collect. In (a) is shown the loose connection with gaps between the wire and hole providing easy access for corrodents. In (b) is shown the braze connection, with excess braze; this is an area which will also collect corrodents. The coverband at the tenons. With tenon-attached coverbands, there always exist gaps between the tenon, at its fillet radius, the blade tip platform, and the underside of the coverband. The voids that are produced in this region provide a convenient location for the collection of corrodents, and the possibility of corrosive action in this region is high. Unfortunately, the early stages of any cracking are difficult to detect, as the coverband will not have begun to lift, visual inspection is almost impossible, and nondestructive examination from the outer surface of the tenon is unlikely to indicate damage. 283 Turbine Steam Path Maintenance and Repair—Volume Two Figure 8.11.3 shows a section through a vane and cover, indicating the voids at the underside where the corrosive ions will collect. They can become chemically aggressive when the environment exists to support corrosive attack on the coverband and blade material. Fig. 8.11.3—The voids formed between a tenon head and cover band. These voids can act as collection points for corrosive ions. In the event any of these forms of damage exists, cannot be seen, but are suspected from the observed accumulation of corrosion compounds on the visible portions of the blade, as shown in Figure 6.2.1 in chapter 6, the blade should be checked by visual and nondestructive means as far as possible to determine if cracks are present. In the event examination indicated cracks are present, it is often necessary to provide replacement blades. However, tie wire hole cracks can be repaired by various methods. Coverbands can be removed, the tenons rebuilt, and the coverband (new or the original) reattached. Unfortunately, root cracks cannot be easily corrected on either the root or wheel portion of the fastening, and more radical correction is required. 284 Refurbishment Techniques for Rotating Blades Should corrosion pits, such as those shown in Figure 8.11.4, be found on the blade vane, efforts should be made to identify the source of the corrosive ingress into the unit. No efforts should be made to dress these pits. A typical section through such a pit is shown in Figure 8.11.5. It can be seen there are no “lips” of deformed material that could be dressed to restore the profile. These pits can represent sources of stress concentration, which could lead to cracking and ultimately blade failure. If after examination no cracks are found in the region of the pits, the unit can normally be returned to service. However, for rotating blades, it must be recognized the possibility of failure exists, and consideration should be given to the provision of replacement blades that can be installed at some convenient outage. The need for replacement is dependent upon the local steam conditions and the extent of damage. Corrosion pits will also introduce turbulence at the surface and therefore induce energy losses. Fig. 8.11.4—A blade row having suffered extensive corrosion pitting. 285 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.11.5—A corrosion pit in the root of a 12% chrome blade root. The most obvious methods of preventing, or minimizing corrosion damage, are the use of materials that are better able to resist corrosive attack, and secondly to eliminate or minimize the ingress into and the formation of the corrosive compounds within the steam path. To achieve this, it is necessary for the operator to ensure that as far as possible, corrodents are excluded from the unit. This is done by improving steam purity. Having stated this, it is also known that such action is practically impossible in a normal operating unit, and the most successful method of eliminating significant damage and forced outage, is vigilant examination when a unit is removed from service and available for inspection. The rotating blade path is particularly susceptible to corrosive attack, because once corrosive ions are present, hideouts always 286 Refurbishment Techniques for Rotating Blades exist as a consequence of general stage geometry and blade design. Also the rotating components are almost always subject to continual tensile stress during operation and often the stage temperatures are high. Therefore, if a corrosive compound is present, then corrosive attack will occur. The operator is responsible for determining the presence of conditions that will allow failure to occur, and for minimizing the possibility of their occurrence. The forms of rotating blade corrosion The various forms of corrosion damage are covered in chapter 6, but the significant forms of corrosive damage affecting rotating blades are summarized as follows: Stress corrosion cracking. Turbine blades in the low-pressure section operate alternately in the superheated and saturated condition as the load changes; they are the most likely to be affected by stress corrosion cracking. This is because during the drying process corrosive ions concentrate in the solution and become chemically aggressive. Materials such as AISI 403 and 410 are susceptible when the material has been hardened above its normal value (BHN 260245), and AISI 422 is susceptible due to its higher hardness. In the event cracks are found, and from other observations it is anticipated these are due to stress corrosion, then it is necessary to take corrective action. Weld repair methods allow tie wire holes to be corrected, depending upon the extent of damage, and tenons can be rebuilt. Cracking in the roots require a totally different approach to correct the condition. These are described in chapter 9. Corrosion pitting. Corrosion pitting is probably the most common form of attack found on rotating blades. Fortunately it does not always represent a serious situation, and many rows of blades are in operation with extensive pitting of the type shown in Figure 8.11.1. While such pitting is not a desirable situation, this form of damage 287 Turbine Steam Path Maintenance and Repair—Volume Two can be tolerated, unless and until there is evidence of cracks emanating from the pits, due probably to stress concentration. The greatest difficulty in deciding to return pitted blades to service is the fact that the initiation of cracks cannot be predicted. After determining that the condition of crack initiation and probable propagation is present on a blade, then corrective action becomes essential. Fig. 8.11.6—Pitting corrosion at the closing window on a rotor. The most serious consequence of corrosion pitting occurs when pits form in the root of the blade. At these locations stress levels tend to be higher, and the consequences of stress concentration are more severe. Figure 8.11.6 shows a wheel rim at the closing window where there have been high levels of corrodent ingress and corrosive attack. This damage was found as a consequence of removing the blades to refurbish tenons that had been damaged by corrosion- 288 Refurbishment Techniques for Rotating Blades induced cracking. When corrosion is found to exist, it is unfortunate that the rotor material is normally more susceptible to damage than the blade material (see Fig. 6.8.19 in chapter 6), and this condition poses a far more significant challenge to the correction of the damage, if correction is necessary. Corrosion fatigue. These are failures attributable to corrosion fatigue occurring in the blade, and other susceptible regions that are subject to corrosive attack, with a level of alternating stress sufficient to initiate cracking. The regions most susceptible to cracking are those where there is some stress concentration. If this condition occurs, depending upon the location, it may be possible to undertake some refurbishment, however the possibility of this is remote, as the damage is normally found after rupture has occurred. Coating protection against corrosion Although controlling station water chemistry, and preventing the ingress and formation of corrosive compounds still represent the optimum solution to preventing corrosion, such measures cannot be guaranteed. Relatively minor leaks into the water/steam system still represent the potential for significant damage under the influence of mechanisms for the concentration of ions that occur within the power plant. The most serious forms of corrosion tend to be those that are generally referred to as “aqueous corrosion,” and that occur most commonly in the wet region of the expansion. However, such damage can also initiate during periods of shutdown, when steam will condense and lie in the lower portion of a unit, or even during construction when the unit components in all portions of the steam path can be exposed to humid atmospheric conditions. Damage during periods of extended shutdown can be reduced, or entirely eliminated, by using “nitrogen blanketing.” If a unit is 289 Turbine Steam Path Maintenance and Repair—Volume Two removed from service and “layed-up” with the intention of being returned to service at some future date, then blanketing, while incurring an additional cost, represents a justified investment. The costs associated with blanketing can be more than justified both in terms of reducing the amount of work necessary to return the unit to a serviceable condition, and even in the retained efficiency level from the retention of the original surface conditions. (This assumes the surfaces were in an acceptable condition when the unit was shutdown, or had been cleaned and, as necessary, dressed.) During construction, components of the steam path are exposed for considerable periods to local atmospheres. Such atmospheres can be both humid and also contain airborne corrodents. This is particularly the case with construction at coastal locations. While turbine manufacturers protect their components with suitable coating materials prior to shipment, there is little to prevent either the accidental removal of these coatings during the erection phase, or the conscious removal to facilitate assembly and alignment. With the protective coating removed, the components can be subject to various forms of corrosive attack. Many components produced from susceptible materials can be damaged in the high-temperature, highpressure region of the steam path. Because of the potentially serious consequences of corrosive action within the steam path, there has been continuous research into coatings that can be applied to the blades to protect them against such damage. Various materials have produced encouraging results, and concerns for this form of damage can be reduced with their application. Research continues and undoubtedly more (and possibly improved) coatings will be developed. At this time encouraging results have been obtained using: 290 • cadmium electroplate per AMS 2416 (there may be limitation to the use of this compound and process in some jurisdictions) • sulfamate electroplate per AMS 2424 Refurbishment Techniques for Rotating Blades • aluminide diffusion alloy per TMT 2813L • ion vapor deposited aluminum These coatings have been developed, and have proven successful for use on the blades. What is not certain is their adequacy on the wheel portion of the fastening, and the effect of mechanical deformation of tenons. ROTATING BLADE REFURBISHMENT The rotating blade is subject to various operating phenomena, and damaging mechanisms that will degrade its condition. Some of this damage, such as the deposition of compounds from the steam, will not be significant unless the compounds are corrosive. The blades can be cleaned by blasting, and the surface restored to a satisfactory condition. Other situations require corrective action to prevent a deteriorating condition from worsening, or to return the blade to an efficient condition. Some of the more common and proven techniques applied to the blade will be reviewed. Erosion damage, repairs, and control When examining erosion damage on a blade, it is necessary to consider its magnitude in relation to the number of hours the blade has been in service. For example, if damage occurs that is beyond level (3) (as defined in Table 3.8.1 in chapter 3) during the first year (or 7,000 hours) of operation, then it is possible this blade will be subject to excessive damage unless remedial action is taken. In the event of suspected or obvious secondary erosion, it is advisable to look for the cause of the damage. The damage could be caused by 291 Turbine Steam Path Maintenance and Repair—Volume Two a blocked water collection belt and/or drainage system, or could be a collection point existing within the steam path. In the case of excessive secondary erosion, it will be necessary to evaluate means of correcting this situation. For new units. It is necessary to ensure the turbine drainage system is connected and operational. Failure to connect the drain system at assembly could cause water to collect at some point where it is drawn over into the steam path. Damage caused by a failure to drain normally occurs in the outer regions of the blade. Note: In the case of a blocked drain, where water has been collecting, any sudden drop in unit load will cause water (in large drops or slugs) to enter the steam path, probably causing extensive blade damage. If primary erosion is severe, examine the adequacy of the drain systems one or two rows upstream of the affected rotating blade row. If collection drains just ahead of a rotating blade row are not working as intended, they will cause localized or secondary erosion. Upstream drains have the ability to disperse the moisture in a greater radial direction, and therefore will tend to affect a greater radial distance of the blade inlet edge. This in effect, appears to increase the primary erosion. For older units. When older units are examined, and before any decisions on repairs are taken, it is necessary to examine the extent of erosion damage the turbine has sustained in terms of its operating hours. If it is determined repairs are necessary and justified, several options are available. Field repairs can be undertaken for both brazed and welded erosion shields. Brazed shields can be removed and replaced. However, these repairs must be performed under controlled conditions; but they can be completed without the need to remove the blades from the rotor. The methods of weld attaching an erosion shield were described previously in this chapter. The method of making a temporary weld repair by the deposition of raw Stellite 6B® was described earlier in this chapter. 292 Refurbishment Techniques for Rotating Blades Vane weld repairs at the tie wires holes Cracking that originates at tie wire holes can affect the integrity of the blade vane. These cracks are induced by the high combined stresses often produced as a consequence of the stress concentration associated with dynamic loading within the region of the hole itself. Corrodents, which migrate into the gap between the tie wire and the hole, also affect the cracks and cause corrosion fatigue cracks to initiate. It is possible to introduce stress corrosion cracking where brazed tie wires have been used, if the brazing process has overheated either the vanes or wires. These effects, together with the residual stresses caused by poor wire hole alignment (which results from deflecting the wire and vane during assembly) often initiate cracks, which will then propagate by high cycle fatigue (HCF). Fig. 8.12.1—A crack initiating at a tie wire hole and progressing across the blade vane. 293 Turbine Steam Path Maintenance and Repair—Volume Two The combined effect of the normal operating stresses due to centrifugal loading and dynamic effects, together with the residual stresses, act to produce cracks that initiate at the wire hole corner and run across the chord width of the blade vane. Such a crack is shown in Figure 8.12.1. Here the crack has initiated under the braze material and is extending along the chord, and across the blade. This type of damage is, depending upon the length and depth of the crack, capable of being weld repaired. It is necessary first to excavate the crack to establish its extent. Such excavation is shown in Figure 8.12.2. The following are three distinct methods of repairing these cracks: Fig. 8.12.2—Grinding away a crack in a blade vane prior to weld repair. 294 Refurbishment Techniques for Rotating Blades By weld repairing in-situ. This process is acceptable when the depth of crack does not exceed more than 30% of the vane thickness at the location of the crack. (These cracks are normally present in the low-pressure end of turbines, on those blade rows that employ tie wires.) Such a repair is undertaken using an Inconel 82 rod. The repair is affected after complete excavation of the crack, which must be undertaken by successive grindings and nondestructive examination to be sure the crack is completely removed. After cleaning the crack, remove all traces of the braze material if the wire is braze connected to the vane. Not to do so would contaminate the weld. A suitable means of removing the braze material is a “soak and wash” process using a 50% hydrochloric acid solution. After removing the braze material, the surfaces must be thoroughly cleaned before the weld process is begun. When the surface is clean the weld process builds the blade material back to its original form. To undertake this repair it is necessary to preheat the blade material, and then apply “post weld” heat treatment at a temperature that is about 50°F below the austenetizing temperature. The period should be about 30 minutes. It is also common when undertaking this repair to weld connect the wire to the vane. One advantage of this weld repair technique for last stage blades is that it can be undertaken inside the exhaust hood by gaining access through the low pressure (LP) manways, assuming the crack is visible and accessible. Therefore, it can be undertaken during a weekend outage. Removal of the blade from the unit, or to have unlimited access to it. If the blades are removed from the rotor for other purposes, and cracks are known to exist or are found to be present, the crack can be removed as described earlier, and the vane repaired. The requirements of crack excavation, pre- and post-weld heat treatment are the same as those described above. The steps in the process require the 295 Turbine Steam Path Maintenance and Repair—Volume Two crack be excavated after careful removal of any original braze material that might be present. The vane is then pre-heated, the weld deposited, and the normal stress relief undertaken. The hole is then reamed after any drilling, which might be required. The hole placement requirements are defined in Figure 8.12.3. G G + + A - h h n n - R G G Dr Fig. 8.12.3—The tolerances defining the position of a tie wire hole in the vane in both the chordal and radial directions. Hole plugging and re-drilling. Another repair method requires the hole in the vane be filled with weld deposit, the vane is then stress relieved, and the hole redrilled. This method will normally be completed with the blades removed from the rotor or wheel, and the hole is filled with a weld metal compatible with the blade material. The requirements of stress relief after welding the hole are the same as described for the first option. 296 Refurbishment Techniques for Rotating Blades The first method of repair is considered temporary, suitable for a short period of operation, and the blade should ultimately be replaced. The second and third methods are more complex in terms of pre-repair preparation, but the quality of the repair is considerably higher and it is considered permanent. At completion of any of these three methods of repair, the hole position requirements are the same as those required for a new blade. These are shown in Figure 8.12.3. To pass through the blade vane, an access hole must be provided. Producing such a hole introduces a high stress concentrating effect within the vane, and will increase local stress levels. There are therefore special requirements associated with hole production that must be observed. The chordal and radial location and internal hole finish are selected and produced with careful attention being paid to the main dimensions for its location. Suggested hole location tolerances are shown in Figure 8.12.3. s k k t Fig. 8.12.4—A blade vane showing the tie wire hole, with reinforcement to replace the material removed to permit the wire assembly. 297 Turbine Steam Path Maintenance and Repair—Volume Two Here the diametral location is “2R,” with a tolerance of +/-0.010". In the axial direction the hole should be located at some distance “A” from a radial line such as “G-G” that passes through the vane center of gravity. This location should be within +/-0.015". Small deviations outside these axial locations can be accepted, but only if the entire blade row has the same level of error. It should be noted that the blade vane will have a bending stress induced in it equivalent to the product of the centrifugal mass of the wire, and the hole displacement. The theoretical alignment of the wire to the hole is shown in Figure 8.12.4 If the production of the access hole induces unacceptably high stresses in the vane, local reinforcement can be used. Such reinforcement is shown in Figure 8.12.5, where a vane having a thickness “s” at the hole center, is reinforced locally by the addition of areas “E1” and “E2,” shown in Figure 8.12.5, which approximate the area removed for the wire. This reduced area is equal to “S.(D+2k),” where “D” and “k” are the wire diameter and radial clearance. E1 S k D k E2 Fig. 8.12.5—A tie wire hole showing the dimensional and clearance requirements for a satisfactory admission hole. Engineering dimensional and surface finish requirements of the hole must be observed, to prevent local stresses reaching unaccept- 298 Refurbishment Techniques for Rotating Blades ably high levels. In producing the hole, as shown in Figure 8.12.4, the wire has a diameter “D,” and the vane hole a diameter “D+2k,” where “k” is the mean radial clearance from the wire to the hole surface. After assembly there should be no vane deflection on the larger radial height elements, and no wire distortion on any shorter vanes. The hole must also have either a smooth radius at the entry and exit point from the vane (detail “s”), or a smooth chamfer (detail “t”). The surface finish of the hole should be 32-64 √µ-inches, preferably produced by reaming. Weld repair of integral snubbers. Under high frequency loading, and possibly aggravated by high degrees of blade untwist, it is possible for integral snubbers to fail. This is not a common occurrence, but the blade cannot be reused, should it occur. To replace a blade is expensive. However, techniques exist to permit the weld rebuild of the snubber. After this weld rebuild, the vane must be stressrelieved (temperatures depending upon the base blade material). Vane tip cracks If the vane tip has integral tenons, situations can arise in which high amplitude vibratory loads on the tenon can induce failure. In certain circumstances, the resulting crack can extend into the vane tip. Most often the crack initiates in the fillet radius at the base of the tenon, and is caused by either: • A poor finish of the fillet radius, the radius not blending correctly with the tenon vertical face and the blade tip platform • Interference between the underside of the coverband chamfer and the fillet radius Figure 8.12.6 shows such a failure, with the tenon and a small portion of the blade tip missing. Here the crack at the tip has progressed into the vane to the extent the tenon and a small portion of the blade vane have detached. 299 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 8.12.6—A tip crack showing the tenon having failed as a consequence of high cycle fatigue. With failures of this type, it is possible to rebuild the tenon and any vane material that has been removed. Figures 8.12.7(a) through (d) show the steps in completing such a repair. In (a) is shown the blade as removed from the turbine, with a crack extending under the entire chord of the tenon. It will also be seen that the tenon and coverband are missing. The only reason there had not been complete separation from the vane was that adjacent blades had not allowed further outward movement of the coverband and tenon. In fact, the tenon had failed at the fillet radius starting on the inlet edge. In (b) the tenon has been removed, exposing the failure surface. At (c) the tenon has been rebuilt and stress relieved. Before rebuilding, grind away the failure surface to provide a clean area that has no oxides or other contaminants on the surface, which would pro- 300 Chromalloy HIT Division Refurbishment Techniques for Rotating Blades Chromalloy HIT Division Fig. 8.12.7(a)—The cracked blades as removed from the rotor. Fig. 8.12.7(b)—The blade after removal of the tenon region. 301 Chromalloy HIT Division Turbine Steam Path Maintenance and Repair—Volume Two Chromalloy HIT Division Fig. 8.12.7(c)—The weld rebuilt tenon, and vane tip material. Fig. 8.12.7(d)—The weld rebuilt tenon after finish machining. 302 Refurbishment Techniques for Rotating Blades duce a less than optimum weld attachment. In (d) the tenon has been reformed to the original dimensions. The rebuilding of tenons is discussed in greater detail in chapter 9. When undertaking this type of repair in the region of an erosion shield, it is necessary to check the braze integrity of the shield at completion of the weld, if it had not been removed prior to the weld rebuild. Forming new tenons for coverband attachment A common region for damage to occur is in the tenons, which attach a coverband to the tip section of the blade vane. This coverband is required principally to form one surface of the expansion passage. The forms of damage suffered by tenons are discussed in detail in chapter 9. WATER INDUCTION In terms of the damage caused to rotating blades by water ingress, it is likely that a “slug” of water being ingested into the steam path will cause unrepairable blade damage. The blades will most likely be bent, and therefore destroyed. However, if the water reenters as a steady flow in small quantities, as occurs with a blocked drain, it will cause a concentrated level of erosion, which while removing material from the inlet edge, will not cause mechanical deformation of the vane. Under these circumstances of localized damage, the condition may be acceptable, and will not require refurbishment to allow it to be returned to service and operate in a satisfactory manner. Howev- 303 Turbine Steam Path Maintenance and Repair—Volume Two er, if this type of damage is found to be the reason for the water concentration, it should be investigated and corrected if possible. Such damage is often the consequence of a blocked drain or a closed extraction valve. It is possible this situation could deteriorate and a “slug” be generated, which would destroy the blades. FRETTING CORROSION Fretting corrosion occurs in the blade rows, most often at the interface between the roots and between integral coverbands. It has also been observed at tie wire holes where there is a tight fit, and by contact between a deformed wire and vane. The possibility of refurbishment or correction will depend upon the form and extent of corrosion. The conditions must be evaluated independently. 304 Refurbishment Techniques for Rotating Blades REFERENCES 1. Cotton, K.C., and J. Angelo. Observed Effects of Deposits on Steam Turbine Efficiency, ASME Paper 57-A-116 2. Fraser, M.J. Weld Repair Procedure for Refurbishment of Steam Turbine Blade Elements, Turbine Blading (USA), Inc. 3. Beaudry, R.J., and K.S. McLeod. The Development and Application of Welded Cobalt-Free Erosion Shields for LowPressure Steam Turbine Blades, IJPGC, Atlanta, Georgia, 1992 4. ASME Radiographic Inspection Standard Section VIII, Para. UW-51 5. Swetz, S.D., M.J. Fraser, and G.J. Russell. Major Weld Repair to Tuned L-0 Turbine Blades, EPRI Steam Turbine Blade Reliability Workshop, Los Angeles California, March, 1986 6. EPRI C5085 Project 1408.2 Proceedings: EPRI Steam Turbine Blade Reliability Workshop, Los Angeles, California, March, 1986 7. Bollman, P., R. Tewes, and H. Viertmann. On-Site Weld Repair, without Disassembly to Low Pressure, Last Stage Blading of a 300MW Condensing Unit, presented at VGB Conference, Maintenance Within Power Plants, February, 1992 8. Jordan, S., and M.J. Fraser. Design Modifications and Repairs to Existing Steam Path Components to Improve their Existing In-service Performance, Turbomachinery Congress, Berlin, October, 1991 9. General Electric Company Publication GEA-11850: There are Solutions to Solid Particle Erosion Damage 305 Turbine Steam Path Maintenance and Repair—Volume Two 10. Tran, M.H., and J.R. Kadambi. Characteristics of New Control Stage Section with Contoured Endwall, AMSE Publication PWR-Vol. 7, Latest Advances in Steam Turbine Design, Blading Repairs, Condition Assessment, and Condenser Interaction, papers presented at the JPGC, Dallas, Texas, October, 1989 11. EPRI Research Project Report TR-107021: State-of-the-Art Weld Repair Technology for Rotating Components, Volume 2, December, 1997 12. Protective Coatings for Steam Turbine Components, Sermatch International, Inc., Publication 13. Kramer, L.D., J.I. Qureshi, R.A. Rousseau, and R.J. Ortolano. Improvement of Steam Turbine Hard Particle Eroded Nozzles using Metallurgical Coatings, ASME Publication, PWR-29 14. Power: Coating Technology may help Fight Steam Turbine Corrosion, February, 1982 15. EPRI Project Report CS-5415: Erosion Resistant Coatings for Steam Turbines, September 1987 16. EPRI Report CS-5415: Protective Coatings, prepared by the General Electric Company, Schenectady, New York, September, 1987 17. EPRI Proceedings Project CS-4683: Solid Particle Erosion of Utility Steam Turbines, Chattanooga, Tennessee, 1985 Workshop, August, 1986 18. EPRI Proceedings Project GS-6535: Solid Particle Erosion of Steam Turbine Components, New Orleans, Louisiana, 1989 Workshop, September, 1989 19. Solution to Solid-Particle Erosion, EPRI Journal October/ November, 1990 306 Refurbishment Techniques for Rotating Blades 20. Specialized Coatings, Chromalloy Research and Technology Publication CRT 06913500 21. EPRI Project Report CS-2932: Corrosion Fatigue of Steam Turbine Blading Alloys in Operational Environments, September, 1984 22. Ortolano, R.J. Users Guide for the Use of Corrosion Resistant Coatings on Steam Turbine Blades, EPRI Report, December, 1986 23. Sanders, W.P. Moisture Damage in the Turbine Steam Path and its Impact on Life Extension, Turbomachinery International, Vol. 33, No. 1, January/February, 1992 307 Chapter 9 Damage Mechanisms Arising from Operation and Refurbishment Techniques for Rotating Components INTRODUCTION When a rotating component has been damaged, the maintenance engineer will review the situation and reach a decision concerning the most appropriate action to be taken to allow the condition to be corrected, so that the unit can be returned to a structurally safe and serviceable condition, in the shortest time possible. These corrective actions must be consistent with retaining component and stage reliability. There are instances where utilities and other owners have spare 309 Turbine Steam Path Troubleshooting and Repair—Volume Two rotors or replacement components, which allow necessary replacements to be made. However, this is not a common practice; it is normally limited to installations where duplicate units exist, or in some instances where the financial consequences of a forced or extended outage are considerably greater than the cost of the rotor. The concern with rotating components is that they are normally subject to high levels of centrifugal stress due to their own weight. The rotating components may also be subjected to other forms of stress, which in a damaged condition may be aggravated, thereby increasing total stress levels and reducing the “factor of safety” (FofS) below an acceptable level. When a damage condition has been determined to exist, two actions are required of the engineering group responsible for correcting the situation: • First, a review should be made of the condition to determine the cause of the damage. This review should include, as far as possible, identifying the initiating and driving mechanisms involved (see chapter 1). In many situations this is a relatively simple process, and can be resolved by visual and nondestructive examination. However, there are other situations that can only be established as the result of careful evaluation of both the component, and the manner in which it has been operated Often it is not sufficient to identify the mechanisms of failure or damage. This is because these mechanisms, particularly those initiating the condition, are the consequence of some abnormality in the unit components or mode of operation. The maintenance engineer must identify the cause rather than the effects. This evaluation is an essential step, as it may be possible to take corrective action for future operation, which could prevent a reoccurrence of the damage. 310 The Repair of Rotating Components • Secondly, the most appropriate corrective action must be determined, from an analysis and comparison of the available alternatives, their cost, and time to completion The damaging mechanisms leading to failure in a rotating component are essentially those involved in damaging the stationary components. There are also the effects of complex stresses introduced as a consequence of rotation. It is necessary to consider these effects on the stress levels of the components, and the possible consequence if corrective or remedial action is not taken to return them to an acceptable condition, and then continue to operate. This chapter considers the damaging mechanisms that can affect the rotating components (other than the rotating blades, which are discussed in the chapter 7). It also discusses some of the refurbishment techniques that are currently in use. THE ROTATING COMPONENTS In considering the repair/refurbishment of the rotating components, in addition to the blades, it is necessary to consider their possible forms. The most commonly damaged components, and those most often in need of repair are the rotor, coverbands, and tie wires. Many stages of the steam turbine utilize either coverbands or tie wires. These two groups of components have similar, but functionally different purposes within the stage. The tie wires transmit vibratory loads developed in one element to adjacent members. By accomplishing this they dampen the magnitude of vibration of the blade, and therefore lower the stress levels to which the blades are subjected. Coverbands have a primary function of limiting the steam flow around and through the tip section of the blade. But they are 311 Turbine Steam Path Troubleshooting and Repair—Volume Two also used to fulfill a secondary function, as the tie wire, of helping dampen the amplitude of vibration experienced by the blade vane. These two groups of components, while they perform necessary functions within the stage, add some measure of complexity both to its design and means of construction. They do this by changing the natural frequencies, and by adding additional centrifugal load, which must be supported by the vane and root. When these two components are used, they require a defined method of attachment to the vanes, again possibly adding some level of design and assembly complication complexity. These components are normally batched into discrete lengths within the row, with a specific number of blades in each. In recent designs there have been attempts to increase the length of these batches to the extent of making them span the complete 360 degrees of the blade arc, i.e., the ties through the blades are continuous. There are distinct advantages to this, but design complexity is often added, because it is still necessary for the wires and coverband to grow in the tangential direction during operation as the result of both radial extension of the blades, as a consequence of stress, and as they expand with temperature as the steam is admitted. Because the tie wires and coverbands will heat and cool at different rates from the blades and rotors, they will change dimensionally at different rates. These effects must be considered and accounted for in the design. The turbine rotors are the major rotating component of the turbine, to which are attached the rotating blades, designed to extract energy from the steam as it expands through the steam path. This energy as a force is then transmitted to the turbine rotor, which drives the generator, causing it to rotate against a magnetic field and produce electrical power at the generator terminals. Alternatively, this rotor can be attached to, and drive some mechanical device. Rotors may be manufactured to one of several forms, these forms depending upon the manufacturer and the expertise that has 312 The Repair of Rotating Components been developed and utilized in their particular manufacturing processes. The details of the rotor construction are also influenced by the proportions and dimensional requirements of size, and the number of blades it must carry. The form of the rotor is also influenced by the environmental condition within which the blades and rotor must operate. In sizing a rotor, the torque it is expected to transmit during full load conditions, and the centrifugal stresses induced in it at both normal and emergency overspeeds are perhaps the most critical considerations. In addition to these principal considerations, are the influences of stresses due to blade loading, thermal transients, and the effects of shrink-fit stresses in regions where such methods of construction are used. Also, there are bending stresses induced in the rotor due to its own weight supported between the bearings. The rotor, like all high speed rotating components, requires very careful dynamic balancing. The rotor must be carefully analyzed for critical speeds to ensure these do not occur at values close to, or at multiples of the operating speed. Functions of the coverband The coverband can be used to provide more than one functional advantage to the stage. These can be summarized as the following: Principal function. To provide the fourth side of an expansion passage, and in so doing prevent the steam from expanding through, or being centrifuged through the tip section of the vanes. For rotating blades, the coverband ensures the flow is through the blade expansion passage, and there is no excess leakage due to centrifugal action. Secondary functions. The coverband, however, provides other secondary functions: 313 Turbine Steam Path Troubleshooting and Repair—Volume Two • The coverbands on large radial height blades, which produce a large discharge area, provide a means of controlling that area at the design value, by restraining the vanes from tangential and axial distortion • The coverband can be designed to incorporate part of a steam leakage control system. The coverband can provide either a seal strip, or a platform against which a seal strip, located within the stationary portion of the stage, will seal • The coverband is normally designed to attach to the blade inner (stationary) or outer (rotating) extremities, and it helps to guide the steam through the vanes with a minimum of aerodynamic loss • By being firmly attached to the blade ends, the coverband acts to tie the blades together and dampen the magnitude of vibrations in the same manner as the tie wires Forms of the coverband There are three forms or combinations of coverband design that are used in the stationary and rotating blade rows of the steam turbine, which should be considered: Attached coverbands. These coverbands are produced separately from the blade in multi-pitch lengths and attached to the blades through tenons. The accepted practice is to produce the tenon integral with the vane and to pass this tenon through specially prepared holes in the coverband. The tenon head is then formed. Integral coverbands. The integral type coverband is generally more robust than the attached coverband, and is produced as an integral part of the manufactured blade vane. 314 The Repair of Rotating Components Multi-layer coverbands. There are multi-layered coverbands, in which more than one coverband is used. This form of coverband (in two or three layers) is used for mechanical/structural reasons. The inner coverband is normally produced integrally with the vane, with outer layers being attached through tenons, the tenons being produced with the vane/inner coverband machining. To simplify considerations of the coverbands they can be placed into six groups, depending upon their form and secondary function. These groups are defined, according to the manner in which they form a seal, or provide a concentric platform against which seals can be formed with another steam path components. These groups are: Type A – Coverbands that do not provide any seal or seal platform Type B – Coverbands that provide an axial seal Type C – Coverbands that provide single or multiple radial seals Type D – Coverbands that provide radial seal platforms only Type E – Coverbands that provide an axial seal and a radial seal platform Type F – Special design coverbands In considering the possible shapes of the coverbands, several modifying or influencing factors that affect the overall performance of the stage should be noted: • Ideally, the coverband will act to guide the steam from the stationary blade row or diaphragm into the rotating blade rows through which the steam is expanding. Figure 9.2.1 shows how the coverband helps deflect, or guide the steam into the rotating blade row. It does this by having an overhang, as shown in Figure 9.2.1(a). In some cases a radial seal is provided at the inlet edge, as shown in Figure 9.2.1(b), but 315 Turbine Steam Path Troubleshooting and Repair—Volume Two with adequate radius on the underside to help divert most of the steam down and through the rotating blade row • On the discharge side, the inner edge should be sharp enough to have a minimal directional influence on the steam exhausting from the rotating blade. A comparison between a sharp and rounded discharge edge is shown in Figure 9.2.2 • The blade passage can have either parallel or tapered outer walls, as shown in Figure 9.2.3(a) and Figure 9.2.3(b). The final shape is influenced by stress consideration, and the possible need to produce an adequate seal surface • The coverband can have considerable axial movement during operation, due to differential expansion between the rotating and stationary portions of the unit. This will help determine whether it is possible to use an axial seal, or whether radial seals would be preferable These modifying factors should be recognized as secondary, but valuable considerations in the coverband design. Casing Diaphragm Outer Ring Stationary Vane Rotating Blade Vane Casing Diaphragm Outer Ring Rotating Blade Vane Stationary Vane (a) (b) Figure 9.2.1 Fig. 9.2.1—coverbands and their ability to help deflect the steam discharging from the Cover and their row. ability to help deflect the steam discharging from the nozzle into the bands rotating blade nozzle into the rotating blde row. 316 The Repair of Rotating Components Effect of rounded corner on the underside Rotating Blade Fig. 9.2.2—The effect of a rounded under surface. Casing Casing Stationary Blade Rotating Blade Rotating Blade Axial running clearance Radial clearance Radial clearance above blade. (a) (b) 9.2.3 Fig. 9.2.3—Showing a cylindrical outerFigure sidewall in (a), and an outward tapered design in (b). Showing a cylindrical outer sidewall in (a), and an outward tapered design in (b). Functions of the tie wires Tie wires are included in many stages where the rotating blades have a large radial height vane. Their functions follow: Principal function. The tie wire has a single function within the stage, which is to mechanically link the blades. This must be done with sufficient rigidity, because the tangential vibrations developed in one blade are transmitted to the other connected elements. 317 Turbine Steam Path Troubleshooting and Repair—Volume Two However, during start-up, shutdown, and transient operation, there are periods when the tie wires will be at temperatures that are different to other components within a row. During these times the wires should have sufficient tangential flexibility, or be connected in such a pattern that the temperature differentials will not induce excessive stresses in either the blades or tie wires. Secondary function. The tie wires have a secondary function of helping dampen both axial and torsional modes of vibration. Their ability to do this will be a function of their form. Those tie wires that pass through holes produced in the vane and are not brazed to it, will have little influence on torsional modes, but possibly some effect on axial modes. If the tie wires are brazed to the vane, there will be some dampening to both modes, the actual damping being dependent upon the strength of the braze. However, a brazed joint is not particularly strong against vibratory motion. Therefore, tie wires and coverbands, while having certain similarities in terms of their relationship to the blade, do have different functions within the stage. Forms of the tie wires Because of the energy and therefore efficiency losses associated with the use of tie wires, the designer will avoid their inclusion as much as possible. However, the structural value of these wires and their ability to dampen vibrations has been well documented and the majority of manufacturers continue to use them. These wires not only dampen the magnitude of vibration, they can also modify the vibratory characteristics of an entire stage. Note: Some manufacturers continue to produce long latter stage blades without tie wires, preferring to rely upon the accuracy of the manufacturing process used to produce blades of known vibratory 318 The Repair of Rotating Components characteristics, and by tuning, adjust the blades to avoid dangerous levels and harmonics of vibratory motion. A blade row may contain one, two, or three wires, which can be constructed as a continuous band or arranged in a segmental pattern. The number of wires selected, their diameter, and the pattern in which they are arranged depends upon the designer, who has available experimental data indicating which arrangement will provide the most effective source of damping. However, the final arrangement of any stage will also be influenced by manufacturing considerations and the ability to assemble the wires in the row in the factory and field. Note: The use of three wires has now been discontinued, but many units are still in service with such an arrangement, and these will require maintenance until they are decommissioned. Irrespective of the form of the tie wires used in any row, they will need to be mechanically attached to the blade by some suitable means. This is necessary so they can transmit vibratory stimuli induced in one blade to those to which they are connected. In addition to these requirements, the form of the wire is selected to help satisfy certain other requirements of the stage: • That the aerodynamic losses induced by the wires are minimized, and therefore stage output is maximized • To help to ensure the wire can be manufactured and assembled in the row by economical means. For fieldwork this may require considerations than those required during initial manufacture. The ability to undertake work in the field without the need to disassemble too much of the stage, or to make in-situ repairs is important to the operator, when trying to minimize down time 319 Turbine Steam Path Troubleshooting and Repair—Volume Two • To ensure the design is sufficiently robust, the wire will be able to operate within the stage environment without introducing operational problems • To ensure the stresses induced in the wire due to its own mass, and any pieces it carries, such as ferrules or spacer washers, will not exceed a value that will force the unit from service There are two basic forms of wire. First is the integral type, in which the wire, or a stub of the wire is formed integral with the vane of the blade, and then mechanically connected by some bridging piece. The second type of wire is the continuous wire, which passes through a hole produced in the vane. Both forms of wire fulfill the same function. However, there can be differences in their method of connection, and certain complexities are often introduced into the stage by their form. Wire cross sections The wire can be produced to any cross section that adequately ties the blades together, can be assembled to the row, and can carry the centrifugal load induced in it by its own weight. The most common form of wire in the steam turbine is the circular cross section type. Such a wire can be either solid or hollow. The hollow wire is often used to reduce the bending stresses induced in it by its own weight. This high stress can occur with course pitched blades. Under these circumstances a hollow wire is used, which produces a high section bending modulus, and induces a minimum centrifugal force. If the designer determines, from considerations of required damping, that a tie wire must be used in a blade row, the elliptical shape is preferable to the circular form (from efficiency considerations). It has been demonstrated that generated turbulence, and therefore losses associated with a circular wire are about three times 320 The Repair of Rotating Components greater than those of an elliptical wire, whose minor axis is of the same dimension as the diameter of the circular wire, and whose major axis is four times the minor. Such wires are shown in Figure 9.2.4(a), and a hollow wire in Figure 9.2.4(b). In fact it is difficult or impossible in a steam turbine to use an elliptical wire with a major to minor diameter ratio of 4.0 with the continuous form of wire, as this wire must pass through a hole produced in the blade vane. However, this may be achieved or approximated in the integral wire. The circular wire is also used in the hollow format, as shown in Figures 9.2.4(c) and (d). This configuration is particularly useful when local stresses in the vane are high, and it is necessary to reduce wire centrifugal loading without reducing wire section modulus by a significant amount. It is also possible to employ wires produced from titanium, this being a metal of lower density and higher mechanical strength. However, this does add to unit cost. When replacing wires, a solid wire should not be substituted for a hollow wire without a careful analysis of the implications for both the wire and blade vane. While the true elliptical wire may not be practical in many applications, the turbine manufacturers have approximated this form to a degree sufficient to eliminate some of the losses that are induced by the generation of turbulence behind the wires. Dx is the major elliptical diameter. Dn is the minor elliptical diameter. Dx Dn (a) (b) (c) Figurewire 9.2.4cross (a), (b) and (c) Fig. 9.2.4 (a), (b) and (c)—Alternate sections. Alternate wire cross sections. 321 Turbine Steam Path Troubleshooting and Repair—Volume Two Dx is the major elliptical diameter. Dn is the minor elliptical diameter. Dx Dn (a) (b) (c) Fig. 9.2.4(d)—A hollow outer wire of the form shown in figure 9.4.2(c) brazed into position on the blade vane. Some common wire forms approaching the elliptical are shown in Figure 9.2.5. In Figure 9.2.5(a) is a wire section formed by the straight portion, length “L,” and two rounded ends of radius “R.” In Figure 9.2.5(b) the form is similar, but the radii are not equal; they compromise two radii “R1” and “R2” between a length “L” between centers. These forms of wire shown in Figure 9.2.5 are suited to the integral form of wire only. L R (a) L R R1 R2 (b) Fig. 9.2.5—Various wire sections approximating the elliptical. The steam in the latter stages of low-pressure sections can have a large radial flow component at some radial locations, in addition to the axial component of velocity. Such a flow pattern is shown in Figure 9.2.6, where the streamlines at the tie wire positions are inclined to the axial or horizontal direction by “α.” Because of this 322 The Repair of Rotating Components radial flow effect, if an elliptical wire is used, it is advisable to adjust its inclination so that there is no large degree of steam incidence at the wires, and the steam will approach the elliptical form at an angle consistent with steam direction. It is difficult to predict exactly the angle of the streamline at the wire position, and the effects of flow patterns at part load, when the steam is flowing at a reduced rate, and possibly in some stages with a modified volumetric flow compound this difficulty. However, there are advantages to this elliptical (or semi-elliptical) wire, and manufacturers use it to good advantage in their units. α1 α2 Fig. 9.2.6—Stream line flow pattern around an elliptical tie wire. Batching of coverbands and tie wires Blades are batched (connected in groups) by both tie wires and coverbands. This is done so that each batch forms a continu- 323 Turbine Steam Path Troubleshooting and Repair—Volume Two ous, although complex, mechanical structure. This interconnection of the blades is capable of dampening the amplitude of vibration induced in individual blades, by vibratory forces from one blade on all those to which it is connected. When mechanically connected, the blades within each group act together to form a total damping arrangement. The batching patterns used by different manufacturers, and used at different times in their technology development to establish the pattern they will use to batch the blades together can be placed into two broad categories: • Connecting the blades of a row into separate and discrete batches. Such a discrete batching is shown in Figure 9.2.7(a). Normally each group contains an equal number of blades, which for larger blades are selected after tests have been conducted to ensure these will operate within a safe frequency range For smaller radial height blades, this batching may be only the grouping of coverbands. If blades are to be replaced the same pattern must be achieved in the replacement blades, with particular attention being paid to the position of any closing blade or root block. While the number of blades in each batch should be about the same, the number of batches and any differences in blade count from one batch to the next must be observed. In this batching pattern all ties (wires and covers) connect the blades together. These batches are free to vibrate as separate assemblies without any influence being transferred from adjacent batches within the row. • 324 Arranging the blades into a staggered (or random) pattern, as shown in Figure 9.2.7(b). In this case an attempt is made to tie the blades in such a manner as to transmit the stimuli from any blade throughout the entire row. This is done in The Repair of Rotating Components such a way that it can maintain sufficient tangential flexibility. Blades can accept temperature changes without excessive distortion. (a) Groups of six blades in discrete groups (b) Random batching Fig. 9.2.7—Alternate batching of tied blades. In (a) is shown discrete batching with elements in groups of six elements, tied by a coverband and two tie wires. In (b) is shown a random batching with a coverband and two tie wires. For multiple ties (2, 3, or 4 connections), the pattern should be selected so that at no point are there less than two ties transmitting the motion between any two batches. It is also necessary for the manufacturer to specify the minimum number of blades that must be present between adjacent batch ends. The staggered pattern is usually specified by the designer, and must be adhered to in the manufacturing phase and assembly. If such 325 Turbine Steam Path Troubleshooting and Repair—Volume Two blades are disassembled for maintenance, details of the batching pattern should be recorded and repeated on reassembly. Single tie connections Manufacturers still disagree as to the most suitable pattern to batch the blades together when a single tie connection is used (a coverband or single tie wire). However, there are certain basic requirements that should be considered and evaluated for the assembly: • An odd, preferably prime, number of batches should be used • The manufacturer can elect to use a varying or constant number of blades per batch. The actual pattern must also be influenced by the number of blades in the row, and the stage temperature (because of equalizing expansion and end thrust with temperature changes) • Batch length for the blades should exceed three stationary blade pitch lengths for the row from which they are receiving steam. This is considered necessary to prevent a swinging mode of vibration developing in a tied batch. Such a swing would be due to the impulse received by the blade vane as the batch moves past the nozzle element, which does not have an even pressure distribution across its discharge pitch Some blades have their mechanical design improved by the use of long arc coverbands (shrouding), which ties the blade with sufficient constraint, considerably reducing the amplitude of vibration and therefore vibratory stress. 326 The Repair of Rotating Components Functions of the rotor As the primary rotating component of the steam turbine, the rotor has a number of functions that it must fulfill in order to allow the turbine to operate successfully. The main functions are: • The rotor must be capable of withstanding the centrifugal loading developed on it due to its own weight. In the case of monobloc and welded designs, the rotor must also be able to withstand the centrifugal loading from components such as blades, tie wires, and coverbands that it carries • In operation, torque is developed on the rotor due to expansion of steam and the work done by the blades during this expansion. The rotor must be capable of transmitting this torque to the generator. It must also be capable of transmitting torque developed on other rotors further from the generator, or driven equipment With multiple flows, low-pressure rotors, and when more than one similar unit is used in any station or system, it is often advantageous to produce these rotors as dimensionally close to each other as possible so they may be interchanged between both units and stations. This often requires the couplings and journals on low-pressure rotors to be sized to meet the total torque transmission and the load requirements of the final rotor in the multiple rotor train. • The rotor must be sized and manufactured from materials that are capable of withstanding high temperatures and pressures, and be able to operate for long periods in these conditions. Such rotors must possess high resistance to both creep and rupture • During operation, it is possible rubs will occur between the rotating and the stationary parts of the turbine. These rubs can generate very high local temperatures. The rotor must be 327 Turbine Steam Path Troubleshooting and Repair—Volume Two able to operate to the greatest extent possible without sustaining undue damage from this occurrence • Many rotors operate in an environment having relatively high moisture content. The rotor material must offer resistance to both impact erosion and washing erosion • Although station chemistry limits, as far as possible, the ingress of aggressive chemical compounds to the steam path, complete freedom from chemical attack cannot be guaranteed. It is therefore necessary for rotors to be manufactured from a material having as great a resistance as possible to chemical reaction with any compounds that may be introduced with the steam The rotors must be thermally stable and resist any tendency to bow or sag as a consequence of temperature or temperature changes. The rotors must exhibit good journal characteristics and be capable of producing a high quality surface finish in the vicinity of journals. Rotor construction The form of rotor selected for any particular turbine section or application depends upon several factors: • 328 The manufacturing techniques developed and proven by the manufacturer. It is reasonable to anticipate that different manufacturers should develop different techniques based upon an extrapolation of their design philosophy, sources of raw material, and availability. Such final selection will also be influenced by their in-house manufacturing capabilities The Repair of Rotating Components • The stress levels produced by the operating torques, centrifugal loads, and bending stresses. The stresses a rotor experiences must be considered relative to the temperature at which the components will operate, and also the temperature transients they are likely to experience during operation. A rotor construction unsuited to high temperature operation may well be acceptable, and structurally preferable at lower temperatures, where requirements have changed or are significantly different • The rotor’s operating environment can also dictate that certain forms of construction and materials would be unsuitable • Raw material sources, purity, and integrity are a significant factor in establishing rotor form. Recently there has been a trend towards increasing the diameters of rotors. This increase is normally limited by the forge master’s capability of producing larger rotor forgings of sufficient mechanical strength • Shipping limitations have not yet caused a limitation to rotor size. It has, however, meant many low-pressure rotors, particularly those driving four-pole generators (1,800 and 1,500 rpm), are shipped without final assembly of the long, last stage blades. For the majority of manufacturers, this has required shop assembly of the long blades for testing and balancing, and then disassembly at the manufacturing plant before shipment. Final reassembly is made at site There are three basic forms of rotor constructions used in modern turbines, and certain combinations of these forms in older units, many of which are still in use. The three basic forms are: 329 Turbine Steam Path Troubleshooting and Repair—Volume Two Monobloc. The monobloc rotor is produced as a solid forging. After forging, the rotor may be gashed to form individual discs for each stage or group of stages. These discs will then be machined at their outer diameters to allow blades to be attached. The decision to gash normally depends upon the pitch, or the ability to admit a diaphragm in the axial spacing between the rotating blades rows. This form of construction using a diaphragm enables steam leakage sealing to be affected at a smaller diameter, and therefore, reduces the leakage area. In general, because the impulse type turbine has fewer stages than the reaction turbine, and there is a greater pitch between them, it is possible to arrange access for the diaphragm, and provide sealing at a smaller diameter. Figure 9.2.8 shows the outlines of a typical gashed monobloc rotor. Inspection bore hole Local chambering to remove impurities and inclusions Fig. 9.2.8—The “gashed” monobloc rotor forging, used for the wheel and diaphragm type unit. Figure 9.2.9 shows the outline form of a monobloc rotor more characteristic of the reaction design, which requires more stages to expand the steam between the same pressure limits. This arrangement does not therefore have sufficient axial space to permit the use of a diaphragm. Because of its greater central diameter, this rotor has certain advantages associated with its greater stiffness and load carrying capability. 330 The Repair of Rotating Components Typical root form machined into rotor surface Inspection bore hole Fig. 9.2.9—The “drum” type monobloc rotor forging, as used in reaction type rotors. With this form of construction, the blades are attached into root form grooves machined directly into the outer surface of the rotor body. Because of the lower pitching, these individual stages are closer together than those associated with the impulse design. Built-up. When the diameter required at the blade root exceeds what the forge master can produce, and ensuring adequate material properties, the rotors are often produced by a shrink assembly of individual wheels onto a central forged spindle. These wheels are, in addition to their shrink fit, keyed to the spindle. This keying is intended to prevent any movement of the wheels during transient operation particularly during “emergency” overspeed conditions when the shrink fit could become loose. If the wheel does lose its shrink fit, it would be able to rotate and/or move in an axial direction. Therefore, the key is not required to transmit any torque produced on the blading by the expanding steam, but is a simple locating device intended to limit any movement of the disc during conditions that would eliminate the shrink fit. The basic geometry of such a built up rotor is shown as Figure 9.2.10. 331 Turbine Steam Path Troubleshooting and Repair—Volume Two Fig. 9.2.10—The “built-up” rotor. This Figure design 9.2.10 has individual discs shrunk onto a central forgedThe spindle. This design is required when rotor diameter "built-up" rotor. This design has individual discs shrunk ontorequirements a central forgedexceed spindle. the ability of the to produce larger diameter. This forge designmaster is required when rotorforgings diameterof requirements exceed the ability of the forgemaster to produce forgings of larger diameter. Because there can be problems associated with stress concentration and the accumulation of corrosive ions at the key way and shrink fit regions, there has been continuing pressure to develop a means of producing the larger diameter rotors from a monobloc forging. This can now be achieved to increasing diameters. However, there are a large number of rotors in operation of built-up construction, and they are subject to the problems associated with this form of construction. Welded. The welded rotor consists of a series of individually forged discs located relative to each other in both the axial and radial direction. Each disc is located by positioning grooves and spigots, or other devices that are used to ensure axial and radial alignment is sufficient to provide concentricity, and help ensure no significant “out-of-balance” forces will exist at completion of the welding process. The individual forgings must be accurately machined on 332 The Repair of Rotating Components their inner surfaces before welding, as access is not possible once welding is complete. Therefore, all surface machining, evaluation, and fit requirements must be complete and acceptable on the inner surfaces before “vertical stacking” to undertake the welding process. Figure 9.2.11(a) shows a rotor constructed from individually welded discs for a double flow low-pressure section. It contains six individual forgings, of which two have shaft end stubs integral with the last stage disc. Figure 9.2.11(b) shows a similar rotor for a smaller unit used in a high-pressure section. In this high-pressure rotor, shown in section (a) can be seen the four individual forging details. In (b) are the individual forged discs before welding, and in (c) is the final rotor after machining the root details. These four discs are welded together, forming a rotor containing both an impulse (control stage), and reaction stages. In both Figures 9.2.11(a) and 9.2.11(b) the individual discs are welded together to form a continuous rotor. The preparation of the joint for both location and welding is critical to the success of this form of construction, because once welding is complete, there is no access to make any correction to the inner surface. Weld defects, if they are found to exist, must be removed and the weld remade. A considerable advantage to this form of construction is that there is no requirement for a central inspection borehole, and the welds are completed in regions of low stress, where the weld and rotor components are not subject to the same high levels of stress normally developed in an operating unit. To complete this welding, the individual discs are stacked vertically, and then a root pass weld is applied to each joint, as shown in Figure 9.2.12, with four welding heads placed at 90-degree positions. At completion of this root pass, the rotor is turned to the horizontal position and the welds are completed. Typical weld preparations and locating spigots are shown in Figure 9.2.13. 333 TAPPS Turbine Steam Path Troubleshooting and Repair—Volume Two Fig. 9.2.11(a)—A welded rotor comprising six individual forgings joined by welding. (b) (d) Fig. 9.2.11(b), (c) and (d)—Welded rotor details for a small rated unit. 334 TAPPS (c) TAPPS The Repair of Rotating Components Fig. 9.2.12—Individual forgings ‘stacked’ and root pass welds being applied. To join the individual forgings, they are stacked vertically, the stack being checked for axial alignment, and the root pass welding is completed using four welding heads, which are located at 90degree locations around the circumference. This weld root pass is 335 Turbine Steam Path Troubleshooting and Repair—Volume Two deposited onto the four circumferential locations simultaneously. This is done to eliminate uneven local heating. The root pass is undertaken using TIG methods. After building an adequate root pass, the rotor is turned to the horizontal position, placed in a lathe, and the welding is completed using the submerged arc process. w w w w w w w w Fig. 9.2.13—Various weld preps, in each design the locating diameter is shown as ‘W-W’. The hybrid rotors. It is often convenient for the designer to employ rotors that are composites or hybrids of the monobloc, and built-up rotor forms, as described earlier. It is also possible for some rotors to be of a bolted construction in which interference, rather than shrink fits, are used to locate the individual components relative to each other. The use of these other forms is evaluated and applied where they offer advantages both in terms of performance, or the cost and delivery of material. 336 The Repair of Rotating Components In fact, these forms tend to have been used on older design units, where materials of the size and quality required were not available to support design requirements. However, certain applications are still used to advantage, in terms of costs, delivery, and quality. Rotor forgings A major characteristic of turbine rotors is that the principal components are produced by forging. For the monoblocs and central spindles of the built-up assembled rotors, the forgings are large, require a homogeneous material, and must be produced to meet the requirements for reliable operation for many years. Basic production of the forging. After pouring the melt for a rotor forging, it begins to solidify at its outer surfaces where heat is lost through the walls of the vessel containing the liquid metal. Soon after the melt is poured, it is common to take a ladle sample for chemical analysis as a check on chemical composition. This small sample may also be checked for mechanical properties. As the melt cools from the outside impurities, nonmetallic inclusion and nitrogen bubbles are forced towards the center of the cooling ingot. When solidified the core will be the region where such impurities are concentrated. Note: Modern practice involves vacuum degassing, a process by which the space above the melt is partially evacuated to assist in removing many of the dissolved gasses, which have the probability of producing “blow holes” if they are not effectively removed. After cooling to a suitable temperature, the ingot is removed, and the forging process is undertaken. For rotor forgings, the initial size of the ingot can be taken to have an overall length of “L” and a diameter “D.” The upper surface of the ingot will have deformed to a concave form, caused by the volumetric reduction of the metal as the melt cools, with the cooling having occurred first at the outer 337 Turbine Steam Path Troubleshooting and Repair—Volume Two diameters. Also, as the cooling occurs and the grains of alloy are formed, any impurities will have migrated, or have been driven, to the center portion of the ingot. The production of the basic forging ingot is made in one piece, normally using basic electric vacuum degassed steel. The ingot should be cropped to remove piping and any evidence of segregation, which could exist at its ends. The ingot is then placed in a press, either vertically or horizontally, and worked over its entire cross-section to a diameter ratio of about two to one. Care must be exercised in this forging process to ensure the axial center of the ingot remains common with the axial center of the final rotor. This is necessary to ensure impurities and other undesirable inclusions, which tend to migrate towards the center of the melt as it cools from the outer surfaces, remain central to the rotor and can be removed by boring and chambering. Or if unbored, this ensures that the impurities are in a radial position where the least adverse effects to the stresses are induced in the rotor. Before being shipped from the forge master’s plant, the rotor can be given a central bore, intended to remove any remaining impurities in the material. This central bore can also include the provision of local chambering to an engineering specified maximum diameter to remove any local impurity or blow hole concentration. Inspection of rotor forgings during manufacture. The in-process and final inspection phases are an integral part of the total manufacturing process of a turbine rotor forging. These inspections are conducted at the forge master’s works, which are performed after pouring, and also during and after rough machining. Extensive examination is also undertaken at the turbine manufacturer’s plant, where considerable examination and evaluation is completed during the metal turning. The most significant tests and examinations relative to rotor material integrity are those performed by nondestructive techniques. It is a general requirement of most turbine manufacturers that a magnetic particle examination is made of all exterior surfaces, and if the rotor has an inter- 338 The Repair of Rotating Components nal borehole, these internal surfaces should be inspected visually to the greatest extent possible. If magnetic particle inspection indicates the presence of cracks or linear defects beyond an agreed limiting value, the turbine manufacturer can reject the forging. However, such defects should be examined in terms of their axial position, local stresses, and the material that will be removed during final machining, as some defects could be removed in this manner. Also, the forging position can often be optimized so that significant indications are removed. General Electric Co. Ultrasonic examination, together with magnetic particle examination, must show the forging to be free of cracks, discontinuities, flakes, fissures, seams, and laps. It is normal for the turbine manufacturer to reserve the right, as defined in the material inspection, to reject any forging showing indications that cannot be removed without modifying the mechanical integrity of the forging. Figure 9.2.14 shows a rough machined monobloc rotor being ultrasonically examined. Fig. 9.2.14—The ultrasonic examination of a rotor forging. 339 Turbine Steam Path Troubleshooting and Repair—Volume Two An ultrasonic examination is made of the rough machined forging from all possible and available surfaces. This is done before final heat treatment and the removal of test pieces. At completion of rough machining, final heat treatment, and the removal of test pieces, the rotor is normally given a final ultrasonic examination by the forge master before delivery. If a forging is found to have a recordable defect, it is normally referred to the purchaser (also to the turbine manufacturer, and possibly the user), who has the final responsibility for acceptance, and must make an accept/reject decision based on the predicted duty of the unit, the rotor stress levels, and defect locations. It is difficult to state absolute levels of acceptability for any defect. Acceptance levels are specified by the designer/manufacturer, usually to cover both single and cloud defect clusters. Typically, a turbine manufacturer’s material specification would list acceptable defects as being an isolated defect in a critical area, and should not have its major diameter exceeding 0.04-0.08". In less critical areas, the major diameter should not exceed about 0.150.20 ", and cloud or clustered defects should have no indications whose major diameter is in excess of 0.05". Critical areas of a rotor are considered those adjacent to a blade root fastening, those within 2.0" of the bore, and for monobloc rotors with integral coupling flanges, those occurring within the coupling flange area. Central inspection boreholes. The pouring, solidification, and forging of a turbine rotor all tend towards a concentration of impurities and non-metallic inclusions at, or close to the axial center of the finished forging. In setting up the forging to be machined in the lathe, care is taken to ensure the axial center of the forging will coincide as close as practical with the axial center of the finished rotor. The reason for this effort is to place those inclusions and undesirable constituents in a physical position where they can be removed by boring. Many monobloc rotors and spindles, which are 340 The Repair of Rotating Components intended to have wheels or discs mounted or produced on them, are bored along their entire axial length. This boring is undertaken before final machining to ensure the bored hole remains central during subsequent machining and metal removal, and does nothing to adversely affect the dynamic balance, which can be achieved with the rotor. Normally, during the boring process, portions of the core are trepanned to remove suitable test pieces, which can be analyzed for both chemical composition and mechanical properties. With the introduction of vacuum degassing, and the general improvement in material production technology, some manufacturers now place sufficient confidence in their rotor forgings that they do not bore. However, these manufacturers will undertake extensive NDT to ensure their specifications concerning material integrity are satisfied. Defects can go undetected, as illustrated by the failure (during unit start-up) of a rotor forging in Figure 9.2.15, where a cloud defect has precipitated a massive rupture of a high-pressure forging. Fig. 9.2.15—Showing a rotor forging which has failed, having a defect near its center location. This defect has initiated a major rotor failure. 341 Turbine Steam Path Troubleshooting and Repair—Volume Two Borasonic examinations. The borasonic examination provides a means of determining the condition of rotors, especially those that have been in service for a period of time, at high temperatures, and in which there is some level of concern regarding the possibility of cracks having developed in the rotor material. Many rotors have been placed in service with known “cluster” or “single” void defects close to their center, and even nonmetallic inclusions. These defects met the engineering specifications in place at the time the rotors were produced, and represented good engineering judgment in terms of the manufacturing capabilities of the forge master. While these defects were acceptable at that time, it is necessary for the owner to monitor material condition to help ensure continued satisfactory service. Turbine rotor discs In many low-pressure sections, because of the increase in steam volumetric flow, the blades must have an increased radial height and must also be carried on rotors at a significantly increased diameter. It has for many years been impossible for the forge master to produce a suitable rotor of monobloc form to meet these requirements. One alternative construction discussed previously is shown as Figure 9.2.10. This form of construction was used to increase the effective stage diameter. With current forging technology, monobloc rotors can be produced to meet many of these larger diameter requirements. However, there are currently many units in service with this older construction, and these units must be maintained and must continue to operate for many more years. For the built-up form of construction, many of the considerations concerning rotor design are also applicable to discs. However, a further duty imposed on a disc is the stresses induced in the disc by the shrink fit between the spindle and disc. 342 The Repair of Rotating Components Functions of the disc. The discs form part of the rotor, and are attached to it through a shrink fit. This fit is sufficient to locate the disc both axially and radially, and hold it in intimate contact with the central spindle. The functions and requirements of the disc can be summarized as follows: • To carry the blades of those rows attached to it, and to transmit the force developed on them to the central spindle. The torque developed on the discs is transmitted to the central spindle by the frictional fit of the disc on the spindle. Locating keys are not designed, or intended, for this torque transmission function • To be capable of withstanding the stresses induced in it by its own weight, and the weight of the components it carries. The disc must also be able to withstand the shrink fit stresses • To be capable of withstanding the temperature and pressure gradients developed across it, the material having a high resistance to both rupture and creep • Because discs are used in the low-pressure, low-temperature regions of the unit with large volumetric flows, the steam conditions in this region are normally operating in a wet steam environment in their latter stages. Therefore, it is necessary for the material from which the discs are produced to be capable of resisting water erosion and also, as far as possible, resist any corrosive action associated with the corrodents that come out of solution in the wet region. Figure 9.2.16 shows the face of a disc in the moisture region of a geothermal unit, which has suffered material loss from a combination of moisture washing, erosion, and corrosive attack Forms of the discs. The disc is assembled onto the central spindle through a shrink fit, and is held in the correct axial and tangential location by keys, plugs, and/or retaining rings. These restraining 343 Turbine Steam Path Troubleshooting and Repair—Volume Two Fig. 9.2.16—The face of a wheel from a geothermal unit that has suffered material loss due to water being centrifuged radially out along the wheel face. devices make no contribution towards transmitting the torque developed on the disc to the central spindle. The torque transfer is achieved through the frictional shrink fit at the disc/spindle interface. The disc shape is influenced, and to a degree, defined by the loads developed on the stage and the general design requirements of the unit. In its simplest form, the disc connects the blades to the spindle, shown in Figure 9.2.17(a), with a shrink fit existing at the surface “aa.” In this figure the shrink fit is shown to exist along the complete axial width of the disc. There are, however, designs having a shrink fit only at the outer edges of the interface, as shown in Figure 9.2.17(b), with a central relief of the disc bore. For multi stage rotors it is necessary to provide stationary blade rows between the rotating rows. For this reason, it is necessary to design these stationary rows, which have a pressure drop across them, 344 The Repair of Rotating Components center relief a a c a a a Shaft seal (a) Center of shaft (b) Fig. 9.2.17—Characteristics of “shrunk-on”wheels. Figure 9.2.17 In (a) is a single wheel, and in (b) a “pair” with a diaphragm containing a shaft seal between them. and in (b) a "pair" with Characteristics of "shrunk-on" wheels. In (a) is a single wheel, a diaphragm containing a shaft seal between them. to provide a steam sealing system between the rotating rows or discs. These seals are intended principally to minimize leakage, and to a lesser extent to guide the steam and ensure it follows the designed steam passage. The location of a typical seal system attached to a stationary blade row is shown in Figure 9.2.17(b) and 9.2.18. Interstage Seal Systems Figure .2.18seals. Fig. 9.2.18—The interstage The interstage seals. 345 Turbine Steam Path Troubleshooting and Repair—Volume Two Due to the stress levels in the discs of modern, large rated units, it is rarely acceptable to use a straight or parallel-sided disc, as shown in Figure 9.2.17(a). For these larger output units, the disc profiles are shaped to minimize stress levels. For multi disc designs, a step change in spindle diameter normally exists between discs, as shown in Figure 9.2.18. This step provides a vertical shoulder, which aids in locating the disc in the correct axial position. This radial step also reduces the distance over which the disc must travel when being assembled, when contact between the disc inner surface and spindle diameter (at interferencefit diameter) could cause “chilling,” and cause the disc to “grip” the wheel, making the completion of assembly impossible without first removing the disc and reheating. This is a long, expensive, and difficult process, and one that can cause damage to the shrink surfaces. There are other considerations of disc geometry, which need to be considered: • Interstage seals—To permit a satisfactory seal to be provided between the rotating blade rows, the discs are designed so a seal surface can be provided. This seal can be produced at the hub of the discs, as shown in Figure 9.2.19, or on an axial projection from the wheel, as shown in Figure 9.2.20. The sealing device used depends upon the design details of the stage. However, the most effective seal design will employ a large number of seal strips, and have the seal formed at the smallest diameter possible In selecting the seal form to be used, the design engineer will evaluate the alternatives, and select the one that provides the best return on manufacturing costs against leakage loss. As shown in Figure 9.2.20, one possible advantage of using a blade root seal on either the inlet or discharge side, is that it could be possible to reduce the “rotor span,” by the elimination of a diaphragm inner 346 The Repair of Rotating Components Diaphragm Inner Web Disc 1 Interstage Seals Disc 2 Shaft Shrunk on Locking Ring Fig. 9.2.19—The interstage seal system. In this design forming a seal on the cylindrical faces of adjacent discs. Fixed Blade Row Moving Blade Row Radial Clearance Fig. 9.2.20—The basic seal formed on an axial projection from the disc. This design is used in certain low pressure applications. 347 Turbine Steam Path Troubleshooting and Repair—Volume Two web. However, this can only be done when the disc stresses permit a reduction of the axial span of the discs. • The disc rim—At the rim of the disc, provision is made by the production of a root-fastening slot to enable the rotating blades to be securely attached to it. Most discs carry only one blade row. However, in some designs, the earlier stages can have more than one blade row on a single disc, with the stationary blades and their seal located between the rotating blades. Such an arrangement is shown as Figure 9.2.21. This is a suitable construction and permits several rotating blade rows to be carried in a shorter axial span Axial Clearance Axial Clearance Radial Clearance Fig. 9.2.21—Two rotating blade rows carried on a single disc. This arrangement helps reduce the overall length of the rotor, therefore reducing its bending stress. The disadvantages with this form of construction are the steam leakage seal between the blade rows is made at a larger diameter, thus providing greater leakage area, and the pressure and temperature differentials across the disc 348 The Repair of Rotating Components are increased from a “1 stage condition” to a “2 stage condition.” While these factors represent disadvantages, they introduce no significant problems if geometries and stage arrangements are chosen and fully evaluated. If the discs are in an axial plane that is used for rotor balancing, and are to have balance weights attached to them, such weights are normally attached at an outer radius close to the blade root. • Pressure balance holes—The discs, if they are to be produced with pressure balance holes, the hole geometry must meet the same requirements for finish and location as for a monobloc rotor. They must also be produced without discontinuities or surface gouges in the holes One advantage for the individual discs, however, is that there are no constraints as to the diameter at which the holes are produced, since these holes can be produced before the discs are mounted to the central spindle. Many built-up assembled discs do not have pressure balance holes. These are omitted for two reasons: – The first being that these are often on double flow, lowpressure sections, where the axial thrust is balanced, and sudden changes in pressure would not significantly modify the axial thrust from one end to the other. The exception to this is that should mechanical damage to the flow on one end occur, there would be a thrust unbalance, but the thrust block would not be sized to cope with a single incident – The second consideration is that on many low-pressure stages, the stage is designed with a higher degree of reaction in them than is normally experienced on a rotating blade row. Therefore, including a balance hole would promote the flow of a considerable amount of steam 349 Turbine Steam Path Troubleshooting and Repair—Volume Two through the hole, causing it to bypass the rotating blades and generate no power • Disc to hub fillet radii—Certain discs carrying large centrifugally heavy blades have high stresses developed in them as a consequence of their size, and the blades they carry. For this reason the disc profile conformance is critical in establishing the stress distribution Machining for shrink fits. The disc forging is received from the forge master in a rough machined condition, and will have been examined by nondestructive methods to ensure its mechanical integrity. The turbine manufacturer has to finish machining the disc and the spindle to ensure the correct shrink fit is obtained. The disc is bored to the design diameter, and care is taken to ensure the surface finish is at design specification. It is necessary to match individual disc bores to spindle at diameters, ensuring the bore is circular. This is done by measuring at least two, and probably four diametral positions. Readings are taken from the machined bore to ensure it is both concentric and perpendicular to the disc wall faces. These requirements are important to ensure the final assembly will be concentric, disc to spindle, and the disc will be at right angles to the spindle diameter, as shown in Figure 9.2.22. An alternate method is to heat the disc, and then to lower the spindle into the disc. Both methods are successful. Machining the shrink fit is normally an individual machining operation. The disc bore is machined, measured to establish a mean bore diameter, then the spindle is finished to achieve the required shrink fit. Discs with either cracks initiating at the “keyway,” or discs requiring a modification to the key and keyway geometry to avoid cracking, have commonly been repaired by removing the disc, mounting a collar with a small shrink fit sufficient to transmit the steam induced forces in the blades and disc, and then remounting the re-bored disc over the collar. This arrangement is shown in Figure 9.2.23. Such a design 350 The Repair of Rotating Components 90° ° Hot Clearance Fig. 9.2.22—Lowering a disc onto its central spindle. The disc must be maintained at a 90° position to permit assembly. change cannot be initiated without an investigation of the stresses induced in the re-bored wheel, and if material is removed from it, an investigation of the spindle. However, this does offer an opportunity to correct what could become a significant stress and high-risk situation. Disc shrink fits. The shrink fit used for any stage is selected to achieve certain design requirements: • The stresses must not exceed the allowable values of the material at its operating temperature 351 Turbine Steam Path Troubleshooting and Repair—Volume Two • The shrink fit must be sufficient to maintain contact at all loads that will be developed in the blade row. The shrink fit reduces as speed increases • The shrink fit must be sufficient to secure the disc onto the spindle, and transmit torque without any tangential slipping • The shrink fit must be maintained at all rotor speeds up to emergency overspeed (118-120%) before the fit is lost Disc Central spindle Shrink collar Keys 9.2.23 to allow a disc with center bore Fig. 9.2.23—Showing the use of a Figure “shrink collar” cracking to beShowing reused. the use of a "shrink collar" to allow a disc with center bore cracking to be reused. Vertical assembly of the disc onto the spindle. To assemble the discs onto the central spindle requires a shrinking operation. This process consists of expanding the discs sufficiently, by heating, to overcome the interference fit (diametral difference) between the spindle diameter and the disc bore, then allowing the disc to settle over the spindle in a correct axial alignment. In order to achieve the correct alignment of the disc to the spindle, it is necessary to arrange the spindle in a vertical plane, for long 352 The Repair of Rotating Components double flow rotors in a pit, with sufficient headroom above, for the crane to lower the disc without any form of interference. This general arrangement for such an operation is shown in Figure 9.2.24. (An alternative method, but one rarely used now, is to lower the spindle vertically into the heated disc where the spindle weight ensures contact between the disc hub face and the shoulder machined onto the spindle.) The assembly then cools. Crane hook. 90° Clearance between spindle and disc Disc heated to achieve an acceptable clearance from disc to spindle Central spindle held vertically Fig. 9.2.24—The assembly requirements for lowering a disc onto the central spindle after temperature requirements have been met. 353 Turbine Steam Path Troubleshooting and Repair—Volume Two To achieve an acceptable assembly, it is necessary to arrange the spindle vertically. Any attempt to shrink in the horizontal plane would eventually mean contact between the cold spindle and the hot disc at one tangential position, and almost certainly before the disc was in the correct axial position. This would cool the disc locally over the contacting surface, due to heat conduction, which would cause uneven cooling, local shrinkage, and distortion of the disc. Any building that is used for this disc assembly process, whether a pit is used or not, should be designed to be relatively free from drafts, as these could cause uneven cooling and cause the disc to “cock” on cooling. The disc is preferably heated in an oven under controlled temperature conditions. The disc should be raised in temperature slowly, in an attempt to keep the temperature of the whole disc fairly even. Under no circumstances should the disc be heated above a specified maximum value. This value is normally about 750°F, the exact value being determined by design requirements. In the event this is the reassembly of an existing disc, with blades already fitted to the disc, and with brazed coverbands, tie wire, or erosion shields, these components should be shielded by suitable insulating baffles, and no direct impingement of hot air on them should be permitted. The disc is heated until the desired bore expansion is achieved. At completion of this heat soak, there should be no temperature differential from one part of the disc to another in excess of about 45°F. When the desired expansion of the disc has been achieved, it is removed from the oven, and connected by previously selected slings to ensure it can be adjusted to a level condition. The crane then raises the disc above the spindle. Some manufacturers will oil the spindle surface on the area to have the disc assembled. This is done to facilitate final positioning. If oil is used, care must be taken to ensure 354 The Repair of Rotating Components the use of mineral or vegetable oil contains no corrosive compounds, or other elements that could degrade or modify to form them. The disc is located above, and then lowered slowly into place, avoiding any tilting, ensuring the disc is at 90 degrees to the axis of the spindle. The crane positioning is critical in this operation, and should any tilting occur, the disc should not be allowed to touch the spindle surface for any significant time, otherwise it will quench locally. If significant contact occurs because of poor crane position, the disc must be raised, the tilt corrected, and lowered again. When the disc hub seats onto the shaft shoulder or retaining ring, it should be determined that the spindle diameter and disc bore are concentric, alignment is correct, and any locating keys or pins that are not accessible from the upper face are in place before the final lowering of the disc. As soon as the disc is in its correct position, the underface can be cooled. This cooling should not be undertaken too rapidly and any air used must cause even cooling around the entire circumference. By cooling from the underside, the lower edge of the disc hub grips the spindle first. This cooling, if undertaken, should commence within 15 to 20 minutes of the disc seating on the spindle. This will allow sufficient time to adjust the hot disc to a concentric position, and make any necessary shim adjustment. To aid in locating the disc and maintaining a correct axial gap between discs, axial shims can be used to ensure correct alignment. These shims are preferably heated to within 100°F of the disc temperature, and should not project down beyond the mid portion of the hub. A typical arrangement for shim placement is shown in Figure 9.2.25. These shims must be removed at completion of shrinking. As the assembly cools the shims will normally become loose. When the disc has cooled sufficiently, it should be checked to ensure squareness from hub to spindle faces. 355 Turbine Steam Path Troubleshooting and Repair—Volume Two Spacer shims Fig. 9.2.25—Assembling discs to the central spindle and using ‘shims’ to locate the disc. Removal of discs. There are certain situations where discs previously assembled to spindles will have to be removed to allow some form of corrective action to be taken. These situations include: • If, after the initial assembly, the disc does not meet engineering limitations for squareness, an attempt to resettle the disc can be made. This can require the removal of the disc, and its repositioning To resettle a disc, it is necessary to heat it until the shrink fit is released. The disc should be heated uniformly with gas ring burners. The number and location of gas rings depends upon the proportions of the disc. Figure 9.2.26 shows two typical arrangements. During the heating process, the maximum temperature differential from one 356 The Repair of Rotating Components part of the disc to another should not be in excess of 300ºF to 600ºF. After re-settling, the disc should again be checked for squareness. If on resettling, the disc does not meet squareness specifications, it must be removed. Note: There are limits to the amount of time a disc can be heated before temperature embrittlement of the disc occurs; the higher the temperature, the shorter the exposure time. The time and temperature selected will depend upon the disc material specification. Gas heaters Gas heaters Fig. 9.2.26—Arrangements for heating ‘shrunk-on’ discs prior to removal. • In addition to the requirement for the removal of initially assembled “cocked” discs, there are also instances where, after some years of operation, it may be necessary to remove and replace, or refurbish a disc. This occurs when the disc of a unit that has been in service needs to be removed to allow inspection or re-boring to employ a collar, as shown in Figure 9.2.23. Unfortunately, it is occasionally necessary to remove several unaffected discs in order to gain access the disc in need of repair or inspection 357 Turbine Steam Path Troubleshooting and Repair—Volume Two For those units that have been in operation for some years, it is necessary first to clean the rotor of any chemical deposits, dirt, or grease that may have accumulated on it and could interfere with the removal of the disc. This interference can occur because it makes the task of heating the disc, without excessive heat being transferred to the central spindle that would make it expand, more difficult. In addition, any shrunk on details, such as gland sleeves or couplings that would interfere with the removal of the disc, must be removed. There are situations where a tight shrink is present, or large quantities of scale or other deposits exist, and it may be necessary to cool the central spindle of the rotor in order to release the shrink fit. This is particularly so for rotors that have been in operation for some time, and on which chemical deposits may act to conduct heat from the disc to the spindle, making it difficult to achieve the desired temperature differential between the two components. In such a situation it might be necessary to remove one plug from the end of a bored spindle, and connect clean water inlet and outlet pipes, as shown in Figure 9.2.27. When the disc has been removed, the spindle must be drained and dried with suitable materials and the plugs replaced. Occasionally, for tight interference fits, or in the case of excessive scale, a liquefied gas, such as nitrogen, has been used to cool the spindle. However, this practice is not recommended and should be avoided if at all possible. If the central spindle does not contain a borehole, the cooling of the spindles becomes difficult, or impossible, in the case of discs that have only a small axial clearance between them. Disc keyways and securing. The discs transmit the force as torque from the blades to the central spindle by means of the shrink fit friction. No means are required to maintain the disc in its original assembled location. However, a key (or other tangential locating device) is required to locate the disc on the spindle, and to provide positive tangential location of the disc during an overspeed transient when the shrink fit could be lost, and when the driving torque will be removed. 358 The Repair of Rotating Components Cooling Medium Gas Vent Coupling Removed Gas Burners Bore Plug Support Brackets Fig. 9.2.27—Method for removing ‘tight’ discs by cooling the central spindle. Most discs are designed so that up to about 118-120% overspeed they will expand elastically, and will maintain their shrink fit. Above this speed, the shrink can be lost. Therefore at this speed, a key or locating device must continue to hold the disc in a correct 359 Turbine Steam Path Troubleshooting and Repair—Volume Two axial alignment, and also prevent tangential migration, which could affect the balance of the rotor. At speeds considerably in excess of the loss of shrink fit, the disc becomes eccentric, causing a large rotating imbalance, which could cause rubs, and eventually disc and rotor failure. With early designs, the method used for locating the disc to the shaft was a keyway design, such as shown in Figure 9.2.28. The portion in the disc being square, as shown in Figure 9.2.28(a), although the key could have a small chamfer. Figure 9.2.28(b) shows another form where the keyway has radiused corners, and the key is chamfered. Figure 9.2.28(c) shows the final form, using the same form of keyway within the spindle, which was not at risk, and a semi-circular form in the disc. Unfortunately, these forms of keyway have high stress concentration at their corners, and have led to numerous failures. These keyway constructions are also a common “hideout” for corrosive products, which are capable of causing stress corrosion cracks to initiate at the corners of the disc keyway. In modern units the tendency is now away from the simple keyseat, and towards replacing systems using low stress areas for the point of attachment. The button type locator (shown in Fig. 9.2.29) uses a circular pin to locate the disc to a central collar produced integral with the spindle. Subsequent, downstream discs are located to the inner ones. The locking screw (shown in Fig. 9.2.30) attaches discs together. A key or button would be used with the inner disc. The surface finish and fillet radii of the keyseat are critical. Attention must be paid to ensure all fillet radii are smooth and continuous. Any machining marks produced on the disc bore surfaces must be dressed smooth. The internal finish of buttonholes is not nearly so critical. However, any burrs, tool marks, or other surface indications should be dressed. To prevent axial movement, some manufacturers use a retaining ring to locate the disc. This ring, shown in Figure 9.2.31, is either 360 The Repair of Rotating Components Key Shaft (a) Key Shaft (b) Key Shaft (c) Key Shaft Fig. 9.2.28—Forms of the shaft to disc key. Screw Attachment Connecting plug disc 1 disc 2 Disc 1 Disc 2 Spindle Fig. 9.2.29—Button connection from the disk to a central collar. Fig. 9.2.30—A locking screw attaching disc 1 to disc 2. 361 Turbine Steam Path Troubleshooting and Repair—Volume Two Seal System Clearance The retaining ring Fig. 9.2.31—A retaining ring preventing axial migration of the disc. shrunk in place or, as is often done, the ring is of a split type, which is held in radial position by the disc. This retaining ring is positioned as the disc is shrunk onto the central spindle. COVERBAND DAMAGE, REPAIR, AND REFURBISHMENT METHODS During operation, the coverband is subject to a variety of stresses due to its own weight and the forces imposed on it by the attempts of the blade to vibrate. Many of the damage phenomena that affect the blade have a similar influence on the tenons, and tenon hole regions. The most common forms of coverbands and tenon damage will be considered. 362 The Repair of Rotating Components Impact damage The coverbands and fastenings formed from the tenon, which are produced from the blade vane material, can be subject to impact damage from solid-particle debris, and erosion by water or oxide scale. These various impacts occur as the particles are transported over the coverbands by the steam. Such impacts can remove material from the tenon heads, and this material loss can continue until, eventually, it weakens the clamping effect of the tenon, and the head has insufficient material to restrain the coverbands in place against the shear forces introduced by the centrifugal force of the coverbands itself. Solid-particle impact damage. Impact type damage results from particles generated either within the steam path itself from detached Fig. 9.3.1—Impact damage on the tenons and coverband. 363 Turbine Steam Path Troubleshooting and Repair—Volume Two components, or as a consequence of debris carried over from the boiler and steam leads. Damage of this type is shown in Figure 9.3.1, and appears as a removal of tenon material at the front edge of the head with material deformation, affecting the leading edge of the tenon. Had a foxholed tenon been used in this instance, the damage would have been of less significance. Water impact damage. Water carried over from the preceding stationary blade row, as relatively large droplets will impact with the blade vane inlet edge, and also the coverband. This water tends to be centrifuged out, and carried between the coverband and casing inner surface. Water above the coverband may also rebound between the casing inner surface and the coverbands if the radial clearance is small. The water in the radial gap can remove material from both the tenon and coverband. Normally coverband’s material loss, as shown in Figure 9.3.2, is not significant. However, the loss of tenon material can have serious consequences. Figure 9.3.2 shows a tenon with material removed by water-impact erosion. This tenon head of Figure 9.3.2 is Fig. 9.3.2—Moisture impact erosion on the coverband and tenons. The material loss is becoming severe, and could result in the loss of a coverband segment. 364 The Repair of Rotating Components Fig. 9.3.3—Tenon material loss due to moisture impact erosion. not recessed; therefore, the loss of head material can become serious, possibly leading to the loss of a coverband segment. Shown in chapter 3, Figure 3.8.14 shows a similar loss pattern, but in this case with a foxholed tenon. Here the material loss is less significant, but can become serious if there is a loss sufficient to weaken the clamping effect of the tenon head of Figure 3.8.14, or if there has been enough material loss that the vertical surface of the tenon hole is beginning to be uncovered. The same loss in a fox holed tenon is shown in Figure 9.3.3. Solid-particle erosion damage. The considerations relating to the loss of material due to solid-particle erosion (SPE) of the blade vane apply to the coverband also. Figure 4.8.12 of chapter 4 shows the mechanism by which scale migrates to, and is carried over the coverband, and therefore is able to remove from the tenon by SPE. The loss of tenon material and exposure of the tenon hole vertical surface will reduce the clamping effect of the tenon, and therefore weaken the strength of the attachment. The frictional fit between the tenon material and the tenon hole vertical surface is expected to be sufficient to hold the coverbands in place when the unit is new. 365 Turbine Steam Path Troubleshooting and Repair—Volume Two But as the unit ages, if the blades have been subjected to high alternating loads, this can weaken the attachment. The attaching frictional force reduces, and clamping is then maintained by the shear strength of the tenon head material. Therefore, material loss of the tenons can reduce the clamping capability of the tenon. As material loss from the tenons continues, and the clamping effect is lost, the most common observation of this damage is for the coverbands to begin to lift on the leading blades in each group. This is because most often the bending moments of the coverbands are a maximum at this location, and the material loss is normally at the leading surface of the tenons. The gap between the blade platform and the underside of the coverbands should be checked. One method of minimizing these types of material losses, and one used by several manufacturers, is to employ a recessed or “foxholed” rivet, as shown in Figures 9.3.4 and 9.3.5. In this arrangement, the coverband has the rivet head formed internal to the foxhole, which is below the outer surface of the coverbands. Although the outer portion of the tenon head can still be eroded, it will take a longer period of operation to remove enough material so that the integrity of the attachment will be influenced to the extent that corrective action is required. Figure 9.3.5 shows the geometry of such a foxholed tenon, in which the tenon has been protected by the thicker coverband. If material loss has occurred to the extent the integrity of the attachment is in jeopardy, a new design that employs the recessed (foxholed) head can often be introduced. This new attachment can be achieved in one of two forms: Coverbands of the same thickness. This is achieved by using the coverbands that have been removed, without sustaining damage, or the use of new coverbands of the same thickness, and producing the recess or foxhole in the existing thickness. This is shown in Figure 9.3.5(a). If this is intended, the engineer should make certain checks, 366 The Repair of Rotating Components Cl R H Fig. 9.3.4—The recessed coverband. ‘H’ is the thickness, and ‘R’ the depth of the recess. Lip s s t s t To Hn s Ho t Vane t (a) Mb Tn Mb (b) Fig. 9.3.5—Showing the geometry of the ‘fox-holed’ tenon for both normal and thickened coverbands. as the effective thickness of the load bearing portion of the coverbands has been reduced from “To” to “Ho.” These checks include: • The stresses in the reduced thickness section “Ho” of the coverband should be evaluated. This will include the shear stress on the coverband at the inner overhang, across “t-t,” and the bending stress due to “Mb” on the reduced thickness • It will also be necessary to consider the shear stress in the tenon, on section “s-s,” if it is to be formed from the original material, without any form of weld rebuild 367 Turbine Steam Path Troubleshooting and Repair—Volume Two Note: Because of the difficulty in defining the tenon head thickness “s-s,” some manufacturers prefer to use a loading per inch of tenon perimeter. This is calculated by determining the centrifugal force of one pitch of coverband, and dividing by the tenon perimeter. • If the blade material contains tungsten, it may be necessary to undertake the re-peening in the hot condition Coverbands of increased thickness. If the original coverbands aren’t thick enough to allow a recess to be formed, it may be necessary to use a thicker coverband. This increased thickness is shown in Figure 9.3.5(b), where the thickness has been increased to “Tn” with an inner ledge thickness of “Hn.” In this case the thickness has been increased so that “Hn” is equal to the original thickness “To,” but is of a sufficient depth “Hn” that a protected head can be formed. The stresses in the coverband and tenon should be checked for acceptability. Other forms of coverband damage In addition to those forms of damage resulting in the loss of tenon head material, other forms exist. These include: Excessive overspeed. Should a turbine experience an excessive overspeed transient for an extensive period, or the tenon head has deteriorated, it is possible that the coverband could detach. Shown as Figure 9.3.6 are tenons where the cover has detached. In this case, because of the form of the remaining tenon head material, it is possible the original clearance between the tenon and coverband hole was excessive, preventing a full head, and filling the hole that was formed. Heavy radial rubs. Figure 9.3.7 shows a coverband that has sustained a heavy rub at the end pitch. When such a rub occurs, the material is immediately heated by the frictional force that is developed, and just as quickly “quenched.” This heating/quenching cycle causes the material to become brittle and easily cracked. Chapter 4, 368 The Repair of Rotating Components Fig. 9.3.6—Blades in which the coverband has detached. The centrifugal action of the cover has overcome the clamping effect of the tenon head. Fig. 9.3.7—A coverband that has sustained a heavy rub on its thin inlet edge. 369 Turbine Steam Path Troubleshooting and Repair—Volume Two Figure 4.11.6 shows the same row as shown in Figure 9.3.7, where a portion of the coverband inlet edge has detached. Axial rubs. Axial rubs are a common occurrence, and are most often present in the axial gap between the stationary blade discharge and the rotating blade inlet. At this location the coverband can form a knife-edge seal (Type B coverband), and when a rub occurs only the knife-edge is destroyed. This causes only minor damage to the cover, which is unlikely to lead to mechanical rupture. Such a rub is shown in chapter 4, Figure 4.11.5. Cracks initiating at the coverband hole. A relatively common form of failure is the development of a crack from a coverband hole. There are a number of possible causes, which include a poor finish to the coverband hole, a poorly designed tenon form, requiring the Fig. 9.3.8—A crack initiating at the tenon hole, and progressing across the coverband. This coverband will eventually detach if not replaced. 370 The Repair of Rotating Components coverband hole be too sharp, causing high stress concentration, or even over peening when the tenon head was formed. If such a condition is found when a unit is inspected, the coverband should be replaced. Figure 9.3.8 shows such a crack. While it cannot be determined from this figure where the crack originated, it is suspected it formed at the acute angled long tenon, and then propagated by highcycle fatigue. Examination of the fracture surface would confirm this. The reforming of tenons If the tenons have sustained damage, or have eroded to the extent the integrity of the coverbands’ attachment is suspect, there are obvious cost advantages to refurbishing the tenon material so the blades can be reused. There is an even greater advantage if this can be done without the cost of removing the blades from the rotor. Such refurbishment can be undertaken in certain circumstances, and the tenon can be returned to an acceptable condition. To undertake this refurbishment it is necessary to remove the existing coverband, which is done by making a series of axial or semi-axial cuts, and removing the material from under the rivet head. This is an operation requiring considerable care, and must be undertaken using hacksaws or cutting discs to cut through the thickness of the coverbands, without cutting either the tenons or the blade at its outer section. There are then available several options for reforming an adequate tenon: Reworking the original material By reworking the tenon material, it can be reformed so a new coverband can be passed over it. This is shown in Figure 9.3.9. This tenon material is then used to reform a rivet head. After removing the coverbands, the tenon can be reformed to achieve a total height of “Ht,” with a height above the coverbands of “Hs.” It is then necessary to evaluate the stresses and general geometry to determine if foxholing is required. 371 Turbine Steam Path Troubleshooting and Repair—Volume Two Drawn out tenon material used to reform a new rivet head Hs Fig. 9.3.9—Form of the tenon after the coverband has been removed. Dependent upon the blade material, it may be necessary to heat treat the tenon material to ensure it can be reworked without cracking. It may also be necessary to remove sharp edges to minimize the possible effect of crack formation in the “thin” edges. In this situation it may be necessary to preheat the tenon before peening. Weld deposit on the existing tenons With modern welding technology, it is possible to rebuild the tenon as shown in Figure 9.3.10. This process has been a most successful method of extending blade life, and can be undertaken without degrading the integrity of the attachment. This weld repair method can be undertaken using either a material that is compatible with the blade material, or one of the Inconel family of materials. The Inconel materials do not require a stress relief operation. However, there is normally a need to relieve the stresses from the blade material. The weld material chosen in any particular situation depends principally upon the operating temperature of the stage. Inconel repair can be undertaken with the blades in-situ. A steel weld buildup often requires the blades be removed from the unit to allow the level of pre- and post-heat treatment control required to achieve material properties and stresses relief. However, with care even a 372 The Repair of Rotating Components Excess mold thickness Finished form of tenon Weld deposit Copper mold (a) Copper mold. Weld deposit Lip Design tip platform diameter (b) Copper mold ToFinished Ho form of tenon. Weld deposit extends into the blade vane, which is then dressed to achieve the design tip diameter. This method removes the HAZ into the vane material. Fig. 9.3.10—Forms of the weld rebuild. In (a) the tenon has been rebuilt. In (b) it was necessary to remove blade tip platform material also to achieve a satisfactory tenon. steel rebuild can be stress relieved with the blades still mounted to the rotor. The weld rebuild method for tenons The weld rebuilding of tenons is a mature procedure, and can be used with considerable success in many applications. However, 373 Turbine Steam Path Troubleshooting and Repair—Volume Two although the techniques are known, there is still considerable skill required to complete this process and produce tenons that meet the requirements of the original design, in terms of material properties, dimensions, and therefore stress levels. There can be a considerable degree of complexity depending upon the weld material used in any application, and if the process is not adequately controlled, there can be some level of distortion, twisting, or lean of the blades. The details of the weld repair for any blade should be established in terms of the stage material, the row operating temperature, and the form or geometry of the tenons themselves. It is possible the weld can be undertaken using the shielded metal arc, or gas tungsten arc process. The general steps include the following: • Removal of the existing tenon material sufficient to allow the tenon to be rebuilt Cl Original tenon outline Cover band δ Sub height ' δ' required to remove the HAZ from region of deformation, and prevent cracks forming upon riveting Fig. 9.3.11—The material to be removed from the existing tenon, down to a height “δ”. 374 The Repair of Rotating Components The tenon may have lost only a small amount of material. However, after removal of the coverbands, it is considered necessary to remove material down to a sub-height of “δ,” as shown in Figure 9.3.11. This metal removal allows the HAZ (from welding) to be formed low on the neck of the tenon, and in an area that will not be deformed, or at least suffer a minimal amount of deformation from the riveting process. There are instances when it could be advisable to remove material from the blade platform as shown in Figure 9.3.10(b). This is something that should be determined in terms of the tenon shape and the thickness of the coverband. • After grinding the tenon down to the sub-height “δ,” the remaining tenon material and blade outer tip surface should be cleaned. This can conveniently be completed using a grit blast or polishing • The tenon and blade tip region area should then be examined by NDT methods to ensure no cracks are present in the blade vane • Establish the amount of material to be deposited to reform the new tenon. This requires a definition of both the cross section and radial height of the new tenon. The tenon cross section can be determined from the existing material, after grinding down to the sub-height “δ.” The radial height must be established, and be sufficient that a suitable head can be reformed by riveting. The defined requirements are shown in Figure 9.3.12. This height normally provides 0.090" to 0.11" above the coverband • Prepare a copper mold that will be placed over the existing tenon stub and guide the weld deposit process. The molds should be designed to have sufficient clearance around the old stub, Figure 9.3.13, that the reformed tenon can be shaped after the mold is removed. These molds should be thicker than 375 Turbine Steam Path Troubleshooting and Repair—Volume Two Tenon material for forming rivet head Cover thickness Tenon height Clearance tenon to cover Fig. 9.3.12—Details of the material to be left to form a rivet head. Tenon/Mold clearance Required final form of tenon Mold thickness Form of original tenon Fig. 9.3.13—Details of the ‘copper mold’ which is required to form the weld rebuilt tenon. 376 The Repair of Rotating Components the required final height of the tenon by about 0.110" to 0.150". The molds can take several forms, either forming the complete shape of the tenon to be rebuilt with a location to an adjacent blade tip, or simply the tenon form with no location The most applicable form in any stage is dependent upon the process to be used, the stage location, and to a degree the welder preference and experience. • Before welding commences, the blade must be preheated to ensure an acceptable weld. The amount of preheat temperature to be achieved is dependent upon the blade vane and the weld material. It is necessary for this preheat to be maintained throughout the weld deposit process. The blade vane should be preheated for a length (at least 2.0") of the vane that will ensure no cracks form as the vane cools • In making the weld deposit, it might be necessary to employ chill blocks. This can be determined from the vane geometry and mass. In making the deposit, the molds should be placed over the tenon stub at its center, and the weld deposit made around the entire tenon. The procedure for weld deposit will depend in part on the weld process, “shielded metal arc” (stick), or “tig.” It might also be necessary to remove the mold and clean off the slag from the deposited weld metal before proceeding. This is a procedure that can be established before work proceeds • At completion of the weld deposit, remove the mold, deslag, wrap the blade tip in an insulating material, and allow it to cool to room temperature • The tenon may, depending upon the weld metal, require a stress relief process, of which there are several acceptable methods. These include the use of a torch, in which the flame is kept in constant motion over the tip section, the temperature 377 Turbine Steam Path Troubleshooting and Repair—Volume Two being monitored by means of “tempil sticks.” This is normally only considered suitable for short time operations up to about 20 minutes Electric resistance heating is also very suitable, and can be used when a precise temperature is required for a longer period. At completion of the heat treatment, the blades should be covered with an insulating material and allowed to cool to room temperature. • Reshaping the tenons requires care, and is relatively complex, particularly if the blades are still mounted in the rotor. If the blades have been removed, it is possible to use machine tools. But on the rotor handwork is normally necessary to establish the final form, and the clearances between the tenons and coverband holes. It is normal to remove the coverbands intact (to the greatest extent possible) so they may possibly be used as a template in producing the holes in the new coverbands. Prior to producing the tenons and new coverbands, it is necessary to establish the tolerances that are required for the coverband to tenon clearances It is important that fillet radii at the base of the tenon are formed to remove any sharp sections or discontinuities of form, which could induce stress-concentrating regions. Also, the surface finish must be fine, and the radius not so large there will be any interference between the fillet radius and coverband underside chamfer. • At completion of the tenon rebuild and reforming process, it is recommended a final NDE be completed of the tenons before riveting commences. Figure 9.3.14 shows a rebuilt tenon, where the weld has not fused at one radial location. This tenon will need to be ground away and rebuilt When tenons are weld rebuilt by hand methods, it is typical for the blades to remain assembled to the rotor. If the blades have been removed, and tenons require rebuilding, another method is to weld 378 The Repair of Rotating Components Chromalloy Fig. 9.3.14—A weld rebuilt tenon where the weld layers have not fused completely This must be rebuilt. Fig. 9.3.15—The robotic weld rebuild of a tenon. 379 Chromalloy Turbine Steam Path Troubleshooting and Repair—Volume Two Chromalloy Fig. 9.3.16—The raw deposited weld from robotic rebuild. Fig. 9.3.17—The tenons of Figure 9.3.16 after finish machining. 380 The Repair of Rotating Components restore the tenon robotically. This process is shown in Figure 9.3.15, where the welding head is depositing a compatible material over a vane chord length sufficient to form the deposit shown as Figure 9.3.16. This weld deposit is then formed into the required tenons by machining (Fig. 9.3.17). A concern with the refurbishment of some critical stages is the inability to establish, by NDT methods, the condition of the weld, particularly in the HAZ after the tenons have been riveted. A method that can be used in this case, for vane sections that have a small turning angle, and where the vane section is relatively flat, is to weld in a complete tip section and then to reform the airfoil. This method has been used successfully in applications where the blades have a long radial height, and the end user wishes confirmation that material integrity has been maintained after tenon formation. Typical geometry of the tip rebuild is shown in Figure 9.3.18. Weld attached coupon Cut off Outer section Weld Gap "g" Weld preparation (a) Original form of blade tip Heat affected zone (b) Final form of blade after dressing (c) Fig. 9.3.18—The repair of a blade vane by removal of the outer section and the reattachment of a new tip which is then formed into suitable tenons. (a) Is the original design. (b) Is the weld attached coupon. (c) Is the final material form with the HAZ removed from the tenon material. 381 Turbine Steam Path Troubleshooting and Repair—Volume Two Screw attachment of the coverbands In the event only a small number of tenons are involved in failure, it is possible to employ a screw arrangement as shown in Figure 9.3.19. With this method a hole is produced in the blade vane, Figure 9.3.19(a) and (b), through the center of the failed tenon. This hole passes down through the blade vane, providing a means of screwed attachment. After coverband attachment by means of the screw, it is necessary to “stake” the screw to prevent its rotation during operation. Figure 9.3.20 shows an attachment of the screw type. To ensure a suitable attachment, it is advisable to use a screw produced from titanium, and a better attachment is achieved if there is local thickening at the tip section of the blade. Each situation must be evaluated separately, in terms of the local stage geometry. Figure 9.3.19(c) shows another method of attachment, in which a washer is placed over the failed tenon, and welded to the remaining material. Note: Various forms of screw can be used; the most suitable in any situation should be selected in terms of the geometry at the blade vane tip. This form of attachment should only be used when steam temperatures permit the use of titanium. Titanium screw Coverband Blade vane (a) Coverband Blade vane (b) Coverband Weld Washer Blade vane Tenon (c) Fig. 9.3.19—Methods of securing the coverband to the blade vane outer surface when tenon integrity has been reduced. 382 The Repair of Rotating Components Fig. 9.3.20—The titanium screw attached cover, of the type shown in Figure 9.3.19(a). Original vane and tenon outline Modified vane tip diameter Original vane tip diameter Radial position of tenon crack dH Fig. 9.3.21—A blade vane shortened by an amount “dH”. This will provide sufficient material that a new tenon can be formed. 383 Turbine Steam Path Troubleshooting and Repair—Volume Two Blade vane shortening Depending upon the form of the tenon, the blade can be shortened by the small amount “dH” of Figure 9.3.21; this method is often termed “tipping,” with enough material removed to allow the tenon length to be increased without welding. This has the disadvantage that the height of the blade vane is reduced, which modifies the outer “lap,” see Figure 2.12.2(a) and (b) of chapter 2. Also, for those stages in which the coverband provides a radial seal (or seal platform), the radial clearance and leakage will be increased, unless a non-standard sealing device is used, or the effective depth of the platform is increased to the original diameter. Also, if the coverband forms an axial seal, the shortening of the blade could move the seal point into a position where it will be unable to form an acceptable barrier to steam leakage. This method does, however, ensure the tenons can be reformed and the coverband reattached. There are mechanical disadvantages to this procedure: if the coverband segments covers a sufficient number of blades, they may be too long, at the new reduced diameter, and will therefore require new coverbands segments. However, for those stages where this type of repair can be considered, this may not be significant. It will also mean the discharge area from the blade, and therefore the pressure will be modified. It does, however, often allow a unit to be returned to service quickly. Another consideration, which it is difficult for the maintenance engineer to evaluate in a short period of time, is the possible effect on blade natural frequency, and coincidence with the “nozzle passing frequency.” This will only be of concern if the frequency margin of the original design was insufficient. With sufficient care this blade vane machining can be undertaken without the need to remove the blades from the wheel to shape the tenons. This has an obvious cost advantage. Note: With the successful development of weld repair techniques, and the ability to undertake repairs to even the highest temperature stages 384 The Repair of Rotating Components on the rotor without the need to remove the blades, “tipping” should no longer be necessary. Rubs of the outer surface of the coverband Instances occur where there are radial rubs between the coverband and stationary components of the unit, or even between debris trapped between the coverband and stationary components. These rubs are in most instances light, between the coverband and seal strips placed above the coverband to limit tip leakage. Such a light rub, which is relatively common, is shown in Figure 9.3.22. This rub, while almost certainly increasing the leakage steam flow, will have little detrimental effect on the structural integrity of the seal strip or coverband. Fig. 9.3.22—A coverband with light rubs on the inlet edge. Fig. 9.3.23—A coverband with heavy rubs on the inlet edge. 385 Turbine Steam Path Troubleshooting and Repair—Volume Two Other forms of rub have different levels of severity. Figure 9.3.23 shows a coverband that has sustained a damaging rub, causing relatively severe damage to the coverband inlet side, including damage to the tenons. The principle concern with a rub of this nature is the hardening, which could have been caused to the coverband and more particularly the tenons, making them more brittle, and therefore subject to failure. In such a case, hardness checks should be made to verify the adequacy of the material. Local hardening in excess of 15-20 Brinell points is unacceptable. There are, in fact, some major causes of coverband outer surface rubs. The most significant of these are: • debris located above the coverband • excessive overspeed, causing the rotor to grow radially • high levels of rotor vibration, particularly at start-up and shutdown, when the unit passes through critical speeds • distortion of the stationary components of the unit to which a seal strip is attached When a rub has occurred between the coverband/tenon seal surface and the sealing strips, it should be noted what material the seals strips are produced from. In the case of high temperature stages, the strips are normally made from a hard alloy steel with enough mechanical strength to resist the bending stresses induced in it by the pressure drop across it, at a high stage temperature. In this instance the coverbands will have suffered some damage, the extent depending upon the magnitude or severity of the rub. In the case of low temperature stages the seals are often made from a softer copper base alloy, and the seals will in general have been damaged far more significantly than the coverbands. Figure 9.3.24 shows the damage sustained by coverbands and tenons when a seal strip, produced from alloy steel, detached 386 Next Page The Repair of Rotating Components between the coverband and casing, causing a heavy rub on the blade row outer surface. Here there has been significant hardening of the coverband and tenon material requiring a complete rebuild. Similar damage situation is shown in chapter 4, Figure 4.11.7. If the rotor has a large radial clearance above the coverband, there is, on excessive overspeed, a tendency for the coverband to “curl.” This curling effect can occur on both the inlet and discharge side of the coverband, or at the segment end overhangs. Curling can occur to the extent the rivet head is deformed outward, then on resumption of normal operation, the integrity of the tenon will have been reduced, Fig. 9.3.24—Coverband damage sustained when metallic debris has been trapped between the coverband outer surface and the casing. 387 Chapter 10 Seals, Glands, and Sealing Systems INTRODUCTION The steam turbine requires that seals be provided at a number of locations to minimize the leakage of steam between and past stationary and rotating components. These seals can be arranged, or classified, in one of three main groups: • On those stationary components that enclose, and possibly carry steam path components. These include the casings, both inner and outer, and the diaphragms or stationary blades. Such components are used for both high and low (including sub-atmospheric) pressure application • On the rotating components, which compromise basically the rotor, which carries and locates the rotating blades, together with the stage hardware associated with these rows. It can also include some “shrunk-on” components, such as wheels or discs, coupling flanges, and thrust collars 549 Turbine Steam Path Troubleshooting and Repair—Volume Two • The other, basically stationary equipment, which is included to support, connect, and control the unit operation The steam path is defined as those components included in the first two groups, which expand the steam and extract work from it. The components of these two groups operate in close proximity, but must maintain a running clearance sufficient to prevent hard contact. In operation, steam leakage occurs between the stationary and rotating components of the unit. This leakage steam represents a bypass of the blade system, and therefore is a waste of the energy that is available from the working fluid. To limit such leakage, sealing arrangements are made within the unit to prevent or minimize this loss. There are three major locations within the steam path where seals are employed: • Where the rotor passes through the casing, to be supported on the journal bearings. This includes designs utilizing one rotor to carry two expansions of the total steam path. At shaft ends, these seals can be used to minimize the outward leakage of steam and the inward leakage of air This location can also be considered to include the leakage, which would occur along the shaft from one expansion portion to another at a lower pressure 550 • At the stationary blade inner diameters to the rotor. These are seals that, in stages employing a diaphragm construction, are normally carried at the inner diameter of the web inner surface. In stages employing blades located directly in the casing, these seals are arranged on the inner surface of the blade coverband • Those seals located above the rotating blade rows. These seals are used to prevent steam that is discharging from the stationary blade row by passing the rotating blades, and therefore doing no work in the rotating blade row Seals, Glands, and Sealing Systems At those positions within the unit, where the stationary and rotating components are adjacent, and require sealing, the seals can be located in either the stationary or rotating components. The configuration chosen for any location is dependent upon a number of factors, including the experience and preference of the design engineer. Another consideration when selecting the seal form and location is that those seals mounted on the rotor are subjected to the centrifugal forces of rotation. Therefore, if those seal strips normally used to produce the flow constriction are attached to the rotating components by mechanical means, this attachment must be able to resist those forces. STEAM PATH SEALS During operation, many parts of the steam turbine contain highpressure, high-temperature steam, and other portions are subject to vacuum. Also, because it is not possible to locate bearings within either of these steam environments, and it is necessary to ensure shaft continuity, provisions must be made for the shaft ends of the unit to project through the casings. The casings contain these highenergy gases, so they can locate on bearings at atmospheric conditions, and also be coupled to other rotors of the system to form a continuously coupled rotor system. At those points where the rotor passes through the casings, and where significant pressure/temperature differentials exist from internal to atmospheric conditions, it is necessary to provide a sealing system between the two portions to prevent excessive “leak in” of air at the sub-atmospheric or vacuum end, and “leak out” at the above atmospheric locations. Such sealing is usually achieved by the use of strips, which provide a labyrinth seal of small radial and/or axial clearance. These seal points are also constructed to cause a minimum of damage if rubs should occur. 551 Turbine Steam Path Troubleshooting and Repair—Volume Two In the low-pressure sections of the unit, the rotor passes through what is normally a fabricated casing in a region where a vacuum exists. In this case, there is a positive pressure (atmospheric to vacuum) trying to induce an ingress of air to the casing. This air, which is not condensable at system temperatures and pressures present, would tend to degrade the vacuum. In so doing, the air would degrade the energy range of the steam, and therefore reduce the efficiency of the cycle. Also, the oxygen contained within the air could react within the unit to introduce or accelerate corrosive action. At unit “start-up,” it is normal to first pull vacuum in the condenser, whose action produces sub-atmospheric pressures throughout the steam path. This vacuum would be sufficient to induce an inward flow of air into the system, making it difficult to bring the unit to speed, and to begin generating power. For these reasons it is necessary to provide the unit with a steam sealing system, which at all internal to external seal locations is able to provide steam to the seals, and regulate or limit the inward flows of air. To accomplish this sealing, a system must be provided that is able to minimize both air leakage inwards at “start-up” and at all loads from the low-pressure sections; and outwards from the high-pressure locations at “start-up” and during normal operation. This system must be effective under all conditions of load and steam conditions. Therefore, it is necessary for such a system to employ some regulatory devices that controls and limits these flows to an acceptable level. At no load and light loads, a vacuum exists throughout the turbine unit, extending as far up the steam path as the high-pressure section, i.e., at “start-up” the entire unit will be under vacuum. Under these conditions, the sealing system must be capable of providing a sealing capability sufficient to provide positive atmospheric packing, at all sealing locations, and sufficient to prevent the ingress of air, while allowing the vacuum to be maintained in the condenser. 552 Seals, Glands, and Sealing Systems Before the development of the current designs of sophisticated sealing systems, it was normal to provide a positive seal at the lowpressure ends, and an atmospheric leak-off, which vented steam from the gland housings to the atmosphere. With the present design of systems, the “atmospheric leak-off” is maintained at a pressure marginally below atmospheric, and is vented to some suitable location within the system. The lowest pressure leak-off in the system is often taken to separate condensers, defined as “gland steam condensers,” which are maintained at pressures marginally above (e.g., 0.5" Hga), the main condenser pressure. In these condensers the steam is condensed, and any air present is expelled from the system to prevent its promoting corrosion within the system, mainly in the boiler. The constricting seal is normally obtained by the using fine, metallic strips, which are arranged to provide a small clearance (axially or radially), to minimize the leakage area available, and therefore minimize the quantity of steam that is able to flow past them. Such a steam flow, because no work is done, is a “throttling,” or constant enthalpy expansion. These seal strips are normally arranged to provide a series of throttling constrictions. In doing this, there is a continual reduction of steam pressure in successive steps from one section, or from one chamber, to another. Depending upon the differential expansion present at the location of the seal, the strips will be arranged as a labyrinth, in which the steam is continually forced to change direction because the seals are being formed at alternating “high” and “low” diameters. When there is differential expansion between stationary and rotating parts to the extent that too large a distance would be required between the seal strips, a “straightthrough” type of construction is used, i.e., all constrictions are made at the same sealing diameter. Many older units still employ water-sealing glands, but their number is reducing, as these units are retired. However, if such a unit is to be kept in service, there are considerable advantages to replacing the water seal with the now more conventional, and effective steam throttling system. 553 Turbine Steam Path Troubleshooting and Repair—Volume Two The steam sealing system (general requirements) Figure 10.2.1 shows a typical steam sealing system, capable of fulfilling the thermodynamic requirements for efficient operation. To perform these functions adequately, the system requires the incorporation of certain special purpose devices and the availability of an adequate steam supply. It is common practice in many power cycles to use the boiler steam, this being the highest pressure available in the cycle and therefore capable of sealing all the shaft glands, under all those conditions of load and pressure it experienced during operation. Fig. 10.2.1—A typical steam sealing system, capable of regulating sealing steam flow to both the high and low pressure glands of a unit. 554 Seals, Glands, and Sealing Systems Under normal operating conditions, steam is supplied to the sealing system from within the cycle, as required. The pressure of this sealing steam is adjusted by some suitable regulator to a pressure of from 30-45 psi for the 2,400-3,500 psi steam from the boiler, the actual pressure depending upon the cycle design. Figure 10.2.2 shows schematically the essentials of this system regulating high and low-pressure glands on the ends of a shaft. The steam flow “into” and “out of” the glands is shown at the high-pressure end on the left of this diagram. Under normal operating conditions the steam pressure “Pa” is at a relatively high pressure and temperature and will flow out past gland sections “1,” “2,” and “3.” It is possible for steam to be extracted from intermediate pockets at pressure “Pc” or “Pd.” In Figure 10.2.2, steam is shown being extracted to an intermediate point at pressure “Pc.” At position “D” steam at pressure “Pd” is extracted and supplied to the regulator of the system. Its pressure is adjusted to the regulator pressure of about 30-45 psi. This steam is eventually used to seal the low-pressure systems. At position “E” the pressure is maintained at a slightly sub-atmospheric value by use of an Fig. 10.2.2—Schematic arrangement of the steam flow to the high and low pressure glands under normal operating conditions. 555 Turbine Steam Path Troubleshooting and Repair—Volume Two auxiliary gland steam condenser or alternatively a steam jet air ejector. This leakoff takes overflow from the position “D” extraction point and also air that is drawn in from position “F” past constriction “5.” At the low-pressure end steam is supplied to position “H.” This steam is at a pressure controlled by the regulator “R,” and leaks past the seal constriction “6” into the low-pressure section, whose pressure “Pg” is at vacuum corresponding to the turbine exhaust pressure. A portion of the steam leaks past constriction “7,” where it mixes with air at pressure “Pj,” and is taken to the auxiliary steam condenser or steam jet air ejector. Figure 10.2.3 shows the flow conditions at light loads. In this condition, it is assumed the high pressure point “Pa” is sub-atmospheric. In this case, steam is taken from the external source to the regulator, adjusted to a regulator pressure of 30-45 psi, and supplied to both points “D” and “H,” where the steam is at the high-pressure end of the unit. There may or may not be extraction of steam at “Pc” to a point in the cycle. At the low-pressure end the steam will enter at “H” and flow outward to “G” and “J.” At “J” this steam is mixed with air and taken to the steam jet ejector or auxiliary gland steam condenser. Fig. 10.2.3—Schematic of the gland system of Figure 10.2.1 when the unit is operating under light load conditions, and the pressure at “a” is sub-atmospheric. 556 Seals, Glands, and Sealing Systems The steam sealing system (nuclear units) Some nuclear cycles employ “hot steam,” i.e., the working fluid in the turbine has been into the main reactor vessel, and therefore has had the opportunity to have picked up some degree of radioactive matter and become contaminated. In this type of system it is considered prudent to provide for a separate source of clean sealing steam that is free from the possibility of contamination, so that should outward leakage occur, it will not be a health concern. Such a system will use a special boiler or evaporator. This steam raising equipment will use a feedwater quality water source, which produces steam of a quantity sufficient to seal the glands. The heat used to generate this clean steam can be supplied from an external source, or it can employ main steam extracted from some convenient point in the main power cycle, and use this in a heat exchanger vessel internal to the evaporator. FUNCTIONS OF THE STEAM SEALING SYSTEM The steam sealing devices are part of an integrated subsystem within the steam cycle. This subsystem is intended to provide sealing at shaft end points to prevent both the “egress” of high-pressure, high-energy steam, and also the “ingress” of air, at sub-atmospheric pressure locations. Ingressing air, being non-condensable, will reduce vacuum, and therefore degrade the efficiency of the unit. Steam, if required to provide a sealing source for use in these devices, is extracted from some convenient point in the cycle, used, then returned to the main steam cycle when it has performed its duty, and been purged of air. The functions of this subsystem can be considered to be: 557 Turbine Steam Path Troubleshooting and Repair—Volume Two • to maintain an essentially constant pressure steam supply to the sealing system header to supply the shaft end seals. This steam supply must be effective with the sealing strips in either a new or worn condition • the removal of the non-condensable gases, which gain access to the system through the shaft glands • to perform this sealing function over a wide variation of steam conditions and of load requirements placed on the turbine generator set • the gland housings must by capable of adjustment to achieve an optimum alignment to the rotor. These gland housings and seals must be capable of withstanding high internal pressure and temperature, and suited to withstand or resist both erosive, and corrosive attack from within the unit • the system must be capable of accepting steam blowdown, and during operation prevent or minimize the ingress of water to the turbine from its internals • the piping of the system must be capable of supplying adequate quantities of steam to the gland housings, without exceeding acceptable velocities or incurring excessive pressure drops. This piping must be sufficiently flexible to permit thermal expansion, and ensure end point movements do not result in excessive stresses or reactions on the gland housing. Such piping must also be connected so that it does not provide regions where water can collect in large quantities during “shutdown” periods At locations internal to the various portions of the steam path, the seal strips have a single function of minimizing the quantity of steam, bypassing the stationary and rotating blade rows. Such leakage flow represents steam bypassing the blades and therefore doing no work on them. This steam will also, after bypassing the blade row, 558 Seals, Glands, and Sealing Systems re-enter the main steam flow both at high incidence angles and velocities, which will tend to destroy locally the orderly flow of the steam from blade row to blade row. There are various arrangements or configurations for these sealing strips, and depending upon their individual geometry, they can provide some of the following requirements: • The ability to be changed when worn • The capability to be spring loaded, so they are able to “back away” from contacts, or rubs, and return to their sealing position when the contact or rub condition is removed or corrected. During such rubs these seal strips could have suffered some “wear” • The ability to have a sharp or “knife” edge produced on them at the sealing position • The ability to withstand the pressure differential that exists across them, and to withstand the stresses these pressures induce STEAM LEAKAGE THROUGH LABYRINTH SEALS In order to determine the leakage quantities that flow past a labyrinth seal, it is necessary to employ the basic theory of flow. The relationship relating leakage quantity to the physical properties or characteristics of the steam, and the geometric arrangements of the gland, and its dimensions is derived by the relationship of Martin. 559 Turbine Steam Path Troubleshooting and Repair—Volume Two Consider the single seal constriction, shown as Figure 10.4.1. From the equation for steam quantity “m” flowing through any area “Ae”), the flow can be found from: where: R= Inlet pressure/discharge pressure = P1/P2 Ae = Discharge area = 2π R ε This equation can be represented in the more recognizable form of “Martin’s equation,” for labyrinth seals: where: X= P1 = P2 = Vs1 = N= 560 The The The The The pressure ratio across the seals P1/P2 inlet pressure pressure at discharge specific volume corresponding to pressure P1 number of series constrictions Seals, Glands, and Sealing Systems P1 Ae = 2 π R ε dp P2 Steam flow R Fig. 10.4.1—A single sealing strip showing the principal dimensions controlling leakage flow. To determine the value of constant “k,” consider the dimensions of the equation for flow, in foot/pound/second imperial units. This equation assumes the flow coefficient “ψ” is equal to 1.0. Also, “lbf” is pounds force, and “lbm” is pounds mass. For a series of “N” strips, the Martin equation assumes the pressure ratio “x” is constant across each of the series constrictions of the total seal arrangement, and the conditions are in either the superheat or saturated regions for the entire throttling expansion. The leakage flow will be a maximum when the value of the pressure ratio “x” reaches the critical value. Because the flow cannot exceed that associated with the critical pressure ratio, if the value of “x” exceeds the critical value, the critical value from Figure 10.4.2 should be used. In many portions of the steam turbine, labyrinth seals are used in series groups, the pressure falling successively through each expansion (or throttling). At each throttling constriction a portion of the total heat drop across the series gland arrangement is converted to kinetic energy, which is subsequently destroyed in the steam chamber formed between the strips. This kinetic energy is partially 561 Turbine Steam Path Troubleshooting and Repair—Volume Two reconverted to pressure energy as its velocity reduces in the chamber. Most of the remainder of the energy that is converted at each constriction will be converted to heat. Fig. 10.4.2—The critical pressure ratio across a series arrangement of labyrinth seals. Consider this throttling effect of the four series throttling strips in a single gland, shown as Figure 10.4.3. The steam has an initial pressure “Pi” at entry to the gland. This condition is represented on the Mollier diagram (Figure 10.4.4) by point “Ai.” After expanding past this first constriction, the steam will have been reduced to condition “Ao,” pressure “Pa.” In the chamber formed between the first and second seal strips, the kinetic energy of the steam is destroyed, and reconverted at constant pressure “Pa,” to condition “Bi.” From point “Bi,” there is then a further expansion of the steam past the second constriction, with the pressure falling to “Pb,” condition “Bo.” The kinetic energy is again reconverted in the chamber between the second and third seal strips, raising the thermal energy level from condition “Bo” to “Ci,” at constant pressure “Pc.” This process of expansion and kinetic energy reconversion is continued throughout the series of seal strips, until the final expansion takes the steam to condition “Do” at pressure “Pd.” The locus of the points “Ao.......Do” is 562 Seals, Glands, and Sealing Systems called the “Fanno curve.” Note that at exit from the final strip, the steam condition is represented by “Do,” the steam having kinetic energy, at a reduced enthalpy “Hdo.” ε Pi Pb Pa Pc Pd Steam flow Fig. 10.4.3—A series arrangement of four seals with inlet and outlet steam pressures of ‘Pi’ and “pd’. Pi H Pa Ai Bi Ao Hdo Pc Pb Ci Di Pd Ei Bo Co Do s Fig. 10.4.4—The expansion of steam through the four seals of Figure 10.4.3 shown on the Mollier diagram. 563 Turbine Steam Path Troubleshooting and Repair—Volume Two Any group of labyrinth seals has pressures before and after them determined by either the cycle parameters, or the internal arrangement of the steam path parts. There are two types of flow, or pressure distribution that should be considered: • Those groups in which the number of constrictions is sufficiently large, the pressure ratio across each, including the last, is less than critical • Those groups in which there are insufficient constrictions for the pressure ratio, that flow through the last has a pressure ratio “x,” which exceeds the critical value When this critical pressure ratio has been exceeded, the flow through a group of constrictions can be determined using the following equation: where: “λ” is a function of the labyrinth pressure ratio “x” where: N= X= Pi = Pd = Number of effective series constrictions Pressure ratio across seals = Pi/Pd Inlet pressure to the seals Outlet pressure from the seals When the critical value of “x” has been exceeded, this expression can be used, using the critical pressure ratio in place of “x” from Figure 10.4.2. 564 Seals, Glands, and Sealing Systems Calculation of leakage flow past blade constrictions (small number of series seals) As an alternative method of determining the leakage at a stage sealing position where the pressure drop is relatively small, and only one or two constrictions are effective, the following method can be used, with reasonable accuracy, to determine the leakage quantity. The steam velocity “C” through an opening with a pressure drop “dp” across it can be determined by: C2 = 2g 144 Vmean . dP where: V mean is the mean specific volume over the seals dP is in psi For openings 1 and 2 in series: Consider the rotating blade tip sealing arrangement shown in Figure 10.4.5(a). Because of differential expansion, only two of the three seal strips are effective at any one time. Here these two strips form an arrangement of seals in series. It can be assumed the pressure at inlet to the first seal is “Pi,” and at discharge from the second seal strip is reduced to “Pe.” It should be noted, these pressures are set not by the seal, but by the rotating blade pressure drop at the tip section, as determined by the thermodynamic design calculations for the stage. The pressure drop across the blade tip section is “dP.” The velocity at discharge from the series seals is “C.” The specific volumes corresponding to pressures “Pi” and “Pe” are “Vsi” and “Vse.” For such a small total enthalpy drop it can be assumed “Vsi” = “Vse,” or the enthalpy drop across the first seal constriction is the same as that across the second. This makes the velocity of discharge from both strips the same and constant at “C,” on the reasonable assumption that the discharge area through the two clearances is the same. 565 Turbine Steam Path Troubleshooting and Repair—Volume Two Pi ε Pm ∆ Had Pe (a) (b) R Fig. 10.4.5—In (a) is shown a radial seal arrangement, where two seals are effective at any axial position of the moving blade. In (b) is shown the conditions on the Mollier diagram. The steam conditions around the tip are those shown on Figure 10.4.5(b). In this case the total enthalpy drop across the seal strips is “∆had.” For a design such as that shown in Figure 10.4.5(a), it is general for “Ae1” to equal “Ae2,” which is equal to “A.” If “N” is the number of constrictions with the same leakage area “A” and coefficient of discharge “ψ.” However AE = 2. π .R .ε Also the velocity “C” due to an isentropic enthalpy drop through one throttling constriction is: where: “N” is the number of effective constrictions 566 Seals, Glands, and Sealing Systems In Figure 10.4.5(a), this is two at any one operating condition, and “∆Had” is the enthalpy drop across the seal. Inserting equations 10.4.9 and 10.4.10 into equation 10.4.8 for the mass flow through the two series constriction gives: Experiment has shown a mean value of “ψ” can be taken as 0.82, and a mean value of “Vs” between the inlet and outlet conditions yields an accurate result. QUANTIFYING LABYRINTH LEAKAGE (APPLYING THE METHOD OF MARTIN) The following analysis, undertaken for various portions of the steam path, is made using the method of Martin. It is recognized that this equation derived by Martin is theoretical, and must be modified by specified values of the “flow coefficient.” However, this expression has been found to yield good results in practice and therefore provides an acceptable method for operators to estimate leakage quantities for various parts of a unit. In this analysis, leakages are calculated for both an impulse and reaction unit, to demonstrate the level of leakage that occurs in both designs. The actual values for any unit can be determined from information derived from the heat balance, and 567 Turbine Steam Path Troubleshooting and Repair—Volume Two clearance measurements. These values are calculated using the heat balance, and knowledge of the unit cold setting clearances. Shaft end glands Consider the high-pressure end of the turbine steam path, shown in Figure 10.5.1, (which can represent shaft glands of either an impulse or reaction design). In this case, the steam that leaks from the steam path is controlled by the first section of 35 constrictions, and the pressure ratio across this portion of the gland housing. The sections of 20 and 15 constrictions, simply determine the quantity that flows to other, lower-pressure points of the cycle, with the remaining portion flowing to some cycle position at a lower pressure. Therefore, for calculating the leakage flow quantities and the associated losses, attention must be focused on the 35-constriction gland, in terms of controlling the steam leakage quantity, which degrades efficiency. In fact, the points at which the leakage steam is returned to the unit does ultimately effect the output and efficiency, but their influence can be considered as secondary in terms of current considerations, and unit performance. The effective seal diameter at the 35-constriction gland is 20.7125", with a radial clearance of 0.020". The upstream pressure P1=1045 psia, with a specific volume Vs1 of 0.687 cu ft/#. The conClearance Cl1 P1,Vs1 P2,Vs2 Qa-Qb P3,Vs3 Qb-Qc N=35 P4 Qc+Qd N=20 N=15 Q1 N=4 Q2 Q3 Shaft End Dr De1 = Effective seal diameter De2/Cl2 De3/Cl3 Fig. 10.5.1—The shaft end gland arrangement of a high pressure section. 568 Qa Pa,Vsa Seals, Glands, and Sealing Systems ditions at the first leak off are P2=237 psia, at a specific volume Vs2=3.00 cu ft/#. The shaft gland is then arranged to provide a 23constriction gland at a seal diameter of 18.5", again with a radial clearance of 0.020". After this second seal system, there is a second leak off to a pressure P3=21.00 psi with a specific volume of 34.5 cu ft/#. There is then a third group of seals, comprising a 17-constriction gland at a diameter of 15.5", with a radial clearance of 0.020". This leak off is connected to a gland steam condenser at a pressure of 14.5 psia. A fourth seal section is provided to allow the ingress of air, which goes with the leakage steam to the gland steam condenser. Determining the leakage quantity “Q1” past the initial 35 seal strips is: For leakage “Q2” past the second seal system containing 23 strips, which is overcritical, therefore use “x” = 6.12 from Figure 10.4.2. This indicates the latter constrictions will have an “over critical” pressure drop. 569 Turbine Steam Path Troubleshooting and Repair—Volume Two Leakage past the third seal system “Q3” is found from: The first leak off quantity = Q1 - Q2 = 13,936 - 3,53 = 10,483 #/hr The second leak off quantity = Q2 - Q3 = 3,453 - 299 = 3,154 #/hr The third leak off quantity = Q3 + Qa = 0.0683 + Qa In terms of the degradation of unit output, only the first leakage, past the 35 seal strips need to be considered. If the enthalpy at position “1” is 1,413.5 Btu/#, and at no portion is the steam returned to the steam path to generate power, then the output loss is equal to its initial enthalpy, minus that at the condenser, say 956.4 Btu/#. This represents at per mil loss of 1,866.9/20 = 93.35 kW/0.001" of radial clearance. Note: While the loss rate is indicated at 93.35 kW/0.001" clearance, this is misleading since the radial clearance will decrease as the unit goes into operation, and there is adjustment due to radial growth. However, this number is of considerable value when comparing measured clearances at an outage with the design values. This loss rate is anticipated by design and allowed for in determining unit output. It is the opening of clearances above these values that represent a real loss to the unit. 570 Seals, Glands, and Sealing Systems 60 Shaft End Leakage Loss - 'Q1' 9,000 55 8,000 50 7,000 45 6,000 40 5,000 35 4,000 30 3,000 25 2,000 20 1,000 15 0 0 20 30 40 50 60 70 Radial Clearance in 0.001" 80 Leakage frow in 1000#/hr Loss of Output in kilowatts 10,000 10 90 Fig. 10.5.2—The leakage and power losses for the shaft end gland shown as figure 11.5.1 The leakage flow and output loss, as a function of radial clearance, is shown in Figure 10.5.2. These calculations assume the flow coefficient is unchanged as any “rubbing” occurs, and the seals wear uniformly along their length. These shaft-end packings represent a large loss in any unit, and it should be standard practice during maintenance outages to check this region, to determine the need to install new strips or gland rings. The dummy pistons of reaction turbines In reaction-designed units, because of the axial thrust developed on the rotating blade rows by the pressure drop that exists across them, it is often necessary (when reversed flow sections cannot be used) to employ a dummy, or pressure-balance piston to help counter a portion of this axial thrust, and therefore lower the load that must be carried by the thrust block. 571 Turbine Steam Path Troubleshooting and Repair—Volume Two T1 = 2 2 π Dd Dr xP1 4 Dummy Piston Seals (see details) Pb T2 = P1 P2 π 2 2 Dd - Ds xP2 4 Shaft End Packing Blade Thrust Tb Rotating portion of unit (a) Total Thrust = Tb + T2 = T1 Dr Dd Ds Dt Dq Fig. 10.5.3—The effect of the ‘dummy piston’ on total piston thrust. A dummy piston is shown diagrammatically in Figure 10.5.3, which has on it an axial thrust in both directions. The magnitude of these thrusts is determined by the shaft diameters and pressure intensities on the two sides of the piston. The pressures and shaft diameters are selected to minimize, or reduce to an acceptable level the resultant axial thrust. Because there is a net pressure drop across the dummy piston, and there is a rotating/stationary interface at the outer diameter, this system requires a labyrinth seal be produced at a relatively large diameter to limit the amount of leakage steam. The steam leaking over the dummy then passes to the shaft end packings, or another suitable point within the steam path. P k Clo Cli Clo Dummy Piston Cli Dummy Piston (a) Fig. 10.5.4—Various arrangements of dummy pistons. 572 P P (b) Dummy Piston (c) Seals, Glands, and Sealing Systems Figure 10.5.4 shows the basic arrangements of dummy pistons having “N” seal strips. As an example, dimensional and steam characteristics around such a dummy as shown in Table 10.5.1: P1 = 2050psia VS1 = 0.352cu.ft/# x = 2050/612 = 3.350 Dimensional Characteristics: Dd = 25.00" Cl = 0.025" Steam Characteristics: PS = 612psia N = 60 Table 10.5.1—Dummy Piston Steam and Physical Characteristics. From Martin’s equation (equation 10.5.3) the leakage quantity “m” can be found as: If the steam has an initial enthalpy of 1432.4 Btu/#, and a final enthalpy at 612 psia of 1,310.2 Btu/#, the lost kilowatts at 0.025" radial clearance is: Therefore, the loss rate is 1,114.7/25 = 44.6 kW/0.001" of radial clearance, in excess of design-specified clearances. Figure 10.5.5(a) shows the losses at the dummy piston as a function of the radial clearance between stationary and rotating components. The dummy piston is normally located close to the thrust block. This is done so that there are relatively small amounts of differential expansion at this axial position, and a greater number of seal strips can be incorporated into the design. Shown in detail (a) of Figure 10.5.4 is the design where alternate seal strips are fitted into the stationary and rotating components, and in details (b) and (c) where strips are assembled to the rotating and stationary components respectively. 573 Turbine Steam Path Troubleshooting and Repair—Volume Two 2000 60 Dummy piston leakage loss 1600 50 1400 1200 40 1000 800 30 600 Leakage frow in 1000#/hr Loss of Output in kilowatts 1800 400 20 200 0 0 20 30 40 60 70 50 Radial Clearance in 0.001" 80 90 Fig. 10.5.5(a)—Shown are the basic dummy piston losses in terms of leakage flow and kilowatts. 2000 1900 Loss of output in kilowatts 1800 1700 1600 1500 1400 1300 1200 1100 1000 10 20 30 40 50 Number of Series Constrictions "N" 60 Fig. 10.5.5(b)—The effect of reducing the number of effective seals in the dummy piston. 574 Seals, Glands, and Sealing Systems Figure 10.5.5(b) shows, for the same dummy piston, the effect of reducing the number of constrictions. If it is assumed the unit has a design specification of 60 effective strips, with a radial clearance of 0.025", this will give a design leakage loss equivalent to 1,114.7 kW. Unfortunately, these strips are normally located in a “hot” environment, where they can become brittle after a period of operation. While designs exist for spring loading these seals, they do tend to suffer damage due to rubs, and the effects of exposure to high temperatures. It is not uncommon to find strips, or portions of strips, missing when such a dummy is opened for inspection. Figure 10.5.6 shows that in some older design of smaller rated reaction units, the “dummy piston” was arranged to be stepped. This allowed the thrust to be adjusted with greater ease. Outer Dummy Inner Dummy Fig. 10.5.6—The reaction unit with inner and outer dummies. Steam path seals Within the steam path (the stationary and rotating blade system), it is necessary to minimize the leakage past both sets of blade rows. The number of seal strips that can be accommodated at any position is influenced by the seal form and location, the axial width, and arrangement of the stage. These are factors established during the early design phase. To illustrate the effectiveness of these various seals, portions of two high-pressure expansions will be considered. One is an impulse design section having five stages, the other an equivalent reaction section with 575 Turbine Steam Path Troubleshooting and Repair—Volume Two 0.80 0.70 Vs 2200 P 0.60 Steam Pressure psia 2000 1800 1600 1000 T 0.50 1400 1200 900 0.40 600 1300 850 800 1000 800 950 Temperature °F 2400 Specific Volume Cu Ft/# nine stages. These stages, because of the difference in the design velocity ratios, cover approximately the same enthalpy range, and therefore are comparable. In each of the two expansions, the stages have the same vane root diameter, and a velocity ratio consistent with design philosophies. The total stages of the two expansions follow current design practice and have a rotor span of equal axial length. The steam conditions throughout the expansion are shown in Figure 10.5.7, as a function of expansion line enthalpy. It has been assumed in the following calculations, that both expansions have the same state line efficiency (which may not be a fully justified assumption, but will be sufficient to allow a basic comparison of the two design philosophies). 0.30 1320 1340 1360 1380 1400 1420 1440 1460 1480 Enthalpy BTU/# 750 700 Fig. 10.5.7—The variation of steam path conditions in a high pressure section. These conditions relate the state line properties as a function of enthalpy. 576 Seals, Glands, and Sealing Systems P=2002 V=0.386 P=2240 V=0.365 P=1945 V=0.394 P=1725 V=0.426 P=1662 V=0.437 P=1465 V=0.480 P=1240 V=0.552 P=1403 V=0.497 P=1050 V=0.636 P=1180 V=0.574 P=1003 V=0.665 1 1 1 1 1 S1 S2 R1 13 R2 6 P=1971 V=0.391 S4 R3 S3 6 P=1690 V=0.432 S5 R4 4 5 P=1430 V=0.491 P=1200 V=0.566 P = Pressure in psia, V = Specific Volume in cu ft/#, S = Stationary Blade Row, R = Rotating Blade Row. N R5 P=1019 V=0.654 = Number of Effective Seals Fig. 10.5.8—A five stage impulse design steam path, and the steam conditions at various locations that affect the quantity of steam by-passing the steam path blade rows. P=2168 P=2014 P=1864 P=1715 P=1568 P=1425 P=1297 P=1075 P=1175 V=0.366 V=0.381 V=0.414 V=0.429 V=0.457 V=0.492 V=0.530 V=0.618 V=0.574 P=2240 P=2080 P=1925 P=1772 P=1623 P=1480 P=1343 P=1220 P=1109 P=1020 V=0.365 V=0.372 V=0.394 V=0.418 V=0.445 V=0.477 V=0.515 V=0.537 V=0.602 V=0.648 2 2 2 2 2 2 2 2 2 S1 R1 S2 R2 2 2 P=2151 P=1991 S3 R3 2 P=1832 S4 R4 2 P=1683 S5 R5 2 P=1539 P = Pressure in psia, V = Specific Volume in cu ft/#, S = Stationary Blade Row, R = Rotating Blade Row. S6 R6 S7 R7 2 2 P=1264 P=1398 N S8 R8 S9 2 2 P=1148 R9 P=1050 = Number of Effective Seals Fi 10 5 9 Fig. 10.5.9—A nine stage reaction design steam path, and the steam conditions at various locations that affect the quantity of steam bypassing blade rows of the steam path. 577 Turbine Steam Path Troubleshooting and Repair—Volume Two Figures 10.5.8 and 10.5.9 show the basic arrangement of the two steam paths. Figure 10.5.8 shows the impulse unit, and Figure 10.5.9 shows the reaction design. The steam conditions at the various stage points are shown on these figures. The vortex, or radial flow action of the steam will cause a radial pressure gradient at discharge from the stationary blades; but no such gradient exists at discharge from the rotating blades. This is a valid assumption for blades with a small ratio of blade height to mean diameter. The number of effective seals at any location is shown as “N.” These seals are considered to be radial. However, they could be replaced by an axial arrangement; no effort has been made to differentiate between the effectiveness of the two. The clearance with an axial seal is a function of the “hot running clearance,” which varies from stage to stage, and is dependent upon the rotor differential expansion. It has been assumed, with justification in this analysis that cold axial clearances are chosen by design so that both “radial” and “axial” seals are equally effective during normal operation. For an operating unit with rubs on the axial seals, their effectiveness will have deteriorated, and will be difficult (or impossible) to repair if the seal is formed as an integral part of the rotating blade. Therefore, it can be assumed axial effectiveness is variable, and possibly incapable of remedial action once damage has occurred. For this reason, no considerable error will be introduced by assuming radial and axial are equally effective. Impulse unit. The impulse unit, because of the larger axial space available per stage, can be arranged to employ a greater number of seals in series, in an effort to reduce leakage between the diaphragm and rotor, which is where the majority of the pressure drop occurs. In this present example, the number of seals selected is consistent with the type of unit arrangement. (The number of seals used on any diaphragm will depend upon their pitching, and whether castellations or “high-low” formats can be used). The number and arrangement is influenced by diverse factors such as the rotor differential expansion, diaphragm creep, and deflection (see chapter 2). 578 Seals, Glands, and Sealing Systems In this unit, the blade tip sealing has been assumed to employ only one effective seal. This can often be increased to two or more for larger rated units by the use of a combined axial and radial arrangement, the use of a radial seal on both the inlet and outlet of the coverband, or a two strip inserted segment. The axial seal is normally formed as part of the shroud or coverband. When seals are formed as part of the rotating component of the unit, they are difficult to repair once they have suffered material loss due to rubs. Under these conditions, to repair could require replacement of an entire blade row. When rotors are double flow construction, it is often difficult to produce an effective axial seal on the expansions, which move away from the thrust block in operation. If axial seals are produced on both flows of a double flow rotor, it is questionable as to their effectiveness in one flow, and more reliance is normally placed on the use of a radial seal configuration. Shown as Figure 10.5.10 are the calculated losses due to leakage between the diaphragm and rotor. The upper curve shows the total losses for all five stages, and the lower curves show the individual stage losses. These are plotted for radial clearance from 0.020"- 0.060". Using the mean slope of these curves, it can be calculated that the loss associated with a high-pressure section diaphragm is about 10.5 kW/0.001"/stage (kW/mil). The losses associated with blade tip leakage are shown in Figure 10.5.11. In this case, the loss due to increased leakage is at about the 4.0 kW/mil/stage level. If there were two effective seals at each stage, the level would reduce to about 2.8 kW/mil/stage. An estimate of the deterioration in cylinder output for clearances beyond design values can be determined based on the output loss rates in the diaphragms and blade tips. Using these two loss rates, it is possible, when examining the high-pressure section of any unit to be able to assess the total loss associated with rubbed clearances. At that time the most appropriate remedial action can be selected. 579 Turbine Steam Path Troubleshooting and Repair—Volume Two 2200 Total Kilowatt Loss in 5 Stages 2000 1800 1600 1400 Total Loss for 5 impulse stages 1200 1000 800 0.020 0.030 0.040 0.050 0.060 Radial Clearance - inches F2, N= 6 F3, N= 6 Stage Loss - Kilowatts 500 Individual Stage Losses 400 F1, N=13 300 200 100 Fig. 10.5.10—The diaphragm leakage in the impulse stage. 580 F4, N= 5 F5, N= 4 Seals, Glands, and Sealing Systems 1800 Output Loss - Kilowatts 1600 Tip Leakage Loss for 5 Stages (Assumes 1 Effective Constriction/stage) 1400 1200 1000 800 600 0.030 0.040 0.050 0.060 0.070 0.080 Radial Clearance - inches Fig. 10.5.11—Blade tip leakage for the five impulse stges, with one effective seal strip on each row. Similar calculations can be made for an intermediate pressure cylinder, and would yield values of about 3.0 kW/mil/stage for the blade tip leakage, with one seal strip, and 2.0 kW/mil/stage for the diaphragm leakage. For the low-pressure cylinder the corresponding values would be about 1.5 kW/mil/stage, for both the blade tip and diaphragm leakage. Reaction unit. The reaction unit contains a greater number of stages carried by the rotor in the same (or slightly longer) axial span. Therefore, there is not enough axial space to provide a large number of seal constrictions at each stationary blade row as used in the impulse unit. However, as the same enthalpy range and pressure drop occurs across a greater number of rows, the per stage pressure drop is smaller. In addition, in the 50% reaction unit, the pressure drop also occurs in two distinct and equal steps, across both the stationary and rotating blade elements. Therefore, the strip seal duty at each sealing point is somewhat reduced. 581 Turbine Steam Path Troubleshooting and Repair—Volume Two In order to determine the effectiveness of the sealing system, it is necessary to calculate leakage flows, and determine their value on a similar basis to those determined for the impulse unit. In this manner both the loss rate, and the potential for deterioration and refurbishment can be established. The curve in Figure 10.5.12 shows the loss of output for both the stationary and rotating blades for the nine stages. Based on these curves, it can be estimated that steam path losses are at about the 4.0 kW/mil/stage level for both sets of blades. By analysis, a similar set of values can be determined for an intermediate pressure cylinder at about the 2.5 kW/mil/stage level, and a similar value of about 1.5 kW/mil/stage for a low-pressure cylinder. Both values for the intermediate and low-pressure cylinders apply to both the stationary and rotating blades. The values calculated for the high-pressure section assumes two effective seals at each of the stationary and rotating blade coverbands. Some designs of low-pressure units, both impulse and reaction, are arranged to carry more than one row of rotating blades on a single wheel. In such a design, this can modify these values of per stage losses, because the interstage seals are effective at a larger diameter. This fact influences the effective leakage area. In such a case, it would be necessary to make a separate evaluation of the design. From the potential gains, operators might be tempted to close seal clearances below their design value, being prepared to recognize these clearances will open during operation, probably during the first “run-up” after return to service. While this procedure would produce a minimum of clearance, and be consistent with what the unit can tolerate, it must be recognized that clearances could open further during transient operation. It must also be recognized that rubs generate heat, and excessive amounts of heat can cause metal embrittlement, possibly causing the failure of the entire sealing strip. Such a decision to rub clearances must depend upon the amount of material that must be removed by the rub. One serious disadvantage 582 Seals, Glands, and Sealing Systems Rotating & Stationary Blade Losses in Kilowatts 2400 2200 Total Blade Leakage Losses for 9 Stage 2000 1800 Rotating Blade Losses 1600 1400 Stationary Blade Losses 1200 1000 800 0.020 0.030 0.040 0.050 0.060 Stationary and rotating blade radial clearances - inches Fig. 10.5.12—The leakage on the stationary and rotating blades of the reaction unit with two seals effective on each row. of rubbing clearances is that the “coefficient of discharge” of the constriction can be increased when produced by a rub. This can effectively increase the losses to values in excess of what might have been obtained by initially installing seals at the design values. Typically, a “mushroom” edge will have a flow coefficient greater than a knife-edge. The calculated losses at the blade tips in the above examples were determined using the equation of Martin. It is perhaps more correct to use the expression developed as equation 10.5.11. Figure 10.5.13(a) shows a comparison for a typical stage, where the Martin equation and the equation 10.5.11 values are compared for the blade tip section, shown as Figure 10.5.13(b). 583 Turbine Steam Path Troubleshooting and Repair—Volume Two In these calculations it has been assumed that the number of tip seals can be increased, i.e., no attempt has been made to consider the effects of stage geometry. This comparison is undertaken simply 0.060" "N" (2) Position (1) P = 1100 psia Vs = 0.6866 Cu ft/# H = 1444.8 BTU/# (1) 35.15" Position (2) P = 1030 psia Vs = 0.7250 cu ft/# H = 1437.6 BTU/# Fig. 10.5.13(a)—A stage with a variable number of radial seal strips. Leakage Flow #/sec. 45 40 Martin Equation Equation 10.4.11 35 30 25 20 15 10 0 1 2 3 4 5 6 Number of constrictions "N" in series Fig. 10.5.13(b)—A comparison of the calculated leakage quantities for the Martin and enthalpy drop methods. 584 Seals, Glands, and Sealing Systems to demonstrate the magnitude of difference between the two methods. In both cases the coefficient of discharge “ψ” has been assumed to be 1.0. It can be seen from the curves of Figure 10.5.13(b), that the difference between these two curves is relatively constant at about 30%. Example 10.5.1 Consider the impulse stage (diaphragm and disc) construction shown in Figure 10.5.14. Here the steam conditions are shown at the various stage points (1).....(4). Applying Martin’s equation to the leakage under the diaphragm, and above the rotating blade tip, the following leakage quantities can be found (assuming clearances under the diaphragm have a design value of 0.025" at a seal diameter of 33.6", and above the blade tip a clearance of 0.060", on a seal diameter of 41.6"). The flow coefficient should be considered constant at “ψ” = 1.0. Cl =0.060" N=2 (3) Posn. P Vs H (1) 1945 0.390 1442 (1) (2) 1690 1426 (3) 17250.426 1428 (4) 1662 1424 (4) (2) Ds = 41.6" C l= 0.025" N = 12 Ds = 33.6" Fig. 10.5.14—The stage details for example 10.5.1. Showing steam conditions and design dimensions at the seal positions. 585 Turbine Steam Path Troubleshooting and Repair—Volume Two Solution Diaphragm packing leakage The power loss is found from kW The effect of increased clearance can also be calculated, assuming the value of “ψ” does not change as the clearances open due to rubs: Cl 0.025 0.035 0.045 0.055 m 12.496 17.494 22.493 27.491 dh —————— 1442-1426 = 16.0 BTU/# dkW 211.0 295.4 379.8 464.2 0.065 0.075 32.490 37.488 ———————— 548.6 633.0 Blade tip leakage and the effect of clearance changes on leakage flow quantities, assuming an unchanged value of “ψ,” and the power loss is: Cl 0.055 0.065 0.075 m 40.508 47.873 55.238 dh —————— 1428-1424 = 4.0 dkW 170.96 202.04 233.12 586 0.085 62.598 BTU/# 264.19 0.095 0.105 69.963 77.327 ———————— 295.27 326.35 Seals, Glands, and Sealing Systems This variation of leakage loss is shown graphically in Figure 10.5.15. The effect of changing the number of series constrictions “N” can be found using the same equation to as follows: 800 Leakage Loss: kilowatts 700 Diaphragm Leakage 600 500 400 300 200 Blade Tip Leakage 100 0 20 30 40 50 80 60 70 Radial Clearance: 0.001" 90 100 110 Fig. 10.5.15—The leakage loss in kilowatts, as a function of the radial clearance for the diaphragm (12 seals effective), and the blade tip. Diaphragm packing leakage, (sensitivity to “N”). In the event the seals under the diaphragm are damaged, and the number of effective seals are reduced, there is a change in the leakage quantity. This is shown below: 587 Turbine Steam Path Troubleshooting and Repair—Volume Two These losses are shown graphically in Figure 10.5.16. From this numerical example, it can be seen that opened clearances, as in the impulse unit, have the potential to reduce the steam path efficiency, by reducing the total kilowatts developed in the section. This is because any steam that leaks through the seals and bypasses the blade rows is unavailable to do work there. Another loss that cannot be quantified is that associated with the steam reentering the main steam flow, and disrupting the streamline effect after it has leaked past the seals. These losses could be as high as those suffered through leakage, particularly on the small radial height blades. 320 Loss in output - kilowatts 300 280 260 240 220 200 180 5 7 9 11 13 Number of Effective Seals 15 Fig. 10.5.16—The sensitivity of steam leakage quantities to the number of effective seals. 588 Seals, Glands, and Sealing Systems Measuring radial seal clearance When a unit is opened for inspection, a normal maintenance action is to measure the seal radial clearances. This is normally done at the horizontal joint. At this time the unit is cold. In addition, the top half casing has been removed, and therefore the casing could have distorted to an elliptical form without the tension from the bolts. These factors can combine to give false readings, and the field operator is left to interpret these values. It is important to consider what should be measured, and how the values should be interpreted. At the horizontal half joint. The readings recorded at the half joint do not represent the actual running clearances. The clearances at these various positions are affected by three major influences, which include: • The relative radial movement of the steam path components during operation due to temperature and stress effects (discussed in chapter 2) • When stationary, the rotor does not lie on its true center of rotation, and therefore the seal gap may not be measured to be the same on the two sides. It is acceptable to assume a mean • The casing may have assumed an elliptical form as the horizontal joint bolts were released A word of caution, however, after the readings are taken. If they show a significant localized discrepancy on the two sides, it is suggested that when the rotor is removed, the seal strips be examined to determine if there is heavy localized damage affecting the true readings. It is also recommended that a visual inspection be made (before the clearances are measured) to determine if there are any abnormalities in the seal strip condition. The “feeler gauges” should be inserted into the horizontal joint as far as possible, and at a minimum of 1.5". 589 Turbine Steam Path Troubleshooting and Repair—Volume Two Casing distortion. When the studs are removed from the horizontal joint, it is possible for the casing to distort and move either inwards or outwards due to some metallographic change in the casing material structure, and the release of residual stress. In fact, some movement can be so severe the casing will “grip” the rotor, making it difficult to remove. Consider the casing, in its design position (shown diagrammatically in Fig. 10.5.17). In this position, there are clearances at the horizontal joint of “Kl” and “Kr,” and this clearance would also be measured at the vertical centerline as “Kb” at the bottom, and “Kt” at the top dead center. The seals also have a radial height “H” at each tangential location. Also shown are details of the left hand side, showing seal strip height as “Hl.” Kl Kr Dc Hl Kl Dc Design condition with the rotor central, and no casing distortion. Rotor Casing Seal Strip Kb Fig. 10.5.17—The theoretical clearances ‘K’ and seal strip height ‘Hl’ around the steam path. When the casing has distorted, the clearances will change, and possibly the seal height “Hl” will not be equal at all tangential positions, particularly if there have been localized rubs during operation. 590 Seals, Glands, and Sealing Systems When the casing is disassembled, the horizontal joint diameter can change, either increasing or decreasing. When the horizontal joint diameter increases, there will be a corresponding decrease in the vertical centerline height. Similarly, if the horizontal joint decreases there will be a corresponding increase in the vertical centerline height. These moves must be taken into account in determining the radial clearance, and thus in calculating the leakage quantity. However, the amount of horizontal joint movement, “in” or “out,” will not determine the vertical movement. This will be determined only by the degree of distortion. The horizontal joint details must be measured to establish the clearances and the seal heights. The seal strip heights “Hl” and “Hr” are measured on both sides, with the measured clearance “Kl” and “Kr” on the left and right hand side of the casing. The fact that there has been casing movement can be determined from accurate measurements of the half joint diameter “Dc” at the horizontal joint and from a vertical centerline height measurement “Vd,” shown in Figure 10.5.18. These measurements must be made to the casing outer sidewall, not the seal edge. The effective clearances can be determined in the following manner: Mean horizontal joint clearance = (Kl + Kr)/2 From measurements and comparison of the diameter “Dc” and drop “Vd,” the casing distortion can be determined from the fact that for a totally cylindrical casing: 2. Vd = Dc Assuming any movement in the casing is the same magnitude on both sides, gives side joint movement per side of “dK”: dK = (Dc/2) - Vd. Casing width increasing at the horizontal joint dK = Vd - (Dc/2). Casing width decreasing at the horizontal joint 591 Turbine Steam Path Troubleshooting and Repair—Volume Two When these casing half joint measurements are complete and the casing movement “dK” has been determined, it is then necessary to determine the cold clearance at the top and bottom “dead center positions.” If the measured heights of the seal strips at the left and right hand sides are “Hla” and “Hra,” (the “a” indicating actual measurements), then the cold clearances can be determined: Cold clearances “Kla” and “Kra” with casing bolted closed are: Kla = Kl +/- dK, and Kra = Kr +/- dK It is then necessary to establish the seal height “Hl,” “Hr,” “Ht,” and “Hb” at the four quadrant positions. The cold clearances at the top and bottom positions “Kta” and “Kba,” can then be determined in terms of the seal heights at the left “Hl” and right “Hr.” Recognizing that these readings are taken without the casing bolted, and cold, it must be accepted that errors can, and will, exist. However, the most significant consideration in taking clearances and determining leakage losses is to be sure the procedures for measurements are always the same. This is necessary so that differences from one time period to the next are consistent. The leakage loss calculations are used to determine clearance differences, and these differences can only be compared by the use of a repeatable procedure. Clearances at the top and bottom dead center “Kb” and “Kt” are also taken with leads. However, unless the casing is bolted the clearances will not define the ellipticity of the casing. Also, if the casing half joint has increased as the joint bolts were removed, it will be impossible to bolt after the leads are used, as the rotor in the unbolted condition will already compress them. Measuring seal axial clearances The axial movement during operation is controlled by the differential expansion (see chapter 2), and the relative movement of the 592 Seals, Glands, and Sealing Systems Dc L Vd Fig. 10.5.18—The drop check (vertical centerline height) from the horizontal joint, to establish the presence of ovality within the casing. rotor from the thrust block. There are situations where the running axial clearance between the stationary and rotating components of the unit change and decrease. This movement can occur to the extent that the axial seals, possibly produced integrally with the coverband will rub, removing material from the knife-edge, opening the axial clearance, and increasing leakage loss. These clearances must be measured as the steam path becomes available for inspection, and the leakage loss determined. The equation of Martin will allow such determination. Calculated values of incremental leakage loss When a unit is opened for inspection, and clearances are measured, it is clear that even if no wear has occurred, the measured value of clearance does not represent the “hot running” condition. The actual running clearance is modified by those phenomena that modify the spatial relationships within the steam path (see chapter 2). However, these changes are relatively small in the radial 593 Turbine Steam Path Troubleshooting and Repair—Volume Two direction, and tend to be of a value that they can be ignored in terms of their effect at the “hot end” of the unit, which is where leakage is most significant. Therefore, if the losses are calculated on these measured values, they will provide a loss that is in excess of the actual value. This is because the rotor will tend to move towards the casing during operation, reducing this measured value. In terms of determining “increased leakage losses due to rubs,” the measured clearance values at an outage can be used with complete accuracy, when compared to previous and design values, to establish the increased losses. As an example, consider a location within the unit at which the design radial clearance is specified as 0.025", and at installation the measured clearance was 0.027" (which represents design conformance, and is within tolerance). If this same clearance is measured on removal of the unit from service, and the measured value is 0.056", then there will have been an opening of the seals of (0.056-0.027 = 0.029"), if the loss rate is 3.9 kW/mil. Under these circumstances, the losses are: Anticipated design loss: As installed: As removed from service: 25 x 3.91 = 97.75 kW 27 x 3.91 = 105.57 kW 56 x 3.91 = 218.96 kW Therefore, the recoverable losses are: 218.96 - 105.57 113.39kW or alternately: 218.96 - 97.75 = 121.39kW This recoverable loss is dependent upon restoring the clearance to its design value. Therefore, the measured opening represents a true loss of (56 27) = 29 x 3.91 = 113.39 kW, or 121.21 kW, which loss can be recovered by replacing the seals, and re-establishing the clearance at 0.027" or 0.025". If the clearance were re-established at some other value, the output loss would change accordingly. 594 Seals, Glands, and Sealing Systems THE ECONOMICS OF SEAL MAINTENANCE The one category of loss over which station maintenance staff can exert a significant influence, in terms of being able to plan for, gauge and take corrective action, is the control of the internal leakage, allowing the steam to bypass the steam path blade elements. During operation the most likely factor to influence the sealing efficiency is for a rub to occur. Such a rub wears the seal, and has two detrimental effects. First, it will open the clearance, causing an increase in leakage area. Secondly, the rub will modify the form or shape of the seal strip, possibly increasing the “discharge coefficient.” Unfortunately, both these effects tend to increase leakage flow and therefore cause a deterioration of expansion efficiency. It is important for the operator to know what level of increase in leakage area can be tolerated, and the financial penalty associated with such leakage. It is essential to consider these questions, and while it is difficult to provide an all encompassing answer, operators should be aware of the magnitude of this leakage for their units, and what this represents in terms of financial penalties. Some manufacturers provide operators with guidance concerning the loss rate in kW/mil, and this can be used to establish the need for replacement of the seals, assuming they are a type that can be replaced. These recommendations are made usually as a function of seal wear. Such “rules of thumb,” while being an acceptable guide to fuel costs, do not allow the operator to readily assess the effect of changing fuel costs, load factors, and the value of incremental kilowatts. The following analysis, although requiring a greater level of information preparation on the part of the owner, will normally allow him to factor in more variables that influence overall operating costs. 595 Turbine Steam Path Troubleshooting and Repair—Volume Two If a unit is opened only every four to eight years for planned maintenance, there are considerable advantages to having tools available that allow a quick and accurate assessment of the losses, and financial penalties that can be anticipated during the next operating period. The following method allows the owner to predict additional anticipated fuel costs associated with leakage. However, there are several factors that must be considered when using this analysis: • To be of real value, it is necessary for the operator to know the mil loss rate for the various seal positions in the unit. This can best be achieved by the construction of curves, or a calculated loss rate for each position in the unit • The operator should be able to predict, with reasonable accuracy, the fuel cost changes during the next operating period, and before the next planned outage. While difficult, this will allow a more accurate prediction of total revenue loss If the actual fuel costs change in an unpredictable manner during the operating period, such an analysis can be used to help establish the possible economies of reducing the operating period to the next outage, if the system can tolerate a change. • The operator must know the station heat rate (SHR), i.e., the heat rate for the boiler-turbine-generator cycle • The operator must be able to predict the load factor (LF) and possibly its variation, with some accuracy for the next operating period With these factors established, within acceptable limits, it is possible to determine, with reasonable accuracy, what additional fuel costs will be incurred during the next operating period due to excess leakage. Alternatively, it is possible to estimate the fuel cost savings available from making improvements to the seal system, and to justify the cost of changing seal strips or segments. 596 Seals, Glands, and Sealing Systems Example 10.6.1 Consider a unit with an annual load factor of 80% (0.80), having a station heat rate (SHR) of 10,000 Btu/kW-h, and burning fuel that costs 200c/million Btus. It is necessary to predict the annual additional fuel cost per kilowatt, as the result of seal wear and increased leakage. With this unit, it costs 200c to generate 1,000,000/10,000 = 100.00 kW for 1 hour. Therefore, any action that can be taken to improve output by one kilowatt represents a savings. The annual fuel saving per kilowatt is equal to: Fuel Cost = 200c/10,000kW = 0.020c/kW $ (Savings) = 1 x 8,760 x 0.8 x 0.020 = 140.16 $/kW/annum. It should be noted that this savings is independent of the unit rating, i.e., a kilowatt lost from a 30,000 kW unit is just as expensive as a kilowatt lost from a 750,000 kW machine, if the fuel costs are the same. Figure 10.6.1 shows a nomogram, which permits additional costs on a per kilowatt basis to be established for a variety of fuel costs and load factors (LF) for a unit with a station heat rate of 10,000 Btu/kW-h. For other station heat rates, the losses/kilowatt can be determined by the ratio of the actual “SHR” to 10,000 Btu/kW-h. 597 Turbine Steam Path Troubleshooting and Repair—Volume Two Figure on Curves are fuel costs in c/million BTUs SHR = 10,000 BTU/kw-Hr. 220 300 250 200 Incremental Fuel Costs $/Annum/Kilowatt loss 275 225 180 200 160 175 140 150 120 125 100 100 80 75 60 40 20 30 40 60 70 80 90 50 Annual Load Factor %. 100 Fig. 10.6.1—The annual fuel cost savings in $/kW as a function of fuel cost and unit load factor. Example 10.6.2 A unit consumes fuel costing 200c/10E6 Btus, has a load factor of 80%, and a station heat rate of 9,650 BTU/kW-hr. What are the losses per kilowatt of lost output due to leakage? 598 Seals, Glands, and Sealing Systems From Figure 10.6.1, the loss for a unit with a SHR of 10,000 Btu/kW-h is: $140.16/kW/Annum The loss per kilowatt for an SHR of 9,650 Btu/kW-h/year is: = $140.16 x 9,650 = 135.25$/kW/Annum 10,000 In terms of assessing the losses over an extended period of operation, and increasing clearance due to rubs, the following analysis can be made. Predicting kW lost due to excess clearance The normal manner of predicting leakage quantities and converting these to kilowatts is to use Martin’s equation for leakage past labyrinth seals. However, a simplified, although less accurate method, is to determine a mean leakage for various stage and unit configurations, and apply these mean losses to measured clearances. It is difficult to specify an overall loss rate for any location. Typical ranges are shown below; the actual values will depend upon the unit arrangement: 2,400 psi/1,000/1,000°F High-pressure section/flow Reheat section/flow Low-pressure section/flow N1 shaft packing N2 shaft packing N3 shaft packing 4.25 2.25 0.75 6.75 16.00 1.50 to to to to to to 10.00 kW/mil 8.25 kW/mil 1.25 kW/mil 19.00 kW/mil 55.00 kW/mil 5.00 kW/mil 599 Turbine Steam Path Troubleshooting and Repair—Volume Two 3,500 psi/1,000°F/1,000 F High-pressure section/flow Reheat section/flow Low-pressure section/flow N1 shaft packing N2 shaft packing N3 shaft packing 5.00 3.25 0.75 7.00 20.00 1.75 to to to to to to 12.25 kW/mil 9.50 kW/mil 1.25 kW/mil 21.00 kW/mil 55.00 kW/mil 6.25 kW/mil Example 10.6.3 As an example of this estimation, consider a unit having a combined high-pressure (HP)/reheat section, and a single double-flow, low-pressure section. The high-pressure section contains 7 stages (6 diaphragms and a nozzle box), the reheat section has 6 (6 diaphragms), and each low-pressure section has 7 stages (6 diaphragms and an inlet nozzle box). This unit is shown diagrammatically in Figure 10.6.2. Loss Rate/0.001" 8.00 12.50 HP Cl (N1) = 28 1.00 6.25 Double Flow LP Rht Cl (N2) = 28 Cl (HP) = 37 LP 'A' LP 'B' Cl (LPb) = 28 Cl (LPa) = 28 Generator Cl (N3) = 28 CL (Rht) = 30 Clearance in 0.001" 1.00 3.25 40.00 Fig. 10.6.2—The radial clearance and loss rate in kW/0.001" for the unit in example 10.6.3 This unit has steam conditions of 2,400 psia/1,000ºF/1,000ºF an “SHR” of 10,150 Btu/kW-h, and a mean load factor of 75%. The fuel costs 175c/10E6 Btus. 600 Seals, Glands, and Sealing Systems As removed from service, the high-pressure stages have rubbed open an average of 37 mils, the reheat section has an increased clearance of 30 mils, and the low-pressure stages have an increased clearance of 25 mils. The N1, N2, and N3 packings have opened an average of 28 mils. In this unit the assumed loss rates in kilowatt/mil are: HP 8.00, Rht=6.25, LP=1.00, N1=12.50, N2=40.00 and N3=3.25 kilowatts /0.001". The total leakage losses are: HP section: Reheat section: LP sections N1 packing: N2 packing N3 packing: 6 x 37 x 8.00 6 x 30 x 6.25 2 x 6 x 28 X 1.00 1 x 28 x 12.50 1 x 28 x 40.00 1 x 28 x 3.25 Total losses are: = = = = = = 1,776.00 1,125.00 336.00 350.00 1,120.00 91.00 4,798.00 kW kW kW kW kW kW kW With a fuel cost 175c/10E6 Btus, and a load factor of 75%, it can be determined from Figure 10.6.1 that the annual cost of leakage was $115.00/kW. Therefore, for the last year of operation the annual fuel cost increase was: 4,798 x $115 = $551,770 The cost in the previous years would naturally have been less, assuming the damage in those years (immediately after startup) was less severe. 601 Turbine Steam Path Troubleshooting and Repair—Volume Two Example 10.6.4 As an example, consider a unit with an annual load factor of 70% (0.70), having a station heat rate of 10,000 Btu/kW-h, and using fuel that costs 200c/million Btus. It is necessary to predict the annual additional fuel cost as the result of seal wear. It is assumed the SHR and LF will remain constant over the next operating period. Solution With this unit it costs 200c to generate 1,000,000/10,000 = 100 kW for 1 hour. Therefore, any action that can be taken to improve output by one kilowatt represents a savings. The annual fuel saving per kilowatt is therefore equal to: Fuel Cost = 200c/10,000kW $ (Savings) = 1 x 8,760 x 0.7 x 0.02 = 0.020c/kW = 122.64 The curve in Figure 10.6.1 shows the annual fuel cost savings for fuel costs from $0.75 to $3.00 per million Btus (75 to 300 cents), as a function of unit load factor, and for a SHR of 10,000 Btu/kW-h. As an example of the application of this curve, consider a 500,000 kW unit having an SHR of 10,000 Btu/kW-h, and a predicted load factor of 70%. Also assume maintenance work has reduced the leakage loss by 2,350 kW. In this case, this represents an annual fuel saving of: 2,350 x 8,760 x 0.7 x 0.02 = $288,204 (with a fuel cost of 200 cents/million Btus) Note: There are 8,760 hours in a normal 365-day year) This amount represents the savings in the first year after return to service. For subsequent years it is possible the fuel costs will increase. It is also certain the initial improvement in clearance will 602 Seals, Glands, and Sealing Systems not be sustained throughout the operating period to the next maintenance outage. Based on these assumptions, and a five-year maintenance cycle, the following cost savings can be anticipated, when the load factor remains at 70% (0.7). Years after return to service Remaining kW improvement (assumed) 2 3 4 5 2,010 1,860 1,760 1,670 Predicted Fuel Cost Annual Fuel Cost Savings 210 221 232 243 Total savings $1,258,832 $1,252,062 $1,250,382 $1,248,843 $1,010,119 These savings of $1,010,119 together with those in the first year ($288,204) represent a total potential saving of $1,298,323. If these later years are returned to original year dollars at the rate of 8%, this represents a total present day worth of $1,125,000. It is clear from this example that considerable savings in operating costs can be achieved by attention to maintenance of the sealing system. Or, when output is limited by equipment of the cycle other than the turbine, limiting, or controlling the leakage flow can achieve some increase. One factor that needs to be considered, however, is that the clearances can be opened to the five-year level by one bad operating experience. Therefore, it is necessary to limit transient, and other phenomena known to cause seal damage as much as possible. The extent to which such improvement can be maintained, is dependent upon the manner in which the unit is operated, and the transients to which it is subjected. However, if seals have been maintained at, or near, the design or installed values, it is usually possible and economically justifiable to examine the unit to determine what level of improvement can be made. 603 Turbine Steam Path Troubleshooting and Repair—Volume Two It is recommended the unit seal system loss rates be analyzed before the unit is removed from service. In this manner, measured clearances, and the predicted loss rates can be determined. Then when clearances are measured at the outage, decisions can be made immediately in terms of the extent of wear and damage that can be tolerated at the various seal points. It is normal for both “utility” and “industrial” operators to carry, as inventory spares, many of the seal strips and gland segments that could be used in making upgrading repairs. This is a prudent practice and allows remedial action to be taken at short notice. In assessing changes and measuring the effect of any improvement on the unit after maintenance, it is recommended, and considered necessary, to calibrate the unit performance by means of some performance test. It is strongly suggested that in the undertaking of such tests, that instruments and procedures, which accord with a recognized Power Test Code be observed. The most important aspect of such tests is to ensure repeatability of the results, and then to ensure they are run at a frequency that monitors condition, and helps assess the need for further maintenance. FORMS OF THE SEAL KNIFE-EDGE DISCHARGE COEFFICIENTS The form of the seal strip (or tooth edge) at discharge from the upstream to downstream positions has a considerable influence on the total flow that occurs at any seal constriction. In establishing the flow using the equation of Martin, a discharge coefficient of “ψ = 1.0" was assumed. In fact, the coefficient of discharge will be other than this value and a reasonable mean of “ψ,” if no better data are available, is 0.82. This 604 Seals, Glands, and Sealing Systems value represents an acceptable mean and can be used with the knowledge that it represents a value suited for many seal configurations. However, leakage flow is directly proportional to the discharge coefficient. Therefore, there could be situations where the use of a more accurate value is justified. The difficulty in establishing absolute values of “ψ” is that the shape of the seal strip influences the value, and the only absolute method is to undertake expensive tests to determine the most appropriate coefficient in each design. It is also necessary to consider the seal location, the form of the surface on which the seal constriction is produced, and general stage geometry at that location. Knife-edge form The discharge point on the seal has a marked effect on the discharge coefficient. Considerable effort and manufacturing expense can be justified in achieving the design requirements. Figure 10.7.1 shows the simplest form of sealing strip, which is a simple constant section strip of width “d.” Figure 10.7.2 shows various forms of the tapered knife-edge. In (a) is the form where a taper is produced on one side, and in (b) there is a two-edge taper. In (c) is the stepped tapered form, in which the final thickness is “x,” and is maintained over a radial height of “Li.” In each case, from a thickness “d” to a final knife-edge of thickness “x,” the taper occurs over a radial length of “L.” There are two reasons for bringing the discharge point to a knife-edge—first, to reduce the leakage quantity by reducing the flow coefficient, and secondly, to cause less heat to be generated in the event of a rub on the mating surface. In the event of a rub, the knife-edge will be deformed to a shape like that shown in Figure 10.7.3(a). If the inner diameter of the seal has been increased from “Ds” to “Dt,” there are considerable advantages to “dressing” the remaining strip to a form like that shown in Figures (b), (c), and (d), if the preferred option of changing the strip and restoring the original seal diameter “Ds” is not possible. 605 Turbine Steam Path Troubleshooting and Repair—Volume Two The forms of strip described in Figure 10.7.1 and 10.7.2 are inserted, with the seal strip being in a true radial direction. There are also seals where the inclination of the seal strip is at an angle “λ” as shown in Figure 10.7.4. These strips are normally produced in segments, and the final production of the seal diameter “Ds,” can be difficult to machine if the segment is spring loaded, but can be trimmed to the correct diameter to achieve the design clearance. L d Fig. 10.7.1—The simple inserted seal strip Fig. 10.7.2—Various options for forming the tapered knife edge. 606 Seals, Glands, and Sealing Systems Fig. 10.7.3—Methods for finish trimming a damaged seal strip after a ‘rub’. With the exception of the angled strip (shown in Fig. 10.7.4), the strips are inserted into the main body of the carrying component, either the casing, the rotor, or gland ring. However, other forms have the strips formed integral with the ring of the seal gland. Again, these strips are tapered to a knife-edge of thickness “x” and have an included angle “λ.” These seal strips cannot be replaced, and should a significant rub occur, the seal must be replaced, or sharpened, but to a larger diameter, which increases the leakage flow. λ E X Fig. 10.7.4—A formed seal strip. This shape of strip is manufactured using form tools and provides for greater axial strength. 607 Turbine Steam Path Troubleshooting and Repair—Volume Two Seal location geometry The geometry of the component at the seal location has considerable influence on the leakage quantity. The seal operates on the basis of converting the steam thermal potential energy to kinetic energy, and then destroying that kinetic energy to the greatest extent possible. Consider the seal shown in Figure 10.7.5(a). Here a simple, single constriction is provided, and the seal provides a barrier to flow, which reduces the leakage. However, the kinetic energy is not destroyed, and the steam flows freely into the downstream space. In Figure 10.7.5(b), a vertical face is presented to the leaking steam, which effectively reduces the velocity, and therefore lowers the leakage quantity. This is because a portion of the kinetic energy is reconverted to pressure at discharge, reducing the pressure ratio across the seal and therefore reducing the leakage steam quantity. At some locations a vertical face is produced by the inclusion of a diverting device, as shown in Figure 10.7.6 for two locations, in (a) on the coverband, and in (b) as a rotor castellation. In Figure 10.7.7(a) two strips are used, and a chamber is formed between them. This chamber acts to completely destroy the velocity, but some small quantity will be carried through the second leakage area. It can be seen that the horizontal pitching “p” between the strips can become too close to allow the effective destruction of the kinetic energy. The design engineer must consider this, and there is no advantage to reducing the pitch to increase the number of seal strips that can be accommodated in any axial length. In Figure 10.7.7(b) the strips are mounted alternately in the stationary and rotating components of the seal region. The steam is provided with a tortuous path, and a considerable portion of the energy is destroyed by the continual change of stream direction. Here the axial gap between stationary and rotating seals strips is governed by the differential expansion that occurs, and this pitch “p” will change with changing operating conditions. 608 Seals, Glands, and Sealing Systems Cl (b) (a) Fig. 10.7.5—Two arrangements of a single seal. In (a) the seal is formed on a plane cylindrical surface, and in (b) having a vertical face following discharge from the construction. Cl Clu Clu Cl (a) (b) Fig. 10.7.6—A central diverting ‘rib’ in (b) helps destroy the kinetic energy of the steam. Clu p (a) Cr p Cll (b) Fig. 10.7.7—Two seal strips forming a chamber between them. In (a) the seals are mounted in the same component, producing a clearance ‘Cr’. In (b) the seals are mounted in alternate components forming an upper clearance ‘Clu’ and a lower clearance ‘Cll’. 609 Turbine Steam Path Troubleshooting and Repair—Volume Two Fig. 10.7.8—Discharge coefficients for a single seal in terms of the seal geometry. Relative Leakage Flow Sealing Design Cl 1.00 Cl 0.93 Cl Cl Cl 0.92 0.88 0.86 0.73 Cl 0.62 Single Strip Small Groove Medium Groove Large Groove Coned Groove HoneyComb Double Strip Labyrinth 0.40 0.75 Spaced Labyrinth Fig. 10.7.9—Various seal configurations and their flow coefficients. 610 Seals, Glands, and Sealing Systems Figure 10.7.8 shows a base discharge coefficient “ψ” for a straight through seal. These values were established for a single seal, and can be used for different values of the ratio of pressure drop across the seal. This curve makes some attempt to factor in the geometry of the seal. Figure 10.7.9 shows a variety of seal configurations, and relative flow corrections as affected by the seal geometry relative to the straight through type. These are “shape factors” and will influence the flow coefficient. Figure 10.7.10 shows a series of curves representing the flow rates, as a percentage of the straight through type, for different configurations of the strips, and general stage geometry. Similar information has been presented for tip seal configurations, and the form effect on flow coefficient. These various forms and corresponding flow coefficients are shown in Figure 10.7.11. Fig. 10.7.10—Flow factors for various arrangements of the seal strip. 611 Turbine Steam Path Troubleshooting and Repair—Volume Two Flow coefficient 0.58 0.46 0.46 0.38 0.30 Fig. 10.7.11—Flow coefficients for various seal strip configurations at the blade tip. Multi strip seal configurations In addition to the importance of the individual strip form, the arrangement of the seals in a multi strip configuration must also be considered, in terms of their spatial position relative to each other. Some possible arrangements will now be considered. In each design, the arrangement is influenced by certain dimensional and operational factors, and the designer makes the ultimate selection to provide the most effective seal possible at the axial location being considered. The most common configurations are: 612 Seals, Glands, and Sealing Systems The straight through design. The straight through design is the simplest multi strip arrangement, and is shown in Figure 10.7.12(a) and (b). In (a) the strips are mounted to the stationary component of the unit, and in (b) to the rotating component. In each design a series of seal strips, each having a radial clearance “Cl” are arranged along the leakage path, pitched at “P” apart. The distance “P” is selected to provide an adequate chamber between the strips, and enough to destroy the steam velocity at entrance. The seal diameter is marginally different from one arrangement to the other, being a function of the seal strip height “h.” This difference is minimal when determining leakage quantities. P Stationary Cl h (a) Rotating Ds P Stationary (b) h Cl Ds Rotating Fig. 10.7.12—The straight through design, with the seal strips mounted in the stationary portion of the unit in (a), and in the rotating portion in (b). 613 Turbine Steam Path Troubleshooting and Repair—Volume Two The seal arrangement is convenient when the differential expansion between the stationary and rotating components is such that clearances at different diameters could not be accommodated into the arrangement. The alternative stationary/rotating configuration. In this design, the caulked attached strips are located alternately in the stationary and rotating components of the unit, as shown in Figure 10.7.13. The strips are pitched at a distance “P” apart, and are located from stationary to rotating component, a distance “Q,” which is selected so that during operation there will be no contact between them. There is also a reference setting from some point, such as the thrust block or coupling flange, such as “G.” This will assure axial positions are predictable under all operating conditions. It is normal for the upper clearance “Clu” and the lower “Cll” to be of the same value, but the seal pitch does vary. When calculating leakage quantities, the mean of “Dsu” and “Dl” should be used. This is an effective seal, and as seen in Figure 10.7.9, does produce an effective barrier to leakage. This figure illustrates that the last two arrangements of the pitching of the strips has a considerable effect on sealing efficiency. Cu Stationary Cl Dsu Rotating Dsl Q P G Fig. 10.7.13—The alternate stationary and rotating locations of the seal strips produces a tortuous leakage path. 614 Seals, Glands, and Sealing Systems The simple hi-lo staggered configuration. The “hi-lo” configuration can be used where differential expansion is relatively small, and the designer wishes to avoid the need to locate seal strips in the rotating components of the unit. This arrangement is shown in Figure 10.7.14, where the clearances on the rotor body “Cll,” and the castellation “Clu” are the same, and there is a minimal difference in the seal diameters “Dsu” and “Dsl.” The requirement of cold setting “Q” and “G” are the same at the previous arrangement. In designing this arrangement, it is necessary to ensure the castellation width “W” is sufficient, such that under all operating conditions the seal strip is located above the seal platform. The multi high strip staggered configuration. In those locations where differential expansion will not permit the use of the form of seals shown in Figure 10.7.14, the form shown in Figure 10.7.15 can be used. In this design there are two high strips. Dependent upon the extent of load, and temperature distribution throughout the steam path, only one of the two high strips is effective. The high strips are pitched in a way that at all rotor axial positions one strip is active, and the other has a radial clearance equal to “Cu + s,” which is too high to be considered effective in limiting P Stationary Cl Rotating Dsl W G Q Fig. 10.7.14—The simple ‘Hi-Lo’ staggered labyrinth arrangement used on a castellated rotor location. 615 Turbine Steam Path Troubleshooting and Repair—Volume Two steam leakage. Therefore, when determining the value of the number of strips “N” in Martin’s equation, only one of each pair should be counted as limiting leakage. Stationary Clu Cll s Rotating Dsl W Q Dsu G Fig. 10.7.15—The ‘hi-lo’ arrangement for axial locations with a large differential expansion. Here two ‘hi’ seals at each location help ensure one seal is effective at all axial positions of the rotor. A shaft end packing arrangment is shown in figure 10.7.16, which shows a shaft end packing arrangement. This shows the strips that are located in a gland ring, and the horizontal joint securing screws that are required to prevent the gland rings from tangential migration during operation, which would prevent the disassembly of the top half of the gland housing. The multi high strip configuration. Shown in Figure 10.7.17 is the seal system employing what is known as “herringbone seal strips,” which are used by some manufacturers. In this type of seal, the seal strip can be formed as a continuous helix, with seal strips machined in such a manner (on both the stationary and rotating components) that a small positive radial clearance “Cl” is formed between them [see Fig. 10.7.17(a)]. The actual leakage area is in fact controlled by the total clearance shown as “Ct” in Figure (a). However, this a very effective seal due to the geometry, and the fact that there is (during operation) relative movement between the stationary and rotating components. 616 Seals, Glands, and Sealing Systems Fig. 10.7.16—Gland rings at a shaft end sealing position. Stationary Rotating Cl = + ve (a) Ct Stationary Rotating Cl = 0 (b) (c) Cl = - ve Stationary Rotating Fig. 10.7.17—The ‘herringbone seals’. This design can be used with (a) a positive clearance, with (b) zero clearance, and (c) a negative clearance, where there is no large differential expansion. 617 Turbine Steam Path Troubleshooting and Repair—Volume Two The major disadvantage to this seal is that in the event of rotor vibration, the seals will rub and cause damage that is difficult or impossible to repair. The effective clearance at any point “Ct” can be seen from this Figure (a) to be considerably larger than “Cl.” However, it is difficult to determine the actual value of “Ct” at any point because of the relative motion between the parts. If the seal strips are cut to a helical path there is no one value to “Ct,” and the strips’ clearances vary due to their relative and changing pitch position around the circumference. This type of seal has been found to be very effective, and having a tendency to reverse the flow direction, destroys the kinetic energy of the steam. Depending upon the amount of differential expansion in the vicinity of the seals, they may be arranged with a zero (or negative) radial clearance “Cl,” as shown in Figure 10.7.17 (b) and (c). In such a design the seal strips must be rings; the helical form cannot be used in (c). These herringbone seal strips are also used on low-pressure segments where there is large differential expansion. FORM OF THE GLAND RINGS Many of the seals formed in the unit are produced as rings, which can be inserted into the stationary components of the unit, and then located around the rotating components. These rings can be accurately located in both the axial and radial directions. This represents a convenient arrangement, as it allows rings to be replaced with relative ease when the unit is open for maintenance inspection. These rings can be classified into two basic types. First are those in which the seal strips are produced integral with the ring, in which design the strips are generally of the form shown in Figure 618 Seals, Glands, and Sealing Systems 10.7.4. The second type are those in which the seal strips are inserted, and then staked into the ring. The latter type can, in certain circumstances, have the strips removed and replaced. The rings must be split at their horizontal joint to facilitate assembly and disassembly, and to allow the segments to make a radial adjustment during operation. It is normal for the rings to be produced so a complete ring is formed by 4, 6, or 8 equal arc segments. This allows the segments to be capable of individual radial alignment. Depending upon the local environmental temperature, these rings can be produced from either chrome steel, or a copper base alloy. Carbon rings The type of seal used by earlier design units, of smaller output, was a carbon ring. Such a ring arrangement is shown in Figure 10.8.1. In this design, the seal is produced by a three-segment carbon ring. This ring is arranged and broken at the horizontal joint on one side of the rotor. Also, at this same location it is keyed to the gland housing to prevent operational rotation. The ring is held in intimate contact with the rotor with a garter spring. These carbon segments normally have closed-butt joint segments. By providing a large axial taper on the outer diameter of the carbon ring, as shown in Figure 10.8.1, the spring is also able to exert enough axial force on the ring to maintain contact between the ring and housing, and so minimize end leakage over the assembly. The bore of the carbon ring is normally sized to provide a “cold clearance,” which is taken up as the carbon ring reaches operating temperature. This type of carbon gland (on earlier design) had a small butt clearance “g” between the ring elements, as shown in Figure 10.8.2. However, these were suitable only for low-pressure wet steam applications, and were used at low rubbing speeds. 619 Turbine Steam Path Troubleshooting and Repair—Volume Two Load applied by garter spring both radially inward and axially to produce a steam seal. Atmospheric leak off Carbon Rings Turbine Shaft 360° Garter Spring Drain Fig. 10.8.1—The carbon gland ring system with ‘garter springs’. g 8 - 45° segments g Stationary Securing g/2 Screw Gland Ring Fig. 10.8.2—The gland ring, showing the tangential gaps between the segments. Butt clearances and tangential location As gland rings are mounted into the stationary components of the unit, they are adjusted to have sufficient tangential space (butt clearance) between the individual segments. This allows them the ability to move radially to accommodate movement of the rotor, and not bind, or become “arch bound” with changing steam temperatures when they heat at different rates compared to the major components 620 Seals, Glands, and Sealing Systems of the unit. Therefore, it is necessary for them to have carefully selected tangential clearances between the segments, to allow for conditions of temperature transients, both at “start-up,” “shutdown,” and at any prolonged temperature change during operation. It is also necessary to ensure this butt gap should not be so large that excessive bypass leakage will occur. The butt clearance is normally specified by the design engineer, and must be achieved as the rings are assembled to the stationary components of the unit. The butt clearance is normally a “field adjustment.” The individual gland rings are also restrained from tangential migration during operation. Securing devices at the horizontal joint provides tangential locking. Gland ring spring loading During operation the gland ring is held in a radially inward position by the steam force developed behind it. This steam pressure also forces the gland ring axially downstream enough to give it positive location, and to form a steam seal. At “start-up,” and before the steam pressure develops in the steam path, it is necessary to ensure the gland ring moves to a radially inward position so the steam seal can be formed between the stationary carrier and the ring segments. To achieve this initial seal, it is normal to employ a spring behind the gland segments, to initiate the seal, and maintain alignment until the steam pressure can become effective. There are various forms of springs, from leaf to close coiled. The selection for any application depends upon the preference and experience of the designer, and the ease of achieving the initial assembly. 621 Turbine Steam Path Troubleshooting and Repair—Volume Two The gland ring and carrier geometry Figure 10.8.3 shows the basic geometry of a gland ring, and the receiving slot into which it is mounted. In this case there is a closecoiled helical spring, but other spring systems are also used with equal effect. W w1 w2 Co G l H Ci Cl Dr Dt S H Ci Fig. 10.8.3—The basic geometry and principle dimensions of the “T’ head for the gland ring. In this design the gland ring “T” head has a total width “W,” and the “tee” slot has a receptacle head width of “w1+W+w2,” which is sufficient to permit assembly without interference. This provides for clearances of “w2” on the high, and “w1” on the low-pressure side. In fact, the pressure in both slot clearances is identical, and equal to the steam admission pressure to the “T” slot. 622 Seals, Glands, and Sealing Systems In the radial direction, the seal strips of the gland ring achieve a radial clearance “Cl” between a rotor diameter “Dr,” and the seal strip inner diameter “Dt.” In the event the rotor causes a rub, and the gland segments move radially outwards, there must be sufficient clearance between the ring surface and the inner surface of the carrier “Di,” so that contact will not occur. This requires that radical clearance “Ci” is sufficient. There is a further radial clearance between the carrier ring inner surface and the “T” head of the gland ring equal to “Co.” This is of no significance except in establishing the required length or geometry of the springs. To ensure steam is admitted to the space behind the gland segments, a slot “G” is provided in the segments, on the high-pressure side to allow easy admission. The steam seal To achieve a suitable seal, it is necessary to produce seal faces (both axial and tangentially) between the gland segments and housing faces. Figure 10.8.3 shows in detail the downstream portion, where the gland ring mates with the housing surfaces. To achieve an acceptable seal, it is necessary that axial distance “S” and radial distance “H” has a surface that is at least 125√µ-inches, and preferably 64√µ-inches. However, the difficulty of producing these surfaces on the gland housing must be recognized. On the gland rings this becomes relatively easier, and for normal machining methods it presents no difficulties. When replacing gland rings in a unit that has been in service, it is common for there to be oxide scale and other impurities present in the regions where seals are formed. It is a good practice to clean the surfaces intended to form seals. Since it is not possible to blast these surfaces with any expectation of success, a removed gland ring with grinding paste on the seal surfaces can be used to remove a significant portion of any debris and scale. 623 Turbine Steam Path Troubleshooting and Repair—Volume Two The seal and radial/axial steam forces The steam forces developed on the gland segments, and holding them in a radial inward location, need to be considered, because if the seal fails, excessive leakage can influence these total forces, causing the seal to become partially ineffective. Figure 10.8.4 shows a single gland with eight seal strips, each reducing the pressure from inlet conditions “Pa” to discharge “Pj.” In this gland there are forces developed in the radial inwards and outward directions, as well as forces acting both up and downstream. Curves of these forces are also shown. Figure 10.8.4 (a) shows the principal dimensions around a gland segment. In (b), (c), (d), and (e) the principal forces developed by the steam on this ring are active during normal operation. The steam radial inward force “Fi” is due to the steam pressure “Pa” acting on the outer surface of the inlet shoulder, the “T” head, and the exhaust pressure “Pj” acting on the discharge shoulder. These steam forces are shown in (b), and are equal to: Radial Inward Force “Fi” = Pa.[d.Db.(Kt - Ka)] + Pa.[d.Di.(ki - Kn)] + Pj.[D.Di . (Km - Ko)] +/- Segment Weight In addition, the radial inward force is supplemented by the force developed by the spring, which is relatively small compared to the steam force when the steam pressure is at design conditions. This inward force is also affected by the weight of the segment, which can be added for the top section subtracted for the bottom. In operation, these radial inward forces are opposed by the radial outward forces “Fo” produced by the steam in the individual pockets formed between the “N” seal strips. These total steam forces are shown in (c), and are equal to: 624 Seals, Glands, and Sealing Systems Pa Downwards Pressure (b) Pj Axial Length Ki Kt Ka Pa Ko High pressure side (a) Db Dh Low pressure side Km Kn Di Pj Ph Pg Pf Pe Pd Pc Pb Pressure Pressure Pj (e) Pj Pa (d) Dt Pa Stage Pressure Drop = Pa - Pj Upwards Pressure (c) Pj Axial Length Fig. 10.8.4—The steam pressure forces on the gland ring. In (b) is the radial inward steam force, and in (c) is the radial inward direction. These forces are directly opposite. In (d) is the downstream steam force, and opposing these are the axial upstream forces, shown in (e). Note: In determining the “inward “ and “outward” radial forces, it must be recognized that the projected inner surface of the gland segment is smaller than the outer because this surface is formed at a smaller radius. This may need to be taken into account on some lower pressure stages where the weight of the segment represents a larger proportion of the total forces. 625 Turbine Steam Path Troubleshooting and Repair—Volume Two In the axial direction, the steam forces in the downstream (direction of steam flow) direction. “Fd” is determined by the product of the steam pressure at inlet “Pa,” and the total area of the gland ring segments exposed to the steam. This total force is shown in (d), and is equal to: In the upstream direction, the steam force “Fu” is developed on the upper face of the “T” head by the pressure “Pa,” and on the lower shoulder vertical face by the steam pressure “Pj.” These forces are shown in (e), and are equal to: In the event there is steam leakage past the seal surfaces “S” and “H,” shown in Figure 10.8.3, then these values will modify, and while the pressure differential will be upset, it is unlikely the glands will fail to function effectively. However, there will be an additional leakage loss around the gland rings. In addition to consideration of the radial and axial forces produced on the gland segment by the steam pressure, it is necessary in certain applications to consider the turning moments these forces produce. Figure 10.8.5 shows these total forces, and it can be seen that there is a resultant moment, which would cause the gland ring to rotate about the point of contact “T.” Because of the curvature of the segments, the rotation of the segment about “T” would not be too great before the radial clearance “s” was consumed, but this would provide an additional leakage path past the seal surfaces, and if water were present could lead to “washing erosion.” This would also open the clearances between the seal strips and the rotor. 626 Seals, Glands, and Sealing Systems Inward force s T Axial force Pj Ph Pg Pf Pe Pd Pc Pb Pa Weight Outward force Axial force Spring force Fig. 10.8.5—The forces on the gland segment tending to rotate it about the turning point ‘T’. FORMS OF THE SEAL STRIP AND ITS TRIMMING The seal strip is formed in cross-section to achieve two concurrent objectives: • It must have sufficient thickness or depth “d” of Figure 10.9.1, to be able to withstand the stresses induced in it due to the pressure drop “dp,” which occurs across it • The seal strip will preferably have a knife edge or very thin edge formed on it, shown as “x” in Figure 10.9.1. This edge must be able to be machined onto the strip after its assembly without deforming the strip, or changing its axial or radial position. This knife-edge is to ensure that should rubbing 627 Turbine Steam Path Troubleshooting and Repair—Volume Two occur during operation, the strip will wear without causing significant damage either through mechanical deformation, or gouging of the rubbed components, or the generation of significant amounts of heat Seal strip Stationary Carrier Caulking Pressure drop 'dp' Thickness 'd' Steam flow direction x Fig. 10.9.1—The basic dimensions of the seal strip. When seals are inserted using caulking material to fasten them to the carrier, there are several considerations as to how they are arranged: 628 • The insertion so that the seal strip projection is firm against the downstream side of the carrier material as shown in Figure 10.9.1. This provides greater strength to the strip, which could be of excessive length if the caulking material were to form the base • The inserted knife-edges (shown in various figures) are arranged to be at 90 degrees to the seal surface • The taper face, if only one side is tapered, should be on the lower pressure side of the strip. This will produce a minimally smaller flow coefficient Seals, Glands, and Sealing Systems From manufacturing, assembly, and maintenance considerations, there are two types of seal strip that are used. There are those that are finished in the inserted condition, i.e., no further machining operations are required after assembly. Also, there are those requiring trimming after installation. When a strip that requires final trimming has been installed, particularly those at a large diameter, it is necessary to undertake the trimming operation using special tooling. This may include the use of a boring bar, or in the case of a rotor, placing the element in a lathe. Because seals can be either axial or radial, there are different methods used to form the final finished dimensions. The most appropriate method in any situation is a maintenance decision. h Casing x Cr s L Dc Ds Ca Fig. 10.9.2—A combined axial and radial seal above a rotating blade coverband. When the stage design contains a combined axial and radial seal of the type shown in Figure 10.9.2, the seal platform with an outer surface “s” has a radial depth that is required to provide stiffness to the coverband against bending forces. Therefore, the radial trimming of the rotor is undertaken before the seal strips are trimmed. However, these seal strips have a pressure drop across them, which if the 629 Turbine Steam Path Troubleshooting and Repair—Volume Two design requirements of “Dc” and “Ds” are not maintained within design specification, can make the length of the seal strip “h-Cr” to be excessive, causing an increase in its bending stress. Such an increase is normally not likely to exceed the strip capabilities, but this should be considered in any trim machining. INSERTION AND SECURING OF SEAL STRIPS There are various methods of attaching seal strips to stationary and rotating components of the unit. There is obviously a need to make a more secure attachment when the major carrying or locating component is rotating. This is because such strips will then be subjected to the effects of centrifugal load due to their own mass during operation. However, because of their design function there is a pressure drop across the seals, and therefore a bending stress induced in them. Possibly the simplest forms of attachment are those shown in Figures 10.10.1 and 10.10.2. Figure 10.10.1 shows a solid form of strip, in which there are small grooves, at a depth “L” produced on one or both sides of the strip. This strip, when inserted into a prepared groove in the major component, can be staked as shown. In the event these strips require changing, it is sometimes difficult to re-stake, because the major component has already been deformed by the initial staking action, and there could be insufficient material to reattach the strip securely. The caulked strip of Figure 10.10.2 is secured by inserting a strip, as shown, and then using a malleable material to form a tight enclosure, holding the strip securely in place. This type of attachment can be used successfully on both stationary and rotating components of the unit, and provides the capability of having the strips easily removed and changed. The prepared grooves are normally arranged 630 Seals, Glands, and Sealing Systems L Staked x Fig. 10.10.1—The staked seal strip. Stationary carrier Soft caulking material d L α x Fig. 10.10.2—Details of the caulked seal strip. 631 Turbine Steam Path Troubleshooting and Repair—Volume Two to have sloped or tapered walls at an angle “α” as shown. These strips can be trimmed to diameter “Ds,” plus clearance and knifeedged to “x” after installation. However, it often becomes necessary to undertake some straightening of the strips after caulking. This straightening is normally undertaken by hand; if rolls are available they are preferred. The ultimate success of the caulked strip depends upon the selection of a caulking material that can operate satisfactorily at the local environmental temperature of the stage. There are suitable materials available, and these make this means of securing the strip relatively easy, and secure. Locating grooves P Dt Fig. 10.10.3—The inserted and replaceable seal. Figure 10.10.3 shows a more sophisticated means of attaching a simple single, or multiple seal to a stationary component that is accessible from the horizontal joint. In this case, the seals are produced from solid material, and have the strips machined into them with form cutters. The location within the major component is achieved by means of a simple dovetail, as shown. The actual geometry of the strip is dependent upon manufacturer preferences, and can have considerable variation, and yet remain a successful type component. It is normally necessary to secure such strips against tan- 632 Seals, Glands, and Sealing Systems gential migration within its locating slot. These seals are normally fitted as segments that are of a length that allows their easy assembly. A typical seal system for a Curtis stage is shown in Figure 10.10.4, where a total of 11 inserted strips, of the type shown in Figure 10.10.2, are used to provide effective sealing between the nozzle plate, the stationary carrier, and two rotating blade rows. These seals are used for both axial and radial flow control, and are intended both to minimize leakage, and to guide the steam from one row to the next. Each of the 11 seals is located in a stationary component of the stage, and therefore is not subject to centrifugal loading. Ca2 Ca1 Cr1 Cr3 Cr2 Cr5 Cr4 Ca4 Ca3 Fig. 10.10.4—Details of the seal system around a ‘Curtis’ stage. The total system contains 11 strips, providing individual seals between the stationary and rotating blade rows. 633 Turbine Steam Path Troubleshooting and Repair—Volume Two When multiple series seals are required, the type shown in Figure 10.10.5 can be used. Here a number of individual gland rings are employed to provide a tortuous leakage path. A similar arrangement is shown in Figure 10.7.16 with alternate “hi-lo” teeth which can also be used. The inverted “T” type root that located the segment in the stationary component is spring loaded, and prevented from tangential migration by use of a securing screw at the horizontal joint. Fig. 10.10.5—Inserted gland rings seen from the horizontal joint. Tangential migration is prevented by means of a button screw at the horizontal joint. Gland and seal strip assembly There are several aspects of the assembly of seal strips and gland segments that need to be considered to help ensure the seals are as effective as possible, and to minimize leakage quantity. 634 Seals, Glands, and Sealing Systems Staked strips. This strip is inserted either in short segments, or as complete 180-degree strips. The length of strip depends upon the overall geometry. These are then held in position by staking. The staked strip, shown in Figure 10.10.6, shows the principal dimensions that should be controlled. The location within the major component is set by the depth of the groove “G” and its width “w.” These dimensions must be controlled to ensure the strip has the two staking slots located at the correct radial height, and the strip is not loose in the groove. The control dimension of the groove “V,” which is set from a referenced surface, sets the axial position of the seal. w 90° L X Dt Fig. 10.10.6—Dimensional details of the staked seal strip. After installation the strip is checked to ensure it is at 90 degrees to the plane of the surface. The strip is possibly trimmed to a length “L” to achieve the design seal diameter “Dt.” The knife-edge thickness is “x.” (There may need to be some small degree of flexibility in this value, and the shape of the strip required to achieve an acceptable compromise.) Many manufacturers do not trim such shaped strips, as they are able to control the manufacturing process of both 635 Turbine Steam Path Troubleshooting and Repair—Volume Two the major component, and the strip to the extent the design diameter “Dt” is achieved within design tolerances. This is a replaceable strip. However, after a number of replacements it would be advisable to re-machine the casing groove to a new width “w1,” and use a new form (width) of strip. This would provide new caulking material and new groove surfaces to locate the strip. Caulked strips. The row of caulked strips shown in Figure 10.10.7 are arranged to be at a pitch “P” apart, and after assembly the strips are adjusted to achieve the 90 degree projection from the major component. After assembly, these strips are trimmed to a length “L” to achieve the design seal diameter “Dt.” The strips are machined to an angle “α” at their tip, producing a knife-edge thickness “x.” The requirements for caulking are similar to those for the staked strips. The advantage of these strips is that it is normally possible to easily replace rubbed or damaged elements because the assembly process does not affect the major component. Like the staked seals, the axial, or radial position is set by the control position of the groove “V.” α 90° L Dt P V x Fig. 10.10.7—Dimensional details of the caulked seal strip. However, as a unit ages there is always a tendency for the integrity of the groove to deteriorate, particularly if the seals have been changed several times. Also, the general condition deteriorates 636 Seals, Glands, and Sealing Systems if there is water at the seal location. Under such conditions it is sometimes necessary to reform the groves with slightly larger dimensions, ensuring the same axial distance “V” is maintained. Inserted segments. The segments that are inserted into specially prepared dovetail locations have (in addition to the requirements of being correct dimensionally) the need to be assembled correctly so they can achieve both an acceptable seal, and so they can be removed for replacement as they become worn, and/or in certain designs to have inserted strips replaced. Irrespective of the type of tooth or seal strip (inserted or integral), there are two basic assemblies of these segmented glands. There are those designed to have a “spring” backing to them so they are able to move radially towards and away from their mating part as it moves during transient conditions. This “spring back off” is intended to minimize the wear that occurs on the segment as the seal rubs on the mating part. There are also solid segments, which have no spring action, and wear in the same manner as the inserted strip if a rub should occur. Figure 10.10.7 shows the cross section of a segmented gland with four straight through strips. This design has no provision for backing away under spring loading. Therefore, if a rub occurs the strips will be worn. The principal dimensions are shown. These are similar to the inserted strips, having a length “L,” being pitched “P” apart, and having a knife-edge thickness “x.” The seal diameter is “Dt.” Figure 10.10.8(a) shows a gland ring having four strips, but this gland section is designed for spring backing. The strips have a seal diameter “Dt.” Figure 10.10.8(b) shows a similar segment section, but in this design there are four “hi-lo” strips. Therefore, this section has two seal diameters “Dto” and “Dti,” with seal strip lengths “L1” and “L2.” With this (b) design, because of the possible geometry changes with the different seal diameters, and a constant pitching, it is possible there will be differences in the strip thickness “x” or the 637 Turbine Steam Path Troubleshooting and Repair—Volume Two angle of the strips. This is a design detail, and may vary from one manufacturer to another. The one factor that is different on these two segments is the distance “E” from the center of the shoulder to the steam face. This dimension “E” is made purposely different on the “hi-lo” design to prevent segments from being installed incorrectly. If incorrect assembly should occur, it is possible the high-low strips would be in an incorrect axial position, or in the case of straight through strips, there could be pitching difficulties. E E E1 E L L1 Dt Dto L2 X2 X P (a) P X1 Dti (b) Fig. 10.10.8—Details of the inserted gland rings with integral seal teeth. In (a) is shown the straight through form, used when the axial location has large differential expansion. In addition to these general requirements, for the spring loaded segments there are other requirements necessary to ensure they operate as intended by design: • 638 The production of a suitable steam seal face on the downstream side. Such a seal is required to minimize leakage around the segment. This steam seal is produced principally on the axial face, as shown in Figure 10.8.3 as surface “H.” There is also a sealing effect on the axial surface “S” Seals, Glands, and Sealing Systems Some designs ensure the steam seal face is always in relatively hard contact, by providing close-coiled, helical spring, as shown in Figure 10.10.9. These springs provide an axial force sufficient to produce contact at all times. Close coiled helical spring Fig. 10.10.9—A close coiled helical spring applying an axial load to close the steam seal face. This steam seal face is particularly important in stages with highpressure drops across the segment, and free moisture, as occurs in the high-pressure section of nuclear units. Under such circumstances, washing erosion could occur if leakage became excessive. It is important that when gland segments are changed, before installation of the new segments, all damage and dents are removed from the seal face within the diaphragm or packing head grooves. • Positive radially inward pressure is required to maintain axial alignment. This is done initially by a spring located between the housing and gland segment, and then by steam pressure during operation 639 Turbine Steam Path Troubleshooting and Repair—Volume Two For normal metallic seals, some form of leaf or close-coiled design is used. For carbon rings, a garter spring extending 360 degrees around the gland ring is used. • Because the segment will need to move radially during operation, there is a need for a small tangential “butt” gap between the segments. This gap represents a leakage path to the steam. However, without the gap, the segments could “bind” at some loads and steam conditions, holding the segments off their shoulders and preventing effective sealing. Also, the segments and carriers are manufactured from materials that can have different coefficients of expansion and will accept and reject heat at different rates. Therefore, the “butt” gap is required to prevent “binding” These gaps must be checked at final assembly. This is most important when a bronze material is used for the seals, because this material has a coefficient of expansion considerably different from steel, and failure to maintain the gap at design values could introduce problems during operation. • At “shutdown” there can be moisture collecting at various parts within the steam path. Often a provision is made at the bottom dead center of the gland housing for collected water to drain through the seal to some convenient collecting point within the turbine steam path, as shown in Figure 10.10.10. This seal has been in service for an extended period and also shows evidence of rubs The details of the components and seal system will vary from design to design. These variations represent a preference and the experience of the designer. When any maintenance work is undertaken, it is necessary to return the measured characteristics as close as possible to the original design, provide positive location, and a steam seal surface to prevent excess leakage. 640 Seals, Glands, and Sealing Systems Fig. 10.10.10—A caulked strip at a rotating blade row, showing the drainage slot and a ‘rub’ at the bottom dead center. SEAL STRIP AND GLAND RING MATERIALS Two quite distinct classes of material are used for the manufacture of gland segments and seal strips. The material selected for use in any particular application is chosen to be of sufficient strength and mechanical properties, such that it is able to perform adequately within the temperature of the local steam environment, and where necessary, be able to limit the damage that can be caused by moisture deposited from the steam. For the lower temperature, low-pressure applications, a nickel leaded copper alloy is normally used. While this material is normally produced as a sand casting, centrifugally cast elements are preferred. 641 Turbine Steam Path Troubleshooting and Repair—Volume Two These centrifugally cast rings tend to minimize the porosity, which is common in this material. Should porosity occur in the region of the thin strip, it can cause a component to be scrapped after a considerable amount of work and time has been invested into its production. In the high-temperature zones, steel is preferred, as it is better able to withstand the elevated temperatures and pressures to which these components are subjected. The transition from steel to the leaded bronze material usually occurs at temperatures below about 750ºF. Sand-cast materials A typical chemical composition for three copper base alloys is shown in Table 10.11.1. The first two alloys are suitable for sand casting, while the third is the normal composition for the centrifugally cast material. While gland ring segments are not exposed to any considerable stress levels, the basic mechanical properties are shown in Table 10.11.2 for reference. ASTM B584-73 Copper Tin Lead Zinc Antimony Nickel Sulphur Phosphorus Aluminum Manganese Silicon % % % % % % % % % % % 949 Cast 976 Cast Centrifugal Cast 79.0 - 81.0 4.0 - 6.0 4.0 - 6.0 0.30 0.25 4.0 - 6.0 0.80 0.05 0.005 0.10 0.005 63.0 - 67.0 3.5 - 4.5 3.5 - 5.0 1.50 0.25 19.0 - 21.5 0.80 0.08 0.005 1.00 0.15 61.0 - 68.0 1.3 - 3.0 4.5 - 7.0 0.85 10.0 - 16.0 0.35 Max 0.35 “ 0.35 “ 0.35 “ Table 10.11.1—Copper Base Alloys for Gland Segments and Seal Strips. 642 Seals, Glands, and Sealing Systems Tensile Stress psi Yield Stress psi Elongation % 38,000 15,000 15 40,000 17,000 10 34,500 13,500 30 Table 10.11.2—Mechanical Properties of Copper Base Alloys. Alloy steel materials The alloy steel chosen by the manufacturer for the higher temperature regions is again chosen to suit particular applications, such as nuclear and non-nuclear applications, and influenced by moisture content in the steam, etc. Typical chemical compositions are shown in Table 10.11.3. This listing is not complete. Material Carbon Manganese Phosphorus Sulphur Silicon Chromium Molybdenum % % % % % % % 1.25CR 2.25Cr AISI 416 AISI 410 0.12-0.15 0.12-0.15 0.03 0.03 0.03 0.03 1.25 1.00 2.25 1.00 0.15 0.15 1.25Max. 1.0Max 0.06 0.04 0.15 0.03 1.0Max 1.0Max 12.0-4.0 11.5-13.5 0.60Max Table 10.11.3—Typical Alloy Steels for Gland Segment and Seal Strips. GLAND SYSTEM OPERATING PROBLEMS Problems occur and damage is found at various locations within the steam path where seals are used. These problems cause damage to either the seal strip itself, which is the more common, or it can be on the component against which the seal is formed. 643 Turbine Steam Path Troubleshooting and Repair—Volume Two The most common fault that occurs on the seal strip is for “rubs” to occur during operation. These rubs flatten the knife-edge, thus causing both an opening of the running clearance, and also increasing the discharge coefficient. A rubbed knife-edge is shown in Figure 10.12.1, where a rub has occurred on the caulked strips in the diaphragm carrier, located above the blade tip. This piece of strip is located at the “bottom dead center,” and the moisture drainage slot can also be seen. This seal is located above a blade coverband in an impulse unit. This rub is not heavy, and a decision to change will be based on an economic evaluation of the additional losses sustained as a consequence of the increased clearance. Fig. 10.12.1—A caulked seal showing a ‘rub’ at the knife edge. Figure 10.12.2(a) shows a gland segment from a diaphragm where heavy rubs have occurred. In Figure 10.12.2(b) are the measured values of clearance around the total seal. These clearances have been determined from knowledge of the design clearance, the original seal strip height “h,” and the measured value of “h” from the removed gland segments. These measurements, which were taken at three loca644 Seals, Glands, and Sealing Systems Fig. 10.12.2(a)—A lower gland ring segment, showing rubs at the knife edge. A diametral record of this complete ring is shown in Figure 10.12.2(b). tions on each of the six segments, show that the major wear is in the vertical position. This unit has obviously been either subjected to high levels of rotor vibration, or was not aligned to its best advantage when returned to service after the previous outage. When such rubs occur, there are three possible actions available to the operator: • To operate with the damaged seals and accept the additional losses that are induced by the increased clearance produced • To replace the worn strips with new components, then adjust them by some means to re-establish the original clearance. This may involve the expense of removing the seal carrier to a facility with the capability of machining the seal edge, or using a boring machine inside the unit to produce the seal diameter 645 Turbine Steam Path Troubleshooting and Repair—Volume Two • To hand dress the worn parts, to re-establish the knife-edge. It will normally be necessary to return to service with an increased clearance, but the coefficient of discharge will have been re-established at or near its original value. This can normally only be justified when the damage occurs over a relatively small length of the knife-edge The choice depends upon various factors, most importantly the economics of the situation, but also by the ability of the design to accommodate the installation of new parts, or the dressing of the existing. Many seal arrangements do not easily lend themselves to replacement without incurring excessive cost, and one that is greater than any savings anticipated as a result of fuel cost savings. Fig. 10.12.2(b)—A diametral record of the radial clearances around a complete gland ring set. These clearances have been measured at three tangential positions on each of the six segments. The design clearance was set at 0.015". 646 Seals, Glands, and Sealing Systems Fig. 10.12.3—A multi knife edge seal showing the effects of a ‘rub’. These seals are produced integral with the gland ring carrier. The forms of some strips make refurbishment difficult, and the only recourse is replacement. Figure 10.12.3 shows the complex seal form from an impulse unit, where each pitch had an original three knife-edges. After rubs, these edges have been destroyed, and badly deformed. However, these gland rings can be replaced with relative ease. After extended operation at high-steam conditions, some seal strip material tends to become brittle. This is particularly the case if there have been rubs that would have heated the material, and then its having been immediately quenched by the flowing steam. Such quenching will increase the tendency towards embrittlement. Under these conditions, relatively light impacts from mechanical debris (or 647 Turbine Steam Path Troubleshooting and Repair—Volume Two even the steam forces) can cause the thin seal metal section to rupture and deteriorate. Figure 10.12.4 shows a seal where a significant amount of material has been lost, increasing the leakage area. This shows that there have been rubs on these spring-loaded elements. Fig. 10.12.4—A seal showing the loss of seal strip material. This seal has been exposed to high temperatures for extensive periods. Another form of damage sustained by seals is “knock” damage, which is sustained as a consequence of small impacts when the unit is open for inspection or repair. For this reason, operators must take the greatest care to protect these seals when they are removed from, or left exposed in the unit during an outage. It is sometimes possible to straighten bent or deformed seal strips. However, it must be recognized that microcracks may then exist, which could cause rupture after return to service. The missiles generated by subsequent failure of the seals are unlikely to cause impact damage, but there will be a resulting loss in efficiency and output 648 Seals, Glands, and Sealing Systems Fig. 10.12.5—Rubs on the inner coverband of stationary blades. These rubs are not normally damaging to the coverband, but will deform the knife edge of seal strips. The casing carrying these stationary blade elements should be checked for concentricity. The surface against which a seal strip forms its seal can also be damaged, by various mechanisms. Figure 10.12.5 shows a portion of a coverband in which small grooves have been cut by rubs that have occurred during operation. This form of damage is common and can also occur on the rotor body. Often, such rubs will cause localized heating and consequently hardening, making the coverband and/or rotor materials more susceptible to corrosive attack from aggressive ions that have found access to the steam path. When seal strips are caulked into a stationary component, they must be checked to ensure they are secure against the steam forces that will act on and affect them during operation. Figure 10.12.6 shows a portion of a seal system located in a casing above a blade coverband. Here the seal strips have worked loose, detached, and passed down the steam path. Figure 10.12.7 shows a caulked seal strip from above a rotating blade row. This strip has been damaged by mechanical impact, but can be replaced relatively easily at a maintenance outage. 649 Turbine Steam Path Troubleshooting and Repair—Volume Two Fig. 10.12.6—Seal strips, which have detached from their holder grooves, have entered the steam path, causing some gouging and damage to the blade elements. 650 Seals, Glands, and Sealing Systems REFERENCES 1. Martin, H.M. Steam Turbines, published in The Engineer, London 1913, 1,610 2. Egli, A. The Leakage of Steam Through Labyrinth Seals, Transactions ASME, Paper FSP-57-5 3. Kearton, W.J. Leakage of Air through Labyrinth Glands of Staggered Type, Institute of Mechanical Engineers, September, 1950 4. Meyer, C.A., and J.A. Lowrie. The Leakage Thru Straight and Slant Labyrinth and Honeycomb Seals, ASME Paper 74WA/PTC-2 5. Neuman, K., G. Stannowski, and H. Termuehlen. Thirty Year Experience with Integrally Shrouded Blades, The Joint Power Generation Conference, Dallas, Texas, October, 1989 6. Cofer, J.I., S. Koenders, and W.J. Sumner. Advances in Steam Path Technology, GE Power Generation Paper GER-3713C 651 Chapter 11 Quality Assurance for Replacement and Refurbished Steam Turbine Components INTRODUCTION There are various causes that can initiate mechanical failure of the steam path components (see chapter 1). A major, but often-overlooked contributor, is the quality of the manufacture and/or assembly, and their final compliance with design requirements. There are many critical components, each with a number of characteristics that have the potential to affect both efficiency and reliability within the unit. The quality requirements of these components are established by the designer in terms of the tolerances applied to the individual components, the processes that will be used to produce them, and the manner and expertise with which 653 Turbine Steam Path Maintenance and Repair—Volume Two they are applied and assembled. Many of these technical requirements are not obvious from casual observation, and to ensure the design specifications are achieved in the delivered product, it is necessary to investigate these during the process of awarding a contract, and then to monitor their application during manufacture, and also to have access to any instances of nonconformance. This availability of access to (and even involvement in) the disposition of nonconforming situations provides considerably greater confidence in the quality of the final product. Such participation also allows the purchaser to be able to anticipate, and identify manufacture as the cause of any specific mechanical problems subsequent to going into service. The purchaser/user has an implied responsibility to monitor the manufacture and assembly of these components. This monitoring does not require the physical measurement of the components themselves, but can normally be achieved by the monitoring of the supplier’s quality program, and also by directing inspection or surveillance attention to those critical characteristics that must be achieved if the unit is to perform as anticipated. To undertake a “quality review and monitor” it is necessary to compare the manufactured components against the standards defined by the design engineering function. These engineering standards establish component requirements, and the quality of the products is dependent upon the ability of the manufacturing department to achieve an acceptable level of compliance. This synopsis is not complete, as the requirements from one manufacturer to another may be defined by different methods. This is not unreasonable, and these differences reflect the differences in both the manufacturers’ method of manufacture and their experience. It is only by seeing that the manufacturer has (and follows) their own documented standards, that an acceptable component can, from a manufacturing/refurbishment perspective, be achieved. 654 Quality Assurance for Replacement & Refurbished Steam Turbine Components The design specified requirements for components of the steam path are stringent. The engineering specifications include definition of material properties, physical dimensions, process, and performance requirements. Many of these engineering defined requirements will have an adverse effect on the total performance of the turbine if not followed in detail. Component failures resulting in forced or extended outages are common within the steam path, and often the cause of such failures can be attributed to failure on the part of manufacturers to adequately control either the metal forming and manufacturing process, or the assembly of the individual components into the final product. An important aspect for any purchaser of new or replacement parts, is the ability to establish that components are produced in a way that they will assemble to the turbine, and provide performance levels that comply with system requirements. This surveillance, or monitoring of product quality should be an integral part of any procurement program, and could have a direct effect on the quality and suitability of the parts for their intended application. The manufacturer of new and replacement parts will normally undertake inspection and produce records to verify compliance with design requirements. It is important to recognize the requirements of a quality program, and in qualifying a manufacturer determine that their quality program will help ensure components meet the design specification. In addition to design, manufacture, and assembly, the program can also be extended to include the requirements that the components are packed and shipped so they will be in an acceptable condition when mounted to the turbine. In the case of inventory components, the time of installation may not be for several years, and therefore will require to be protected to prevent any deterioration during the storage period. There is, of course, a cost associated with the review and monitoring of a quality program—a real cost. However, if this prevents an 655 Turbine Steam Path Maintenance and Repair—Volume Two incorrect component being delivered, or the extension of an outage so that some defect can be corrected, then the costs are easily justified. Figure 11.1.1 shows the possible involvement of a purchasing engineer, in reviewing and establishing a quality program at the outset of a project. The most appropriate (but costly) approach would be to review the potential supplier Quality Assurance (QA) program prior to placement of the contract, but this can become expensive in terms of a small contract, and would be difficult to justify. Therefore, a normal manner of undertaking this review is after placement of the contract, and at an engineering review, which is held after the supplier engineering function has prepared the engineering specification for the component. At that time, the drawings, material specifications, special process specifications, and any other details (togeth- Engineering Definition Prepare Engineering Specification Review QA Program * Manual Review * Program Implementation Engineering Review * * * * * Original Drawings Reverse Engineering Material Specifications Process Specifications Special Processes * Inspection Records Inspection and Test Plan Product Surveillance Quality Records Fig. 11.1.1—Purchaser’s involvement in establishing project scope and QA involvement. 656 Quality Assurance for Replacement & Refurbished Steam Turbine Components er with the QA program) can be examined for suitability, the contract being set on the basis of acceptability. At the review meeting, the inspection and test plan (I&TP) are reviewed and accepted, or modified, and “hold points” agreed upon. Based on this I&TP, the purchaser will define and prepare his or her surveillance activities. RESPONSIBILITY FOR QUALITY There are various factors that can contribute to poorer-thananticipated reliability of the turbine steam path, and in certain circumstances these factors will combine to degrade the total availability of the unit. To begin to define the technical requirements of any component to operate adequately, it must be recognized that the prime responsibility for the adequacy of the component is vested in the design function. Engineering must ultimately be considered responsible for product quality. Steam path components, many of which are technically complex, are designed and manufactured to stringent standards. Such standards are intended and expected to make the turbines suitable for continued, reliable operation. The stresses induced in, and the environmental operating demands placed on steam path components are such that their margin of safety can be seriously eroded by relatively small manufacturing deviations from the design specified requirements. Many failures of components within the steam path can be attributed directly to incorrect or inadequate control of manufacture, while others are accelerated or contributed to by unacceptable manufacturing quality. Turbine steam path components operate in an environment that tends to degrade their material quality. Also, many components can be subject to excessive stress levels, both direct and alternating, as a 657 Turbine Steam Path Maintenance and Repair—Volume Two consequence of relatively minor manufacturing deviations. These factors combine to reduce the operating life of the components in a number of ways. The cost of a turbine component failure is composed of several elements. There is the cost of purchasing replacement components, the cost of opening the turbines to install them, and the maintenance staff costs associated with their installation. However, these costs are normally minimal when compared to those costs associated with the purchase of replacement power, whether it is purchased from other producers, or generated within the utility by starting older, less-efficient units, or starting units consuming more expensive fuel. For these reasons the purchaser has a direct responsibility to monitor supplier quality to help ensure the possibility of manufacturing errors is minimized. DEFINITION OF QUALITY The quality of any piece of engineered equipment can be defined by its degree of compliance in meeting primary and secondary objectives, as defined by design engineering, and by providing the owner an acceptable return on investment. To define the quality of turbine components, it is necessary to be able to state quality requirements in quantifiable terms, and then be able to measure these against a standard or target value that provides direct comparison with anticipated values. In general, the quality of a component should be defined in terms of its performance against both supplier-predicted levels, and standards for the industry. Establishing what the standard value should realistically be for any technically complex, contract-engineered item is a problem of considerable magnitude. In making 658 Quality Assurance for Replacement & Refurbished Steam Turbine Components comparisons, it is very unfair to attempt a comparison of a component or design that either contains a significant number of newlydeveloped features, or will be operating in an environment or load pattern in which it is untried, to one containing no prototype components and operating in a proven and accepted manner. Therefore, it is reasonable to expect the quality or performance level of any component to be a comparison with a standard that represents a realistic assessment for the design, its degree of design evolution, and mode of operation. As an initial step towards measuring quality, it becomes necessary to define, and then quantify performance, and relate quality to those factors that define it. The quality of any turbine component can be established against standards of performance, and against supplier-predicted levels. It would be appropriate if this could be done in terms of their influence on both efficiency and reliability of the unit, and the effect such components have on these qualities. Optimally, each component should be considered in isolation, but if comparisons are to be made among different but comparable components, it is necessary to reduce the parameter of each piece to a common base to allow meaningful comparisons. It is difficult to establish the efficiency of a particular component. It is sufficient to assume that if components meet their engineering definition (from the original equipment manufacturer, or from another supplier determined by reverse engineering), then it will achieve the efficiency and reliability required to achieve the heat rate and availability of the turbine-generator unit. 659 Turbine Steam Path Maintenance and Repair—Volume Two DEFINITIONS OF PERFORMANCE In describing performance or quality levels, it is valuable to establish or define terms that can be used, and have specific meanings. The most appropriate are: • Guaranteed level—quoted by a supplier during the bidding phase. Such levels could form part of a commercial agreement between supplier and purchaser. If tests or measurements are taken, it is normally against these levels that comparisons will be made. Efficiency can be measured. However, availability is difficult to quantify because it is influenced both by manufacturer and operator determined factors • Target (tolerance) level—represents an adjustment made to the guarantee level, up or down, for experience gained during previous operation with the supplier’s equipment. Because these values are normally adjusted by the purchaser (based on his or her accumulated knowledge of the manufacturer) they form no part of the bid, and have no legal status. These target values are, however, of great importance both in bid evaluation, and for operators who must measure their performance to some established and predictable standard • Actual level—represents the field measured and observed values and can be compared to both guaranteed and target values. Discrepancies, either positive or negative, should be accounted for These first two definitions are often used interchangeably as the predicted level of performance. Most often, the predicted value refers to the guaranteed level when applied to unit efficiency, but more often, the target levels, when applied to unit availability. (The supplier does not normally guarantee availability.) However, most 660 Quality Assurance for Replacement & Refurbished Steam Turbine Components suppliers have statistical values available upon which a purchaser can make predictions of how his or her unit should perform in the environment of his or her power system. The factors of performance In discussing quality or performance, it is necessary to establish what factors contribute to it and how they can realistically be measured. These factors are shown in Figure 11.2.1. It is necessary to define performance in a quantifiable manner, establish a means of comparing it to a base that is common for all components, and allows comparison in such a manner the anticipated differences between them can be eliminated. Steam Turbine Performance Unit Availability Unit Efficiency Reliability Initial Efficiency Sustained Efficiency Safety Mechanical Integrity Maintainability Repairability Accessibility Interchangeability Correctability Fig. 11.2.1—The factors of performance. Performance can be considered as being comprised of two distinct factors—efficiency and availability. Of these two factors, availability comprises two sub factors—reliability, which measures the 661 Turbine Steam Path Maintenance and Repair—Volume Two unavailability or forced outage rate of the unit as a consequence of component problems, and maintainability, which measures the ability or time interval required to return a unit to the available status after a forced outage caused by various components. The influences of availability. The availability of power-generating equipment or components contained within such equipment is a measure of its availability to generate power when required. This definition is independent of the equipment actually being in service; the requirement is that it is available if required. There are two factors of availability that need to be considered in describing the performance or quality level of any unit or component: • Reliability—a measure of the forced outage rate of a unit or component, and is a measure of the period during which, due to component failure or damage, the unit is not available to generate power • Maintainability—a measure of the ease of access to the various component parts of a unit to undertake corrective action, and the individual components’ ability to be reliably repaired. Therefore, this is a measure of the suitability of a unit to be inspected, corrected, and returned to an available status after a planned or forced outage has required some corrective action The influences of efficiency. The efficiency of the thermal cycle, the prime mover, or any significant component of the prime mover, can (for a cost) be determined by established means, and with relative ease. This value can also normally be determined with a high degree of accuracy. • 662 Initial efficiency—It is normal for the builder of power generating equipment to guarantee efficiency for each unit supplied, or in the event of supplying replacement components, that they will perform at the same level as the original components they Quality Assurance for Replacement & Refurbished Steam Turbine Components are replacing. This guarantee level is one predicted by the supplier from an analysis of past operating information, or from development tests undertaken on a new design of components • Sustained efficiency—Degradation of efficiency will occur as many of the components age. This degradation can be due either to deposits formed on them, or any mechanical damage they sustain as a consequence of operation. Any evaluation of the efficiency of the replaced component must be undertaken before degradation to a significant level occurs THE DESIGN SPECIFICATION The turbine steam path utilizes components selected and arranged by design to achieve a predicted level of output, and one that can achieve a level of efficiency consistent with an acceptable level of service reliability. These two requirements of efficiency and reliability define the “performance level” of the turbine. Unfortunately, in many instances of component selection, optimizing these two requirements is difficult because they are often in direct conflict. Therefore, a major responsibility of design is to evaluate alternative components and make a balanced selection, i.e., a selection that will achieve acceptable levels of both efficiency and reliability, and can be utilized in a turbine that can be manufactured at acceptable costs. In evaluating and making the selection, the designer considers the total requirements of the various components, both individually and as an assembled whole. The final component selection is made from a detailed analysis of possible arrangements. After a total evaluation, the turbine is defined, and a specification is prepared that will include information to the manufacturing function of the turbine 663 Turbine Steam Path Maintenance and Repair—Volume Two supplier. This design definition will include at least a portion of the following information for each component: A material specification. The material specification defines the properties of the material from which the components are to be produced. This material specification should include both chemical and mechanical requirements, the method of material production, heat treatment, and the delivered (and possibly the final) microstructure. A material supplier will supply a material test certificate along with the provision of material. The turbine manufacturer will examine the material supplier’s certification as the material is received. It is also possible critical materials will be subjected to microstructure and mechanical properties verification and chemical composition analysis before being used to produce components. The most effective method for a purchaser to ensure materials meet design requirements is to examine the material specification and then compare the material test certified properties with the specification for conformance. Material conformance must be complete in every detail, from physical property and chemical compliance, and heat treatment, to method of manufacture and preparation. Physical dimensions. Dimensional requirements are more difficult to define, and these data are provided to manufacturing departments in a variety of forms. The most obvious form is for dimensions and tolerances to be specified on manufacturing drawings. However, other methods, such as operation sheets and manufacturing standards, are also used. If material is to be left on a component for finishing in the field, the specification should define how much additional stock should be left on the component. Possibly the most effective method of considering the dimensional requirements of the steam path components, from a purchaser’s perspective, is to review them briefly by component, and pro- 664 Quality Assurance for Replacement & Refurbished Steam Turbine Components vide details of the controlling dimensions that must be achieved in the finished product to help ensure a reliable component. Definition of surface finish. The surface finish can be critical on certain components, and where necessary is defined. As part of the engineering requirement, this finish is gauged to ensure the surface finish is correct and has its “lay” direction with the correct orientation. Specifications for special processes. Any engineering specification can include the use of special processes, i.e., processes that cannot be gauged at their completion without destructive examination. Such process specifications will include a definition of the requirements, and a means to calibrate the procedures for their undertaking. Procedures for non-destructive tests. Many components will require special tests to ensure they meet engineering-defined requirements. Such specifications may not contain any special tests, but may define closer tolerances, or the use of specific measuring devices. Such a test procedure will also define the tolerances within which the test results must fall. Special instruction for the assembly and alignment. There may be special assembly sequences and test requirements for main and sub assemblies. The engineering definition could define these requirements, test procedures, and acceptable results. If after assembly an alignment test or checks are to be undertaken to ensure the suitability of the assembly, these will also be identified and defined. This information is supplied from the design function in the form of drawings, specifications, and written procedures. It is the responsibility of the engineering function to control product quality through the quality assurance/inspection department. A major responsibility of this control is for engineering to be involved in the evaluation and disposition of any nonconforming conditions that are reported to them for evaluation. 665 Turbine Steam Path Maintenance and Repair—Volume Two REVERSE ENGINEERING The process of reverse engineering provides an engineering specification for turbine components based on an analysis of components that have been in service, and found to be in a condition such that they must be replaced. This process will allow components to be manufactured by suppliers other than the original equipment manufacturer (OEM). This reversal process is applied either to obtain parts at competitive prices, or to achieve a delivery that will allow a unit to be returned to service in a much faster time. However, acquiring enough data to completely define and specify the part can introduce problems related to operation, if sufficient caution is not exercised during the reversal process, and the operating requirement of the components not adequately considered. A component produced as the result of reversal must meet the critical, dimensional, and material properties of the original component as closely as possible. The component may also require certain proving tests to ensure these critical characteristics, essential to satisfactory operation of the unit, are present in the replacement parts. Reverse engineering can be (and is now) applied to many mechanical components. It is necessary to consider the application of the process of reverse engineering of the more complex components of the steam turbine (although the process as described applies to most mechanical components in all equipment). This section will consider those factors that must be reviewed for these complex components to make them suited to the local environmental conditions. The components can be subject to high stress levels, and can possibly be exposed to severe transients in terms of load, heating, and cooling rates. The environment within the turbine can also contain certain aggressive ions, which can concentrate and become active under certain conditions. 666 Quality Assurance for Replacement & Refurbished Steam Turbine Components The reverse engineering process can begin from one of two bases, in terms of the components that are used to determine the critical characteristics providing its technical definition. These are: • Samples that have been in use are showing damage or deterioration, and are in need of replacement. These samples may require some engineering interpretation to obtain complete definition • Inventory spares, in which case the samples are in a new condition, and provide easy definition for the reversal process. Such components are normally required to replenish the stock of spares before committing the last spares In either case, the reversal process requires modeling from the sample components. If the component or sample that is being copied has been in service, its replacement is generally necessary because its condition has deteriorated. This deterioration will represent either a modification or a change of the physical dimensions of the part, or a deterioration of its metallurgical properties. Therefore, the component will not be in a condition that conforms to the original design requirements, and the form and degree of deterioration will have to be identified and accounted for in the replacement parts. Evaluation and interpretation of the available information from its gauging is necessary. Usually, there is enough material and dimensional detail of the original component remaining that a judgment can be made. If such data are not available, then it is often impossible to undertake reverse engineering. It is usually preferable in such a case of modeling from used samples, to have more than one sample or specimen available. Because of the possibility of severe wear, the engineer responsible for the reversal process is required to have some product knowledge of the components. 667 Turbine Steam Path Maintenance and Repair—Volume Two If the component to be reverse engineered is an inventory spare, the problem of obtaining information of the original design is made much easier, because there should have been no or at least a minimal amount of deterioration while in inventory. However, before the reverse engineering process is begun, it is better to ensure the component does in fact fit the requirements for which it was originally produced. Also, if more than one sample is available, the dimensional and mechanical properties should be established from more than the one sample. A component that is specified by reverse engineering should have properties that are equal to or, because of advanced or improved technologies, superior to the original. A reverse engineered component acts as a direct substitute for the original, in terms of function and quality. It will have a life expectancy equivalent to the original, and will not represent an inferior product whose performance will compromise the efficiency or availability of the unit. Economies in the purchase of replacement parts can be achieved by reverse engineering and competitive bidding. However, such economic advantages cannot be achieved if the reversal process is not undertaken with due regards to the design dimensional and material needs. These considerations should also address both short and long term operating requirements. In addition, certain components may require specialized processes, or manufacturing facilities that may not be readily available. For small quantities it would be difficult to justify the installation of such plant. Each case must be evaluated on its own merit. The concept To successfully reverse engineer any component it is necessary to obtain certain information concerning it. This helps ensure its critical characteristics can be reproduced in the components that are to be provided to the owner. To undertake this reversal process, it is 668 Quality Assurance for Replacement & Refurbished Steam Turbine Components necessary to define the technical requirements of the component based on its operating history, knowledge of the type of component and the type of service it will, or has, experienced. The principal considerations of reverse engineering are: • What are the critical characteristics and functions of the component? • What caused the failure in the original component, and is the situation or set of circumstances likely to re-occur? • Was the failed component a previously reverse-engineered component? If so, are there any differences in the repeat failure and the need to replace the two original and reversed components? • Which dimensions are critical, and should be measured and applied to the components? • What level of tolerances should be applied to both the critical and other dimensions? • What surface finish requirements should be specified for the various parts of the component? • What are the maximum and minimum clearances between the component parts and other parts within the steam path? • What materials should the components be produced from, or is there a need to re-evaluate the materials and possibly make changes? • Are there any special heat treatment or other special processes required to provide additional mechanical properties to the material? • Should coatings be applied to the components to reduce wear and damage, and what advantages will this offer? 669 Turbine Steam Path Maintenance and Repair—Volume Two • What functional tests should be applied to the components to help ensure they will adequately fulfill their purpose during operation? Advanced planning To achieve the maximum gain from the reversal process, in terms of costs and delivery, some advanced planning is required. Such planning permits engineering to be undertaken and manufacturing space to be made available within the supplier’s facilities. In the majority of instances, neither the supplier nor the purchaser will have drawings and manufacturing specifications available. In this case an investigation is often of value to investigate if “sister units” were ever built by the OEM. Several options are available to determine if such units exist: • Inquire with your own utility operations and maintenance department • In the United States, the Fossil Operations and Maintenance Information System (FOMIS) [operated by the NUS Corporation and the Electric Power Research Institute (EPRI)] often are able to supply such information • The manufacturers who are to produce the reverse engineered component, or those who are bidding on its manufacture, may also have information on the component • The OEM may also make such information available, either once a sister unit has been located, or to facilitate manufacture and allow a customer to return to service quickly. This is usually the situation when the OEM cannot support emergency manufacture In the event a sister unit (or a unit with similar or identical components) is located, the operation and maintenance department of 670 Quality Assurance for Replacement & Refurbished Steam Turbine Components the owner of the sister unit should be contacted, and possibly a visit made to the owner to discuss the problem and the needs. It is possible the OEM has a design flaw in the original component on other units and has failed to inform the owners of sister units. Such discussions with other owners can be beneficial to the reverse engineering process. Dimensional requirements When a component is examined, sufficient dimensions must be recorded to define its mechanical extremities and production requirements. These dimensions and tolerances are established by both direct and computer measurement. Also, the use of optical comparators with at least a 10X magnification allows direct comparison. Dimensions must be sufficient to allow the component to be drawn and dimensionally defined. It is necessary to define on the manufacturing drawings, normally from product knowledge, which dimensions are critical, and which control (or establish) the quality and suitability level of the component. Remember that each component will interact with (and possibly be assembled to) others. This will require that the reverse engineered component does not cause either interference or looseness beyond what can be tolerated by these other existing components. If a component has been in operation for extended periods, possibly at high temperatures, it is possible this environment could have modified the dimensions that are measured. Such operation could also have caused deterioration, which precludes the application of the reversal process. This is particularly important when creep stresses could have been a factor causing undue deformation and distortion. Under these circumstances, it may be necessary to make measurements of other components that are unaffected or influenced by the component being reversed, and with which it will interact in its future operation. 671 Turbine Steam Path Maintenance and Repair—Volume Two Many components require a level of dimensional accuracy that is established by the detailing close tolerances. This requirement normally cannot be determined from an examination of one component, although it can be inferred by an examination of a number of components. (This requirement requires some detailed knowledge of the design requirements of the component, its application and operating environment.) Tolerances are applied to any component to ensure it can be interchanged, its stress and efficiency levels are acceptable, and it will meet overall criteria of engineering performance. Tolerances that are too demanding can cause the manufacturing costs to be excessive, and provide marginal or no improvement in the performance of the component. Therefore, it requires mature judgment as to the level of tolerances that are applied to each dimension on any component that is reverse engineered. Surface finishes, like the range of tolerances, must be chosen and applied to allow the component to fulfill its design function without causing manufacturing costs to exceed a level that makes them unattractive. There are some applications where the direction of finish is also important. This again should be followed, and processes that remove material in a different manner cannot be substituted. The total reversed component represents a complex engineering evaluation and re-specification. The final manufacturing specification should, therefore, provide a clear definition of dimensional requirements, provide tolerances within which the components will be acceptable, and the surface requirements at each location where the performance level can be achieved. Material requirements Irrespective of its level of dimensional compliance, a component must also be produced from a material that is compatible with both 672 Quality Assurance for Replacement & Refurbished Steam Turbine Components its operating environment, and the levels of stress to which it will be subjected during operation. The original equipment designer selects the materials of original components to achieve certain characteristics in operation, and to make them suitable for extended life operation. It should also be recognized that some components, such as gland rings and studs, are of finite life and could require periodic replacement. However, this should not be a common practice in blades, and other major components that should be manufactured to be suitable for the life of the unit. Because many components will often be exposed to an aggressive chemical environment, it is also necessary to ensure the application of the component is considered in both determining and specifying the material of the replacement and reversed component. The chemical properties of a material can be determined by spectrographic and chemical analysis. The mechanical properties of the component can be determined by suitable destructive and/or nondestructive examination. In determining the mechanical properties of the material, the location of the removed test specimen must be chosen with care. This is to ensure they are not from a location where there could have been significant deterioration due to operation. It is normally not sufficient for a critical component, which is known to be subject to high operating stresses, to rely upon a single test specimen. In addition, there are some components, such as rotors and “shrunk on wheels” where it is necessary to determine multi-directional mechanical requirements. Material substitution There are, however, occasions when it will be necessary and/or prudent to consider the use of an alternate material for the component being reverse engineered. Generally it will be possible to upgrade the material. This is particularly the case when the component being reversed is of an old design, and it is possible to take advantage of 673 Turbine Steam Path Maintenance and Repair—Volume Two advances in material technology. It is also possible to use an alternative material when the component is being reverse engineered because of original material deficiencies, or poor selection. Such situations could arise where it has been found that the original material of the component has suffered from some form of deterioration that could be eliminated or slowed by using a different material. Within the steam turbine industry, there are often materials having chemical compositions and mechanical properties not present in commercially available materials. Under these circumstances, it may not be possible to manufacture the component from materials with identical properties. Fortunately, there are often other materials available that will serve as well, although it might mean upgrading the material to achieve suitability. With modern material production techniques, it is often possible to use materials, although not conforming to the analysis of the original component, which can provide adequate service. When a change of material is anticipated, it is recommended the important design parameters, or requirements of the component be reviewed, and these compared with known operating requirements and restrictions. The OEMs often use materials with properties that are less common than commercially available products for one of several reasons. The materials that the OEM uses were developed for a specific application, and because of the volume of material required, the OEM can afford the purchase of a special melt or mill run. In addition, it is possible the OEM, because of their volume of production, would have designed special tools or processes that do not affect the properties or quality of the final product, but do facilitate the manufacturing process. Material substitution is often a very real consideration and even a necessity in reverse engineering. Such change may be required through both the desire to improve the quality of the product, and 674 Quality Assurance for Replacement & Refurbished Steam Turbine Components also because it may not always be possible to duplicate the component material properties. The following rules concerning the application and substitution of materials should be observed: • No material change should be made when an examination of the component to be reversed indicates there is no justification for it, even if the change is possible • If a material cannot be replicated, an upgrade should be made using a material with the most compatible chemical and mechanical properties • When a change of material is required, and a change in blade mechanical characteristics has occurred, the replacement material must be examined for its short and long-term mechanical properties, and its ability to operate in the chemical and physical environment of the component it will be replacing If these three criteria are observed, it is unlikely a change of material will cause any significant deterioration in the performance level of the reverse engineered component. However, each substitution must be carefully evaluated and made only after it is determined it is safe to do so. There are instances when superior materials can be used. However, it cannot be assumed that the use of a material of higher mechanical properties will improve the component. This substitution could, in certain instances, place the component into a situation where it is subject to other forms of deterioration. Special processes Many components or sub-assemblies of the steam turbine require the application of special processes to complete their manufacture. These processes include welding, brazing, nitriding, shot 675 Turbine Steam Path Maintenance and Repair—Volume Two peening, coating, and a variety of others that join, coat, harden, and generally modify the components and their characteristics to make them suited for application within the turbine. If a component must be manufactured by reverse engineering, it is necessary that the reversal process should identify and specify these requirements to ensure the replacement components are entirely suited to their intended application. Welding is possibly the most commonly used form of the special processes; it is also the process most likely to cause the reverse engineered component to be unsuited to its intended duty if the welding is not performed correctly. Therefore, the reverse engineering process must establish parameters that will allow those components needing welding to be completed in a manner that will not compromise their quality. The selection of the welding process will depend upon the material of the component that will be involved, and the type of duty anticipated. Other special processes, which are essential to the satisfactory performance and continued operation of the component, will also require careful specification as part of the entire engineering specification of the reverse engineered component. Component evaluation and testing To ensure the adequacy of many components, it will be necessary to undertake some forms of nondestructive testing or examination, and in some instances to undertake destructive testing and examination to ensure the components conform to the requirements of the design. The reversal process specifies the requirements of these examinations and tests, and defines the limits within which these components are considered acceptable. Non-destructive examination must be so defined; the component will achieve a level of acceptance that is consistent with that of the original specimen. This phase of the reversal process is important, and provides a level of assurance that the component produced by reversal will fulfill its intended duty. 676 Quality Assurance for Replacement & Refurbished Steam Turbine Components In some instances, component testing is used to help ensure the less tangible characteristics, such as vibration frequencies and levels of deflection are within acceptable levels. In those instances where the original design criteria are not known, a direct comparison with the original sample is used to ensure compliance. To allow this, a test rig is manufactured, which will allow the reverse engineered components and the original samples to be tested under identical conditions, simulating the operational conditions as much as is practicable. Component installation At completion of manufacture, the reverse engineered component must be capable of being installed into the turbine steam path, while allowing the unit to be returned to service to operate with an efficiency and reliability at least consistent with that of the original components. The installation of the reversed components is often as critical to the performance level of the turbine unit as the actual manufacture of the components. For this reason the owner should take care to ensure that qualified people install the components correctly, and there are records produced of any critical characteristics having the potential to affect the quality of the unit and its short and long-term operation. THE QUALITY ASSURANCE PROGRAM A quality assurance (QA) program is prepared and implemented by an equipment supplier to achieve certain requirements or characteristics in the components produced within his or her facilities. The QA program is normally not limited to measuring and gauging products; this inspection activity represents only a portion of the total QA program. 677 Turbine Steam Path Maintenance and Repair—Volume Two A quality assurance program should include various items controlling total quality. These would typically include: • for an alternative supplier, procedures for reverse engineering. This will ensure the reverse engineered components comply (are equivalent to or better than the original product) and that tolerances are to industry standards. It also includes production planning to ensure delivery dates are met • procedures for controlling the purchase of materials and other items that will affect the quality of the finished components • details of inspection and test methods, their implementation, and a definition of those responsible for this work and their authority. This inspection includes incoming, in-process, and final inspection • methods for controlling the calibration and use of measuring instruments • methods for calibrating and controlling special processes, including the qualification of operators responsible for their application • methods for reporting and evaluating nonconforming conditions as they arise in the manufacturing facility as well as the isolation of nonconforming items from those that comply, and are suited for shipment • methods of packing and preserving the components ready for shipment • details of record storage, retention, and retrieval Such a program should have a person within the supplier’s organization responsible for its implementation. This person must have enough authority to halt production until corrective action is taken (if it is determined the manufacturing process used to manu- 678 Quality Assurance for Replacement & Refurbished Steam Turbine Components facture the components is out of control to the extent they would affect product quality). It is normal for a purchaser, before placing an order, to review the QA manual and undertake an implementation check to ensure the program (as defined in the manual) is operative, and the components have a way of being traced. For small orders, the implementation check is reserved until after placement of the order, and then corrective action requested, if the program does not comply with the purchase specification. THE QA MANUAL An important component of a complete QA program is the availability of a suitable manual that defines the quality philosophy and procedures. The manual is also a description of the working program, and provides procedures that are to be followed to assure consistent and acceptable quality is achieved. The intention of a QA manual is to propound a philosophy. The QA manual is a working document, which must reflect what is done within the component manufacturing plant to assure quality. Therefore, the manual must reflect honestly and accurately what is to be done within the plant, and the minimum standards that will be accepted. It is normal (for major purchases) that the manual will be reviewed before a purchase order is awarded, and an “implementation check” made to gain assurance that the program is being implemented. If discrepancies are found between the written program and the actions within the manufacturing department during the implementation check, these differences are normally resolved before a contract is let. It is necessary to reach a solution that allows practice 679 Turbine Steam Path Maintenance and Repair—Volume Two to comply with the written standard, and yet causes the least disruption to the existing system. However, it is inevitable some of these discrepancies will represent faults, or at least the potential for faults, in the existing program. These need to be acknowledged and addressed. Such areas must be corrected in the preparation of the final procedures to be implemented. THE ENGINEERING REVIEW The purpose of an engineering review is to address those areas where it is possible for errors or nonconforming situations to arise, and where the purchasing engineer wishes to exercise some level of control over the project. Such control normally takes the form of approving specifications that influence product quality, and establishing a means of controlling the disposition (recording and correcting) manufacturing errors, should they occur. During the engineering review, agreement should be reached on the inspection plan, provided by the supplier to the purchasing engineer, and also on hold points, which should be included in a total surveillance plan. The engineering review can also be a pivotal point in the project where decisions are made concerning the need, application, and control of special processes, and agreement is met on matters such as dimensional control of the components themselves. Therefore, the following matters should be addressed at an engineering review: • 680 Quality assurance program—This part of the engineering review should examine the QA program, and its adequacy to help ensure that components are produced to the engineering Quality Assurance for Replacement & Refurbished Steam Turbine Components specification. This includes meeting the purchaser specification and the engineering standards of the supplying company • Drawings—The components to be supplied will be produced to an engineering drawing. These drawings must reflect the following: dimensional tolerances, surface finish requirements, the processes to be used in the production, and any assembly phase (although these processes may be defined in the form of internal standards, which must then be examined) There will need to be agreement as to the availability of drawings to the purchaser engineer for approval and record • Manufacturing processes—There will be a need to examine the proposed manufacturing processes to be used, their method of control, and how the supplier will monitor them. For special processes (those that cannot be gauged for compliance at completion) agreement on methods of process calibration should be defined and agreed upon, if necessary With many processes there is often the involvement of proprietary information, which will need to be considered. The purchaser and supplier will need to reach an agreement on how this aspect of the contact will be managed • Non-conforming components—During the manufacturing process, it will often happen that components are produced outside engineering specification. This will normally be determined by the supplier’s inspection department. When such situations occur, the resulting component will be examined, and an engineering disposition developed. (This is considered in more detail in following sections) • Inspection and test plan—The inspection and test plan is a chronological listing of all tests and examinations with which the components must comply to be considered acceptable. (This is discussed in following sections) 681 Turbine Steam Path Maintenance and Repair—Volume Two • Hold points—The engineer will normally undertake some level of inspection and product verification during the course of the supply contract. There can also be a point beyond which the manufacturing process may not continue before the engineer has had the opportunity to examine the product, its records, and even witness a particular test. This will need to be agreed between the parties to the contract It is normal for the supplier to provide warning of the approach of a “hold point” some time before this point is reached. The engineer can then elect to be present to witness, or “waive” this opportunity or hold point Note: The specifying engineer has the overall responsibility for the quality or compliance with the purchase specification. While the engineer may in fact become involved with a portion of the more critical aspects of production, the responsibility is most often delegated to an inspector or surveyor who will undertake the majority of such verification activities. 682 • Access to sub-supplier plants—Depending upon the nature and extent of the contract, there could be a need for access to sub-supplier plants to inspect subassemblies or material. The need for such access and the availability for other information should be addressed at a design review meeting prior to award of a contract. Such access should also be defined in the purchase specification prepared by the engineer • Quality records—An important consideration for the purchaser of any equipment is the level of recorded information concerning the parts he or she will receive from the supplier. This must be agreed at the engineering review, and will probably have been defined in the purchase specification Quality Assurance for Replacement & Refurbished Steam Turbine Components THE RESPONSIBILITY AND ADMINISTRATION OF A QA PROGRAM In establishing the responsibility for product quality, it is necessary to differentiate between “quality assurance” and “quality control.” Quality assurance (QA) is a program that defines the steps considered necessary to ensure the final product of a manufacturing process meets the requirements of the design specification. Quality control (QC) is a function or element of a quality assurance program, and is the proving or checking of activities associated with such a program. It is necessary at the outset of examining a supplier for program preparation and implementation, to differentiate between these two terms, and what they imply for the product. Without this understanding, there will continue to be a belief that quality will be assured by extensive testing and gauging. This is only partially true. This does not provide complete assurance of quality. The two aspects considered in examining a supplier program are: Program preparation. The QA program is defined in the QA manual, which is a component of the QA program. This definition provides guidance of what steps the supplier will take throughout the material procurement manufacture and shipping phase of a supply contract to ensure the delivered goods meet the specified requirement. This QA program may also, depending upon its level of sophistication, define any internal engineering review and how materials and component parts will be manufactured, controlled, and inspected prior to use. Such a program can include many other procedures and controls depending upon the components being supplied. 683 Turbine Steam Path Maintenance and Repair—Volume Two Program implementation. Once a QA program has been designed, and documented by means of a manual, it must be implemented within the plant facilities. This implementation can, depending upon the program design, extend from marketing through engineering, to the final phases of manufacture, testing and shipping. An implementation review requires that the various functions covered by the program, are adhering to it and follow fully the intention of the total company in achieving a quality product, i.e., one that complies with design specification. Company organization. For a quality program to be effective, the manager responsible for the preparation and implementation of the program must have adequate authority to administer it. This requires: • that this manager must have direct responsibility for all personnel who are responsible for gauging and reporting quality and nonconforming conditions. This will normally include QA staff and all inspection and nondestructive testing operators • that this manager must report to an authority beyond the influence of those people and departments of the organization with a responsibility for production. Such reporting authority can include the works manager, or manager of design engineering A well designed and operated quality program can act in an anticipatory manner, making quality checks early enough within the total manufacturing cycle, so that errors are detected and/or corrected before they become significant and jeopardize a complete product, or procurement schedule. 684 Quality Assurance for Replacement & Refurbished Steam Turbine Components Responsibility for product quality Product quality is a responsibility of design engineering. Design engineering undertakes the basic product analysis and defines the characteristics that must be achieved in the final components to ensure they will operate as expected in the turbine to meet predicted performance. The design engineer provides to the manufacturing department a clear definition of the material properties, possibly the method and sequence of manufacture, the dimensions and tolerances within which the product must fall to be acceptable. This engineering definition will also provide details of any special processes that are to be undertaken, e.g., welding, brazing, heat treating, coating, etc. Design engineering provides a definition of quality of the component. However, the responsibility for ensuring product compliance with this definition is met by the quality assurance department, which monitors the total procurement process and all other aspects of manufacture, and has a responsibility for advising design engineering in the event a nonconforming condition is detected. Many QA programs place a considerable amount of responsibility for quality with the individual shop floor workers, for “in-process” inspection, however the responsibility for final acceptance of each stage of manufacture is placed with the “QC” department for monitoring and design engineering for evaluation and acceptance or rejection of any noncompliance. Acceptance does not imply that at each manufacturing step the design engineering function must become involved. Rather, in the event a nonconforming condition arises, then only design engineering can evaluate this condition and establish an accept/reject decision, and then develop any corrective action that may be necessary and acceptable to total product quality. The implementation of the QA program is one assigned to the QA department, managed by a QA manager. However, the QA manager is merely the keeper and enforcer of the program, with 685 Turbine Steam Path Maintenance and Repair—Volume Two power to monitor its application, and report discrepancies to the design function. The inspection department Many companies have an inspection department located within the manufacturing department, which reports to the manufacturing function. By the definitions and requirements of a QA program, and simple logic, this represents a totally unacceptable situation. This is because there could be a direct conflict, and pressure placed on inspectors, urged by their immediate and ultimate superiors to pass a marginal, or just outside specification piece, rather than declare it as nonconforming. The obvious solution to this problem, if it exists, is to have the QC function report to the QA manager, or alternatively to engineering, which has the final responsibility for product quality. Companies with no separate QA department often have the QC department report to design engineering, or other high level management position that is not concerned with manufacture. In this manner, the quality function is completely divorced from manufacturing, and there is little possibility of conflicts arising. The final quality What quality level must be achieved? To produce a component that is better, produced to tighter dimensional tolerances than specified, or having a surface finish several degrees finer than necessary, or produced from a better material, may be good for the strength or appearance of the component. However, if this is beyond designdefined quality, it is wasteful in terms of the costs of manufacturing the product. For these reasons, the requirements specified by the design function that represent the minimum acceptable should be observed, but no effort or costs should be expended in bettering 686 Quality Assurance for Replacement & Refurbished Steam Turbine Components them. This additional effort will not improve the quality of the product, which meets supplier specifications. Design engineering will assess quality requirements of the products it designs. It will specify for each of these components what it considers to be the quality requirements. These requirements are defined in a number of ways, and presented to the manufacturing department in a number of drawings or specifications. These specifications can also include many specific requirements and other less obvious parameters for more complex elements. The object of the QA program is therefore to direct the attention of the company, and its employees, to ensuring the products meet design specification, which is selected to conform with the purchaser specified, or agreed to requirements. These requirements are considered a minimum, but are intended to ensure the product meets a specified level of performance. It is possible companies are prepared to provide a product that better meets or exceeds these minimum requirements. If this is a conscious policy of the company, this upgraded requirement should be provided by the engineering specification, not left to arbitrary decisions within the manufacturing function, as this could ultimately lead to confusion, and the ability to evaluate and accept nonconforming items. Spending time and money on achieving excess quality characteristics is expensive. However, this could be the price a supplier is prepared to pay to achieve a superior product in appearance or performance. This could be a good business decision, but is not the business of the QA program until it is made so by engineering definition. 687 Turbine Steam Path Maintenance and Repair—Volume Two THE INSPECTION AND TEST PLAN The supplier’s engineering design function is responsible for product quality. It has been stated that the necessary information is supplied to the manufacturing department in the form of drawings, specifications, etc. These establish the “accept and reject” criteria that must be met. In a structured and formal QA program, the engineering function, “the arbiter of quality,” prepares quality instruction to other functions in an “inspection and test plan” (I&TP). The I&TP is a responsibility of design engineering. It can, however (and normally should) have important contributions from other departments within the company organization. However, engineering is ultimately responsible for approving and accepting such contributions. In its simplest form, the I&TP is a chronological listing of the inspections and tests that must be undertaken at each stage of the total production process. The I&TP provides a definition of the standards of acceptability at each stage. In the event a product (at any phase) is found by inspection to be in a nonconforming condition, this must be reported to engineering, which has the responsibility to evaluate the condition, and make a decision to “scrap,” “rework,” “repair,” or “accept-as-is.” The majority of these dispositions are normally quite evident, and “scrap” decisions are often made by manufacturing or inspection, but ultimately are the responsibility of design engineering. In some contracts for complex components, nonconforming conditions may, by agreement, need to be referred to the purchaser, both for information, and in some instances for acceptance, except for a “scrap” decision. The possible exception being that if “scrap” 688 Quality Assurance for Replacement & Refurbished Steam Turbine Components has a serious effect on schedules, even these require purchaser review and acceptance. The I&TP, and the need to refer nonconforming conditions to engineering for resolution, and the purchaser for approval, can be the most contentious issues in the implementation of a QA program. These issues are often resolved only after protracted negotiations. In these instances, the requirements defined in the contract must govern. In the case of nonconforming components, the QA department is not normally qualified to establish acceptability. However, in lower-level programs, the QA manager does often have such responsibility. Manufacturing cannot have such authority. Manufacturing is a department, which by definition, is more concerned with “on time” delivery. Therefore, who within the supplier’s plant is better placed to decide the acceptability of a defect than the designer, or in his or her absence, the QA manager, who monitors to some defined quality requirement? The design engineer or QA manager normally has available information on the performance potential of the components, and is more aware of the probability of failure if a situation is not adequately corrected. The I&TP should also define those records to be developed for the manufactured products. PURCHASER ASSURANCE OF QUALITY The objective for an equipment purchaser in undertaking any form of “surveillance” or “inspection” at a supplier facility is to achieve product quality. But what is quality? A quality product is 689 Turbine Steam Path Maintenance and Repair—Volume Two defined as one that conforms in all respects to the requirements of design. Any effort made by a manufacturer to achieve a product with compliance, in terms of dimensional conformance, surface finish, mechanical properties, or other characteristic beyond design engineering specification, cannot be considered necessary to improve the defined quality level. Certainly, such narrower control may improve the value of the product. But the assumption in accepting the design-specified value of these various component characteristics is that the value defined by design is acceptable from technical considerations. Such a finish may have certain ascetic value, or can even be used as part of a marketing strategy, but if it is beyond the design requirements, no quality value can be attached to it. To obtain assurance that delivered components meet or comply with the design definition, and can be considered a quality product, there are certain steps the purchaser can take. These include an “engineering review” of the design and manufacturing specification, then the undertaking of surveillance of the components during the total procurement process, from material procurement to shipment, to ensure they comply. Alternatively, it can include only that part of the total process considered most likely to influence quality. Product surveillance will help establish that elements are produced in accordance with the design specification, and that all steps, processes, and checks included in the manufacturing processes are undertaken correctly. PRODUCT SURVEILLANCE During the total procurement cycle, there are a number of actions that influence the quality of the product. It is normal for many purchasers to monitor details of production to help ensure components are produced to specification, and will perform ade- 690 Quality Assurance for Replacement & Refurbished Steam Turbine Components quately when installed in the unit. This surveillance normally takes the form of “monitoring” supplier activities to ensure his or her actions produce a product that conforms to his or her design and the purchase specification. It is also normal for the purchaser’s inspection representative to report on any deviations, and to the extent possible verify the error and suggest to the engineer possible corrective actions. In this section, the company inspection technicians will be called inspectors, while the purchaser’s monitoring will be undertaken by the “quality surveyor,” or “surveyor.” Surveillance involves activities beyond what is normally termed inspection, although the most common definition for the person responsible for these activities is normally the surveyor, who directs his or her activities towards critical characteristics of the products. The surveillance (or inspection activities) must be carefully planned to be cost effective. These activities must then be undertaken by a surveyor who has a knowledge of the component, preferably prior experience with the supplier of the equipment, and a clear understanding of those characteristics of the component that are most important to defining its total quality and suitability for use. Preparation for inspection at the supplier facilities Prior to inspection at the supplier facilities, the surveyor must identify the critical characteristics of the components being supplied. While the individual surveyors often have extensive product knowledge, it is not their responsibility to identify these quality requirements. Where then does this responsibility lie, and where will the information and quality definition come from? There are a number of sources including: 691 Turbine Steam Path Maintenance and Repair—Volume Two The inspection and test plan (I&TP). It is normal for each component to be manufactured to a test and inspection plan, prepared by the equipment supplier, and approved by, or negotiated with the requisitioning engineer as part of the “engineering review meeting.” This I&TP will be performed to the standards of acceptable quality. The technical purchase specification. The technical portion of the purchase specification should define overall quality requirements. This is done in terms of performance requirements and the ability of the components to be assembled, maintained, and possibly repaired. However, it is unlikely that such specifications can contain all the information the surveyor will require. Therefore, initial and continuing meetings between the engineer and surveyor are normally necessary to provide the final definition of quality characteristics. The purchasing (requisitioning) engineer. The requisitioning engineer has the primary responsibility for defining, and as needed, identifying those characteristics that require special, or closer attention during the total procurement process. This engineer could also bear the individual responsibility for writing the technical portion of the technical purchase specification. This specification should define the quality requirements in terms of performance (efficiency and reliability), and possibly define individual requirements covered by supplier material and process specifications. Previous operating experience. Records of operating experience with the component, its anticipated life, failure rate and mode, and overall operating performance are normally available at the power plant. These experience factors should be considered in selecting the areas to be given special attention during the surveillance review and recording process. This information can in fact be among the most important information available; also, the requisitioning engineer can often be located at the generating facility. 692 Quality Assurance for Replacement & Refurbished Steam Turbine Components Records of experience with the supplier. When a new supplier is to be used to provide components, there is a learning process involved, as the supplier and surveyor become familiar with the total interaction involved, and with each other’s mode of operation. When a supplier is used with whom the surveyor is unfamiliar, it is often advisable that records of this supplier’s past performance are reviewed for the components involved. This review will help the surveyor to direct his or her attention to areas that have previously proven to be those where nonconforming conditions can arise. Anticipated changes in the operating modes of the components. There are situations where the turbine is to be operated in a changed mode. This changed mode could be coincident with the replacement of worn and damaged components before it is returned to service. Similarly, older units will often be refurbished for return to service, or will have their capacity or steam conditions modified a little. If this occurs, the requirements of the new components could be changed in small, but possibly important ways. The surveyor should be made aware of such changes and be prepared to review any changes this will involve in the total supply process. Minutes of the design review meeting. The design review meeting can be an important part of the award of any contract. The minutes of this meeting will record any technical changes agreed to between the supplier and purchaser, and can include significant changes to the processes involved, the records to be produced and provided by the supplier, and other significant changes to the components to be supplied. The minutes of this meeting will, after agreement, form part of the supply contract, and as such will take precedence over the portions of the specification originally issued to solicit bids. The surveyor must review these minutes. 693 Turbine Steam Path Maintenance and Repair—Volume Two Inspection (surveillance) activities and responsibilities When agreement has been reached between the requisitioning engineer and the surveyor as to the activities representing the most critical operations in the total supply process, and the acceptance standards have been defined, the surveyor must plan his total involvement. This can include, but not necessarily be limited to the following: Quality program review. The quality program within which the components are to be manufactured will have been agreed to at the design review meeting. However, such approval could have been given without any “in plant” audit at the supplier’s facilities, and based on what is perhaps a cursory review of the manual, or on the understanding that the program will be acceptable from previous and other client’s acceptance. Therefore, a preliminary activity of the surveyor can be to undertake a more in-depth review of the quality program, and then to review supplier in-plant activities and records to ensure its implementation. The review and implementation check of the entire program can represent an expensive undertaking. Therefore, in the case of the supply of a short delivery, or critical component requiring the application of only a portion of the total QA program, this initial program review can be limited to those elements of the program that will affect the quality of the components being supplied under the contract. Inspection records. An integral part of any QA program is the documentation of the results of tests, examinations, and measurements made of the components at various stages of the procurement and production process. These documents are made available to the surveyor for review and acceptance. For any document to be considered complete, there are two necessary requirements. The document must be signed and dated, and the person undertaking the supplier 694 Quality Assurance for Replacement & Refurbished Steam Turbine Components inspection must be qualified to complete this work and accept responsibility for accepting it. The signing inspector must be qualified to undertake that work, which can be anything from the use of normal measuring instruments, to the most sophisticated nondestructive examination. The purchaser’s surveyor therefore has the responsibility of ensuring the qualification of the supplier inspector before any determination, review, or acceptance of the results themselves. Inspection instrument calibration. For a supplier inspection to be valid, and to provide acceptable verification of the product or process, the instruments employed to undertake tests and measurements must be calibrated. It is a normal part of any quality program that an internal system exists requiring the periodic recalibration of all instruments used to verify a product. Depending upon the type of instrument, the calibration period can range up to as much as a year for some complex nondestructive testing devices. Instruments employed to measure physical dimensions are normally calibrated on a monthly basis. Each time a device of any type is recalibrated, the calibration date and the result of the recalibration are recorded in a central log. Any tests or measurements taken with an uncalibrated instrument are unacceptable, and must be repeated with a certified instrument. The use of personal measuring instruments within any supplier plant is unacceptable, unless these instruments are entered into the general calibration system and are recalibrated in the same manner as company-supplied instruments. The surveyor has the responsibility to ensure that all instruments used to gauge products are calibrated, and that such calibration is readily observable by the display of a dated calibration sticker or other suitable means. 695 Turbine Steam Path Maintenance and Repair—Volume Two Hold or witness points. Certain critical tests within the total manufacturing and proving cycle of some components or assemblies are of sufficient importance; they must be witnessed by the surveyor. These are known as “hold points.” When these stages of manufacture are reached, it is necessary for the manufacturer to notify the surveyor that such a test is to be performed. It is also necessary to provide this information at some specified time before the test, e.g., 72 hours, to provide sufficient time for the surveyor to be present to witness the test. These are also points beyond which production, manufacture, or assemble cannot proceed without the agreement or approval of the surveyor, or in certain instances where a requisition engineering review of the results are required, without that purchasing engineer’s approval. When such a test is to be undertaken, the surveyor has a responsibility to become aware of the test procedure, details of the results to be achieved, and information of the results to be recorded. If special instrumentation is to be used, the surveyor needs to be sure of its calibration, and that it is in conformance with that identified in the test specification. Many such tests contain what is considered proprietary information, and copies (while not provided to the surveyor or requisitioning engineer), will be available for examination at the test. In the case of this type of procedure, these test specifications will normally have been reviewed by the engineers during the bidding phase, and were accepted. If the test procedure was not available for examination prior to the witness visit, the supplier must allow sufficient time for the surveyor to review, and possibly discuss with the requisitioning engineer what he or she considers to be the most critical steps to monitor. Nonconforming items. It is almost inevitable that during the manufacture of complex components that some nonconforming conditions will occur. When these occur there are certain steps in providing a disposition of the component that must be considered: 696 Quality Assurance for Replacement & Refurbished Steam Turbine Components • The generation of a nonconformance report by the supplier’s inspector, copies being provided to the surveyor • An evaluation of the condition by the supplier’s design engineering function, and the generation of some corrective disposition • Provision of a copy of the engineering disposition to the requisitioning engineer for acceptance/rejection • Undertake corrective action to an agreed plan In the case of a nonconformance, the surveyor is to keep the requisitioning engineer advised, initially of the occurrence (which is often done in parallel with the suppliers notification), then to advise the engineer of any possible corrective action considered appropriate in the circumstances of the error, and finally to monitor the agreed upon corrective action. Special processes. Special processes are those that cannot be examined without destructive testing when the process is complete. In these circumstances it is necessary to calibrate and confirm the process by destructive means, and then monitor its application to ensure the process is performed in an acceptable, and controlled manner when applied to the components. The surveyor should participate in this calibration process and then monitor to ensure the correct procedure is applied throughout the total process. This does not require 100% attendance at the process, but requires, if possible or necessary, repeat calibration processes are undertaken to ensure there has been no “relaxation” of the process during its application. Dimensional and surface conformance. The dimensional checks and review of dimensional inspection records of components probably comprises the major portion of any surveillance activities undertaken by the surveyor. In the early phases of a contract, or 697 Turbine Steam Path Maintenance and Repair—Volume Two when the plant is new to the purchaser, it may be that dimensional compliance is checked continuously. However, as a purchaser company develops confidence in the supplier, it will become less common for the checks to be repeated by the surveyor; rather reliance will be placed on the supplier’s “inspection department records,” and a review of these, together with an examination of any nonconforming material reports. However, sample and “spot” checks are never abandoned. For large supply contracts, “patrol inspection” or arbitrary dimensional checks are still considered a suitable manner of helping ensure quality. Material compliance. These activities will normally involve the review of material test certificates and the comparison of these with supplier material specifications. Such activities should also include the review of heat treatment charts for special application materials, and can, in the case of critical materials, involve the witnessing of destructive and other tests at the plants of material suppliers. For reverse engineered components, it can also require involvement in material identification. Nondestructive testing. An integral part of a supplier’s inspection activities includes the application of nondestructive testing to ensure the material and final components conform to design-specified requirements. The surveyor should review the results of these tests, and on a spot basis witness those that are critical, and, if substandard, could lead to poor performance of the delivered components. It is normal for the surveyor to request copies of all critical material test certificates. Supplier purchaser control. When the manufacturer of turbine components employs sub-suppliers for the procurement of critical components, there is a need for the surveyor to monitor activities within these plants also. This need should be defined in the purchase specification. However, the primary responsibility for monitoring 698 Quality Assurance for Replacement & Refurbished Steam Turbine Components quality remains with the primary supplier, but his checking of subsupplier quality must be documented, and monitored by the surveyor. Documentation and shipment. Prior to shipment, the surveyor should review all documentation relative to the supply of the components, and where appropriate ensure copies of these are made available to the purchaser. The documents supplied to the purchaser are those required for records to allow product quality to be verified at some later time. This will not normally include proprietary information of a design nature, but depends upon what is negotiated as being in the scope of supply at the time the contract was let. It is important to recognize that surveillance does not require the surveyor to undertake double inspection as a normal course (however, it may be more intense in the case of a contract with a new supplier). Rather the surveyor should direct his or her efforts to monitoring the records generated by the supplier inspection and QA staff, to ensure both the production processes and inspections were undertaken by qualified staff and adequate records, providing traceability, are generated. NONCONFORMING SITUATIONS During the total procurement cycle of all components, from material specification to preparation for shipping and packing, it is possible for a situation to occur where the products do not comply with the design specified requirements. This is referred to as a nonconforming situation. In this condition the components are not technically in compliance, and should not be shipped. 699 Turbine Steam Path Maintenance and Repair—Volume Two When the manufacturer or supplier inspection indicates a nonconforming condition exists, this must be evaluated. The logic process of evaluation for performance potential is shown in Figure 11.13.1. This figure outlines the avenues the supplier engineer will explore in deciding if corrective action is possible, and from among several that represent the most appropriate corrective action. With such an evaluation, one of four different decisions can be reached. In some circumstances, the decision is relatively simple to make, and in fact is obvious. In others, options are available, and a decision is made based on the probability of failure or poorer than predicted efficiency, the possible cost of repair, and the ultimate consequences, including the correction of consequential damage, which is the result of not taking appropriate corrective action. In certain situations where delivery is a critical consideration, this must also be considered in the total evaluation. When such a nonconforming condition arises, it is necessary for the supplier to review the total situation, evaluate various corrective actions, and make a recommendation for correction. The responsibility for this correction is one placed on the supplier’s design engineering function, since only the design function is qualified to make such judgment. Once an evaluation has been made and a corrective action decision reached, it is normal for this to be referred to the purchaser for information or approval. This depends upon the form and wording of the contract, which in turn depends upon how the purchase specification was prepared and negotiated. The four basic decisions, and the logic that should be considered are outlined in Figure 11.13.1. The possible corrective actions, in terms of their possible effect on performance are: Scrap and replace. This is a decision that is reached when a manufacturing, process, or testing error has occurred, and the component is no longer suitable for its intended purpose. This component must be replaced, either because it is impossible to correct the situation adequately, or because the cost of correcting it is greater 700 Quality Assurance for Replacement & Refurbished Steam Turbine Components than starting re-manufacture. Often, this is a self-evident decision, and there is little need for evaluation. At other times, this decision is reached only after extensive review of the options. In such a situation it is judged the risk associated with accepting the component is too great. MANUFACTURING NONCONFORMANCE Does it affect availability? Does it affect efficiency? Yes Does it affect maintainability? No No Yes Does it affect short term efficiency? Does it affect long term efficiency? Does it affect reliability? * Accept as is. Yes No Yes Cost of operating inefficient unit. Interchangeability. Long term integrity. Short term integrity. Safety. Accessibility. Correctability. Yes No Accept as is. Yes Scrap and replace. Cost of making correction. Decision to correct. Yes No Rework. No Repair. Figure 11.13.1 Fig. 11.13.1—The evaluation process for a manufacturing nonconformance. This figure The evaluationto process a manufacturing figure should be should be compared Figurefor1.8.1 of Chapter 1nonconformance. for a field foundThis nonconformance. compared to figure 1.8.1 of Chapter 1 for a field found nonconformance. 701 Turbine Steam Path Maintenance and Repair—Volume Two There are occasions when to scrap and replace is prejudicial to the project schedule. Under these circumstances, the evaluation becomes more complex, and there are greater constraints placed on the component supplier. If such a nonconformance occurs, the purchaser must be fully aware of the situation, because to use the component could mean the unit will operate at risk for a period, and until conforming components can be made available and placed in service. There are situations where the nonconforming component can be used for a specified time, but must be replaced at the first opportunity. The replacement part is provided at no cost to the purchaser. Repair. A repair corrects a nonconforming condition, but does not re-establish the original design characteristics within the component. It is often possible to make repairs to manufactured or partially manufactured components, sufficient to allow them to be placed into service. Depending upon the nature of the nonconformance and of the repair, the affected component may or may not ultimately require replacement. Recently, there have been significant advances in many repair and refurbishment techniques (see chapters 7, 8, and 9). This is particularly true in cases involving welding, where new technology has made available materials and techniques capable of extending the useful life of many components, which prior to the development of these fusion techniques would have been scrapped. The technical requirements for performing such repairs are stringent. However, if they allow a component to be saved, and the unit placed in service on schedule, rather than require an extended outage, or until replacement components can be produced from the beginning of the procurement cycle, then the costs and minimal change in risk levels associated with such repairs can often be justified. 702 Quality Assurance for Replacement & Refurbished Steam Turbine Components Again, the repair decision is normally made after a review of the nonconformance, an evaluation of the possible repair procedures, and the level of risk involved. Rework. To rework a component that has some form of nonconforming condition during initial manufacture is to undertake actions that will return it to a condition in which it is equivalent, in terms of dimensional requirements and mechanical properties, to those it would have achieved had the nonconforming condition not occurred. When a nonconforming situation occurs in the manufacture of new parts, if a condition equivalent to those required of the new component can be achieved, such rework should be undertaken. However, it is the responsibility of the design engineer to determine that such equivalence has been achieved. Under most conditions this is an easier decision to reach, if proven procedures for the rework are available, particularly when applied to components in which stress levels are low, and are in a low risk level or environment. Accept-as-is. An accept-as-is decision is one that permits a component to continue in manufacture, or to be used with no effort being made to correct the nonconformance. Two reasons for reaching and deciding upon this course of action are: • There is little need to make any corrections. To make them will add no, or at best marginal improvements to the turbine performance • The cost and time of replacing, repairing, or refurbishing cannot be justified, either for the turbine or for the degree of noncompliance in the components. Often to rework or repair the situation could increase the risk to performance level 703 Turbine Steam Path Maintenance and Repair—Volume Two This “accept-as-is” decision is often based on the experience of the supplier design engineer and can only be made by being aware of any risks that are introduced. Such a decision should not be made as a desperation measure. The risks, if any, should be fully evaluated, and the options, from an extended outage before return to service, and the probability of failure, must be fully considered. The “accept-as-is” decision can often be made, being aware of the risks, while replacement parts are obtained. This decision or evaluation process can be complex. Occasionally the solution is self evident, such as when a nonconformance exists to the extent parts cannot be used, manufacturing must cease, and the parts replaced. In those instances of a nonconforming condition, when there is no time to correct the situation before the scheduled date for return to service, mature judgment is required on the part of the design and operating engineer, and acceptance of the fact that the unit will operate at risk if the nonconforming components are used. AVAILABLE QA PROGRAM When a company installs a QA program, it must represent their actual “in-plant” procedures and activities. These activities should be reflected in both the QA manual, and the procedures developed to define their methodology. There are many national and international standards in existence, many developed to provide guidance to suppliers of equipment to the military or other complex and demanding industries. These programs provide a secure basis from which to develop a company program. However, the supply of turbine components may not be as complex, and while the supply of such components that do not meet design specification can have 704 Quality Assurance for Replacement & Refurbished Steam Turbine Components serious and expensive consequences for the purchaser, the installation of a program with complexities beyond what is required to provide a suitable product is wasteful in both time and costs to the supplier and purchaser. An international standard that is developed and utilized now by many companies as a basis for their internal programs is the ISO 9000 series. This international series of programs is three tiered, having three levels of complexity. The most appropriate level in any component supply contract is dependent upon the complexity of the supply process, and should reflect the degree of any project that is to be undertaken. The complexity of the program should not be confused with the complexity of the component form and material. A complex component can be produced within a relatively simple program if the processes in its production are understood, and there are no anticipated difficulties within the total supply process. THE MACHINING OF TURBINE COMPONENTS The components of the steam path are produced to exacting standards, and design specifications are set so the supplier of the equipment can provide a “quality product” at an acceptable cost. This allows the operator to have an efficient and reliable unit available on the system. Also, the manufacturer has an income that covers cost, provides a profit to his shareholders, and has sufficient funds remaining to support further development of the product and its components. An integral part of establishing the final cost and quality of the steam turbine is the expertise and control of the manufacturing 705 Turbine Steam Path Maintenance and Repair—Volume Two process. This process, while very much influenced by the individual machinists and their attention to detail, is also a product of the engineering specifications, and the machine tools and cutters made available to the machinists. The total production of the steam turbine involves using a number of different metal forming processes and cutting techniques. A number of different techniques are required for different components. Also, the level of finish and tolerances applied can be different, not only from component to component, but within the components themselves. In general there are more stringent requirements placed on the rotating components than those that are stationary, such requirements being related to the effects of stress rather than expansion and flow efficiency. In general, the cutting or forming processes selected for the production of each component of the steam turbine are those most suited to provide the tolerance, and produce the surface finish required. The manufacturer of steam turbines will have available machine tools suited for each major component. However, there is continual development in manufacturing these components, and certainly the introduction of advanced computational fluid dynamic techniques has allowed a more sophisticated aerodynamic form to be defined. For these reasons there can be situations where components, while manufactured with considerable care, are not necessarily produced by the most suitable and economic means in the early phases of production of the particular components. The cutting process Fortunately, the majority of the machining undertaken in the production of turbine components utilizes conventional cutting techniques. Therefore, the necessary machine tools are always available, and the factors to be considered in defining requirements are as follows: 706 Quality Assurance for Replacement & Refurbished Steam Turbine Components Surface integrity The quality of the machined surface is dependent upon a number of factors, the majority of which are related to the quality of the machine tools and cutters used. It is possible that a surface can conform to dimensional, and apparently conform to surface requirements, yet contain deficiencies that have the potential to degrade the surface quality. Such deficiencies may have little or no effect on component efficiency, but can introduce structural weaknesses. Typical of these deficiencies are: Tears and gouges. The tearing or gouging of a surface can introduce microcracking, which (depending upon its location) can be the initiation site for some form of rupture. These surface discontinuities can be caused by a number of independent factors, such as tool sharpness, cutting speed, and the rate at which material is being removed from the surface. Tool chatter. To achieve a consistent conforming surface, the cutting tool must be mounted correctly, securely fixed, and the material to be removed must be removed at the correct rate, in terms of cutting speed, feed, and depth. These are necessary so that the tool is free from any form of tool vibration. Surface burning. Surface overheating can cause carburization of the surface. This can be caused by the cutting rate being too high or by there being insufficient cutting fluid. Built-up edge retention. The cutting action removes material from the machined part, forming a series of “chips,” which can form an apparently continuous strip of material. In fact, this strip can contain a series of chips that are fused together by the heat generated by cutting. It is also common for the chips to build up on the cutting edge of the tool, as shown in Figure 11.15.1. This edge can break loose under the effects of the cutting force, and become embedded in the main material surface. 707 Turbine Steam Path Maintenance and Repair—Volume Two Such embedded “built-up” edges, if they exist in the final surface, represent an undetectable, but present surface discontinuity. Fig. 11.15.1—Showing the “Built-up Edge” (BUE) which detaches periodically and can be buried in the surface of the component being machined. Surface finish The reasons for specifying a particular surface finish on any component, or part of that component can be for one of five reasons: 708 • To achieve a condition with minimal efficiency loss (see chapter 6) • To minimize the possibility of stress concentration • To allow nondestructive examination from the surface • To allow the surface to form a joint that is tight against leakage of some fluid or gas through it Quality Assurance for Replacement & Refurbished Steam Turbine Components • To produce a surface that is finer than the engineering specification will normally cost more in terms of machining and finishing requirements, will add no technical value to the product, and cannot be considered to improve the “quality” of the component Definition and means of quantifying surface finish are covered in chapter 6. Machining rates (speeds and feeds). The machined surfaces of a turbine component can contain many levels of complexity. This is particularly true of blading, where complex vane forms are defined, and where precision is of utmost importance. The rate at which these surfaces have material removed to form these surfaces will influence their quality, and since, as dimensional changes occur, so will the relative cutting speed between the material and cutter. These speed differences therefore must be accommodated in the total machining definition. In producing the final or finish surface on any component, it is possible to use a number of different cutting tools. These tools must be properly sharpened and suited to the material being machined. This is necessary to help ensure the chips that are formed by the cutting process are removed, rather than forced into the surface of the material where they will exist as material discontinuities, capable of causing stress concentration, and forming a “rough” surface. Cutting fluids. Fluids, normally liquid, but occasionally gas, are used to surround the metal cutting edge to achieve certain objectives, including the removal of heat, the lubrication of the surface, and the washing away of the cut off material. Therefore, fluids must be selected to achieve these objectives. However, these fluids must contain no additives or other constituents that are, or could become corrosive, if not completely removed from the components before being assembled in the turbine. 709 Turbine Steam Path Maintenance and Repair—Volume Two REFERENCES 1. Andrew, D.D., and W.P. Sanders. The Concept of Reverse Engineering, Turbomachinery Maintenance Congress, Berlin, October, 1991 2. Surface Texture, published by the American Society of Mechanical Engineers, United Engineering Center, New York, New York 3. Sanders, W.P. Select your Supplier of Steam Turbine Blades Judiciously, Power, April, 1984 4. Kaczmarek, J. Principles of Machining, by Cutting, Abrasion and Erosion, published by Peter Peregrinus Limited, Stevenage, England, 1976 710 Chapter 12 The Manufacture and Inspection Requirements of Steam Turbine Blades INTRODUCTION The components within the steam turbine most susceptible to damage, most often replaced, and with modern repair techniques among the most often repaired or refurbished are the rotating blades. Complex stresses are produced in these components during operation, at both steady loads and operating conditions, and possibly amplified during transient operation. Because of this, and their sensitivity to vibratory type damage, it is important that when these components are replaced, the design engineering specification for each is closely followed. 711 Turbine Steam Path Maintenance and Repair—Volume Two This chapter considers the characteristics of the blades, which must be considered when they are manufactured, and the important aspects of their inspection. The most significant characteristics of any blade row is dependent in part on their operating environment, the pressure and temperature of the steam, and the type of loads to which they are subjected. For this reason there are considerable differences in the type of information required to be checked, and the inspections that must be applied to blades from different rows. The most important checks for any row are dependent upon the size of the elements themselves, and where they will be installed in the turbine. The various manufacturers have developed methods of making characteristic checks of their blades, and in many instances have developed not only manufacturing techniques suited to their products, but have designed inspection devices and techniques for proving compliance with design characteristics. The important characteristics of the blade rows are not always apparent, and therefore it is necessary to consider other means of ensuring that the blades comply with design specifications in terms of material, dimensional requirements (including surface finish) and (for the longer elements) that their vibration characteristics are acceptable. RADIAL ALIGNMENT OF ROTATING BLADES An important characteristic that must be maintained, and becomes more critical as the radial height of the vane increases, is the radial alignment of the blade vane to the root. The problems associated with misalignment of the vane include the distortion of the expansion passage formed between the vanes, worsening along 712 The Manufacture and Inspection of Rotating Blades the radial height towards the tip, and the centrifugal bending stress induced in the blade as a consequence of such misalignment. Shown as Figure 12.2.1 is the root and tip section of a long vane, and the position of the center of gravity of both “G.” Under normal circumstances, these two positions of “G” are located above each other and above the center of gravity of the root platform to the greatest extent possible. There are however, exceptions to this. On large radial height blades there can be large steam bending forces developed on the vanes as a consequence of the change of steam momentum in flowing through the expansion passages between the vanes. In an effort to reduce the steam bending effect, the vanes can be given a tilt in both the axial and tangential directions to counter a portion of the stress induced by steam bending effects. Fig. 12.2.1—The root and tip section of a large rotating blade with the center of gravity ‘G’ above a common point. 713 Turbine Steam Path Maintenance and Repair—Volume Two Figure 12.2.2 shows the steam forces developed on the vane. This steam force “Fsb” can be resolved into two components, one axially “Fas” and one tangentially “Fts.” These forces will induce stress “fas” and “fts” in the blade vane. These forces and stresses are transmitted through the vane to the root, and affect both the blade fastening, and the rotor or wheel to which it is attached. Fig. 12.2.2—The steam forces ‘Fsb’ developed on the blade, divided into tangential ‘Ft’ and axial ‘Fa’ components. Incorrect radial alignment of rotor blades During manufacture of the turbine blade, it is essential the vane be produced in correct radial orientation relative to the root form. There are various manufacturing errors that can cause the design value of alignment not to be achieved. Failure to maintain correct radial positioning of the vane relative to the root can result in the rotor blade, on assembly, being in an incorrect radial alignment within the steam path. Such misalignment can result in incorrect pitching and a maldistribution of the steam flow within the blade row lowering expansion efficiency. 714 The Manufacture and Inspection of Rotating Blades However, a significantly more hazardous consequence of this incorrect alignment is that during operation, the centrifugal force of the vane will attempt to rotate the blade about its root section to correct, and achieve radial alignment. As discussed previously, many manufacturers calculate blade stresses due to steam flow, and then take advantage of the centrifugal bending effect to counter a portion of these steam-bending effects. Therefore it is obvious that the magnitude of this centrifugal bending stress can become significant, and if blade vane location is not adequately controlled during manufacture, this can introduce excessive stresses on the blade, capable of inducing premature failure. Consider a blade shown in Figure 12.2.3, in which the locus of the center of gravity of the vane is shown as line “Gr-Gt” from the root to tip. Here the center of gravity of the vane is coincident with the center of gravity of the root platform at the point of attachment to the platform. This locus can be straight or curved, curved being more consistent with a varying profile vane. Also shown is the radial line “R-R.” This line passes through the root center of gravity at the root platform top. With this blade, the vane is set forward by an amount “da” at the center of mass “m” that occurs at a radius “Rx,” and increases to a tip movement of “ϕa” in the axial direction. Similarly, the blade vane requires a tangential adjusting tilt of “dt” at “Rx” which is equivalent to “ϕt” at the tip. During operation the blade will have steam forces developed on it due to the change of momentum as the steam flows or expands across it. This steam momentum force can be resolved into two components, one of these in the tangential direction, and the other in the axial. During the design phase the designer calculates these forces as a function of blade height, and then selectively adjusts the tilt of the blade in both the axial and tangential directions to counter a portion of these steam forces. In this manner it is possible to lower the operating stress levels. The calculated tilt “ϕa” and “ϕt” is 715 Turbine Steam Path Maintenance and Repair—Volume Two normally small, of the order of 2-4°. Therefore it can be seen that even marginal differences in vane radial orientation can have a significant effect on the blade stresses. R da Mac da R Mas Gt Steam flow m dt Mts Mtc Gt Dt dt m f as fts Rx Gr Gr Dr (a) R (b) R Fig. 12.2.3—The steam and centrifugal bending moments on a blade. The bending moments in both the tangential and axial directions due to the steam bending effect, and the centrifugal bending effect are shown in Figure 12.2.3. Mac Mas Mtc Mts = = = = Axial moment due to blade centrifugal load Axial moment due to blade steam load Tangential moment due to blade centrifugal load Tangential moment due to blade steam load It is recommended, especially for longer radial blades, to undertake gauging (an audit) of the pitch and throat at various blade 716 The Manufacture and Inspection of Rotating Blades heights. Table 12.2.1 shows the results of such an audit on a partial row of rotating blades. This audit measured 20 passages on a 14.215" vane, the readings being taken 2.0" below the tip. Figure 12.2.4 shows the results of this audit on a partial row. The variation in the measured data can be seen, and the level of variation established in the ratio “O/P.” Passage 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Throat Height Pitch O/P ’α‘ 0.671 0.658 0.672 0.674 0.670 0.676 0.674 0.675 0.673 0.668 0.679 0.685 0.672 0.672 0.668 0.670 0.675 0.667 0.665 0.679 0.675 14.045 14.193 14.234 14.214 14.260 14.219 14.188 14.183 14.193 14.219 14.209 14.239 14.209 14.198 14.183 14.193 14.173 14.204 14.234 14.255 14.222 1.754 1.744 1.749 1.759 1.732 1.776 1.753 1.734 1.771 1.733 1.746 1.756 1.744 1.751 1.760 1.723 1.753 1.748 1.746 1.731 1.759 0.3826 0.3773 0.3842 0.3832 0.3868 0.3806 0.3845 0.3893 0.3800 0.3855 0.3889 0.3901 0.3853 0.3838 0.3795 0.3889 0.3851 0.3816 0.3809 0.3923 0.3837 22.49 22.17 22.60 22.53 22.76 22.37 22.61 22.91 22.33 22.67 22.89 22.96 22.66 22.57 22.31 22.88 22.65 22.43 22.39 23.10 22.57 Table 12.2.1—The audit values taken from a 14.215" radial height rotating blade: 2.0" below the tip section. Typical tolerances for these parameters in this rotating blade are: Pitch “P” +/- 4% Throat “O” +/- 4% Ratio “O/P” +/- 2% 717 Turbine Steam Path Maintenance and Repair—Volume Two 1.76 0.67 1.75 1.74 0.66 1.73 1.72 0.65 O/P - 2" Below Tip 14.215" 1-2 (b) 1.78 1.77 0.395 0.390 6-7 O/P Sin β 2 11-12 16-17 Pitch - 2" Below Tip. Inches Opening Pitch 0.68 21-22 23.0° 0.385 22.8° 0.380 22.6° 0.375 22.4° 0.370 22.2° 22.0° Sin β2 - 2" Below Tip 2.0" (a) Opening - 2" Below Tip. Inches Note: The actual tolerances are normally set in terms of a percentage of the vane radial height, but as such no industry standard exists. However, exceeding the above values is considered excessive. (c) Vane Height. Inches 14.30 14.25 14.20 Design Height 14.15 14.10 14.05 14.00 Fig. 12.2.4—The results of an audit of 20 expansion passages in rotating blades. These readings were taken at their discharge point 2.0" below the tip. At batch end positions, these values may be relaxed a little, causing a permissible 25-50% increase in the values. However, this should not be necessary for quality-manufactured blades. For the stationary blades the tolerances are discussed in chapter 7, and the results of an audit are shown. The stationary blades, because stress levels are significantly lower are capable of being adjusted to some extent, and the effect of adjusting repaired stationary blades is discussed as well. 718 The Manufacture and Inspection of Rotating Blades Blade vane tilt This bending effect can be introduced into the vane by its geometry, and an offset bending moment induced in the vane due to its tilt from the true radial position “R-R.” This offset can be in the tangential and/or axial direction. When tilt is specified by design, it can be selected in both the axial and tangential direction to balance a portion of the steam bending effects considered earlier. However, the steam bending effects in the axial direction are often sufficiently small that no attempt is necessary to make a bending balance. It is common practice for the design process to specify a degree of tilt in the larger radial height blades. Where this tilt or offset is calculated and produced in a direction that will induce a bending moment capable of countering a portion of the steam bending moment. Figure 12.2.3 shows the effects of offset in the blade vane. In this figure are shown the effects of the total bending moment in the tangential direction “Mts” due to the steam, and those due to the centrifugal effects “Mtc.” Similarly, in the axial direction the bending moments are “Mas” due to the steam, and “Mac” due to the centrifugal effects. In the tangential direction, there is a total bending moment acting on the vane of “Mts-Mtc,” and in the axial direction there is a total bending moment of “Mas-Mac.” These bending moments cause a stress in the axial direction of “fas,” and in the tangential direction of “fts.” If the vane is then manufactured to be given total tilts of the tip section center of gravity “Gt” by amounts “ϕt” and “ϕa” in the tangential and axial directions respectively, as shown in Figure 12.2.5, these tilts will produce bending moments in opposition to the steam and pressure moments, thereby canceling a portion of the stresses induced in the vane. In this figure the “tilt shift” from “Gr” to “Gt” is shown (the tip originally being centered over “Gr”), and the tip movement in the axial and tangential directions “ϕa” and “ϕt” can be seen. 719 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.2.5—The tip shift due to axial ‘ϕa’ and tangential ‘ϕt’ tilt. Consider a blade in which the steam bending moment in the tangential direction is “Mts,” inducing a bending stress at the blade root of “fts.” In this same direction, the blade can be given a backward tilt at the tip, giving a bending moment of “Mtc,” which is equivalent to a fraction of the steam bending moment, the fraction selected being designated “-dMts,” if the steam moment is taken as “Mts-dMts.” Unit load “M ts” “M-dMts” “sMts” “fstr” Case 1 100 50 0 100 50 0 100 100 100 0 -50 -100 0 -50 -100 Case 2 100 50 0 100 50 0 50 50 50 +50 0 -50 +50 0 -50 Case 3 100 50 0 100 50 0 0 0 0 +100 +50 0 +100 +50 0 Table 12.2.2—Resultant stresses from tilt variations (alternative tangential displacements ‘dtg’). 720 The Manufacture and Inspection of Rotating Blades Note: In this table three different cases of tilt levels have been considered. In the first, the blade is tilted to balance the steam moment at full load. However, when the unit operates at no or low load levels there is a high bending stress induced. In the second case, the tilt is adjusted to balance 50% of the steam moment, and in this case the bending stress does not exceed 50% of the potential maximum. In the third case, the blade is given no tilt, therefore the steam moment is unbalanced. The resulting stress can in this instance be a maximum when the unit is running at full load. In making this tangential tilt adjustment, there is a need to modify the tangential correction tilt to account for the pitch or chordal enlargement, which occurs as a function of radius. This is done as a ratio of the diameters “Dr” to “Dt” in Figure 12.2.3. A similar tilt adjustment can be made in the axial direction, if the stress levels justify such an adjustment. The extent of the adjustment in either direction represents relatively small adjustments in the order of 0.050" to 0.015". For this reason it can be seen that achieving the correct radial alignment when mounting blades on a rotor is critical, in terms of reducing excessive and unknown levels of bending stress. No radial position adjustment is required in the case of axial tilt. Center of gravity shift of short blades When a centrifugal bending effect is produced on short blades, no effort is made to account for this as the total effect on the stresses is small. However, as shown in Figure 12.2.6 there are blades that can have a considerable “offset.” This offset is accounted for in the design phase, normally from experience. But, if the blades are not mounted securely (packed) to the wheel, and looseness exists, this will allow the blades to tilt in operation, resulting in the load-bearing surface in the root being loaded heavily at one end, and this can result in root damage. 721 Turbine Steam Path Maintenance and Repair—Volume Two da Gv Gr dt Bending moment BMv χ Fig. 12.2.6—The bending moment ‘BMv’ caused by the displacement of the vane CofG Gv’ above the root CofG ‘Gr’. Blade tilt and the effect on tenons If blades that have a manufacturing (as opposed to a design) with a tilt occurring randomly, then there will be some difficulty in assembling any coverbands required on the stage. For short blades, on which it is not possible to deflect the vane, it will be necessary to custom cut the tenon holes in the coverband. However, on long vanes it is possible that efforts will be made to deflect the blades by some small amount, sufficient to allow the coverbands to be passed over the rivet heads. While this allows assembly of the rotating row components that cannot be detected, it will cause an elastic deformation of the vane, and high contact pressures between the tenon head and hole. This introduces complex stresses, possibly of low magnitude in both the coverband and tenons. 722 The Manufacture and Inspection of Rotating Blades In operation this could eventually cause high-cycle fatigue or fretting, or some other phenomena at the interface that could develop into mechanical failure. BLADE MANUFACTURING TECHNIQUES Among the many complex components of the steam turbine, blades are a main cause of performance deterioration, in terms of both structural integrity/mechanical failure, and degradation of their efficiency. For this reason it is valuable to consider aspects of the manufacture of these components, as the majority of the technologies employed in their production and final acceptance are also used in other critical components. Possibly the one process not used at this time in the production of steam turbine blades is casting. The remainder of this chapter considers the manufacture and quality requirements of blades. This is convenient, as it is recognized that these components contain some of the most demanding requirements in the production of steam turbines. The designer of steam turbine blading spends a considerable amount of time, and applies technological expertise, calculating stage requirements and establishing suitable vane forms and root geometries for the individual stages. At completion of the design phase, a specification for the blading is prepared, which is intended to ensure it is manufactured to acceptable standards so it is able to operate for a sufficient period, and at a performance level consistent with the requirements of the system into which the turbine is being installed. The blade design specification can be considered to be an engineering definition of the technical requirements of the product, which 723 Turbine Steam Path Maintenance and Repair—Volume Two will also define the standards to which it will be manufactured to achieve the correct quality. The blade specification will establish the following parameters, which must be complied within the final product: 724 • Material specifications—this includes material from which the blade is to be manufactured. This portion of the specification will normally identify the material by specification number, the specification being either a generic type material, or more likely an “in house” material developed by the manufacturer to meet his or her particular requirements, for long-term operation, within the local steam environment • Dimensional requirements—the dimensional requirements will be a specification for the form of the vane, its length, and any profile variation as a function of height. The root form will also be defined, normally as a standard root profile for which cutters and other manufacturing capabilities exist • Manufacturing tolerances—the tolerances within which the vane is to be manufactured, the tolerances of the blade root and any stage hardware required to mount the blades to the rotor • Spatial relationships—this is the spatial relationship between the vane and the root on which it is to be mounted. This will include tolerances for position on the root platform, and any radial lean, which will be required to counter steam-bending stresses • Surface finish requirements—the surface finish requirements can include not only the degree of finish, but in certain instances the direction of finish • Manufacturing techniques—the manufacturing technique will be used to produce the blade, and can include the basic metal forming techniques, such as forging or cutting, and in certain instances, the machine tools to be used to make the cutting or forming The Manufacture and Inspection of Rotating Blades • Special processes—the special processes are used to improve the operating characteristics of the blade. This can include welding or brazing techniques to be used for the attachment of stage hardware or erosion shields, or processes used to join the blades together in groups. It can also include processes such as hardening, coating, or shot peening. In each case the engineering definition should establish the standards against which these processes are to be calibrated • Nondestructive examination—the non-destructive, and even destructive tests are used to establish acceptability of the blades and stage hardware. Such testing techniques can also be applied to the results of the special processes that are used on the blades • Blade mounting—the blade mounting to the rotor is an assembly that is an integral part of the total manufacturing process. Any special requirements for mounting and alignment for all stages will be defined, as will the arrangement for any special closing requirements • Any post assembly requirements—this defines any machining or finishing requirements after the blades are mounted to the rotor. This will include the tolerances that are to be achieved to ensure correct position of the blades and coverbands on the completed rotor to achieve axial and radial clearances and stage lap Many of these engineering definitions, or quality criteria are set as standards of the industry, or manufacturing standards of the manufacturer, and are not redefined for each blade or row. In addition, many of these requirements are established in agreement with the manufacturing engineers responsible for producing the blades. However, in the final analysis the quality of the blade is a design engineering responsibility, and it is the designer who must adjudicate in the situation of any nonconformance that arises during the manufacturing process. 725 Turbine Steam Path Maintenance and Repair—Volume Two THE BLADE MANUFACTURING PROCESSES The turbine manufacturer normally obtains the materials used to manufacture steam blades from the materials supplier in one of several specified forms. The form of the blade material required for any stage is defined by the design engineer, and selected on the basis of material suitability, and to a degree on material availability. The defined material also considers the most economical method of producing the blade. These selections of material form are always made with the requirements of unit availability in mind. The material delivery form is defined by the design engineer, who also coordinates with the manufacturing engineer to define the manufacturing process, or processes that will be use to produce the final form. There are obviously different requirements in specifying various blade materials, either as bar stock, precision, or envelope forgings or other forms, although the materials will have similar mechanical properties. However, any specification defining the requirements for replacement blades should identify the preferred material production requirements. The specification should be sufficiently specific that if bar stock is used in place of forgings, there is no degradation of the mechanical properties or microstructure of the material itself. Blades are manufactured using various processes, or basic metal forming procedures. Principal among these are: Metal cutting. Blades produced entirely from bar stock by metal cutting are normally those that are short, and have no change, or relatively small change in vane profile along their length. The bar stock is cut and shaped to the final form by metal removal alone. With modern material production techniques, bar stock can normally be produced to provide the same material properties and quality characteristics as forgings. However, for many blades normally 726 The Manufacture and Inspection of Rotating Blades produced from forgings there is a considerable twist in the profile, and to produce these blades from bar stock would require the removal of substantial quantities of material. This can require an extension of manufacturing time and increased costs on a per blade basis. For small quantities of the vortex blades, bar stock is often acceptable and economical, and allows short period deliveries. For larger quantities, the cost of the forging dies can be offset by both material and the production costs, and by the time required to remove the excess material from the bar. Even for forged blades the dies do wear, requiring maintenance. This is particularly true of precision forging dies. Envelope forging. An “envelope” forging is one that requires the removal of material from all surfaces to produce the final blade form. This type of material delivery is normally specified for blades with a large radial length, which to produce from bar stock would require the removal of considerable quantities of material. Material is removed from all surfaces of the envelope forging, and it is normal for the forging to provide sufficient material both for the production of test pieces, and machining location positions on the main body of the forging. To produce these blades from forgings requires the production of dies, which are expensive, and therefore the blades produced by forging are normally a standard type, in which the same vane and root form are used. An exhaust stage blade produced from an envelope forging is shown in Figure 12.4.1. Precision forging. The precision forging does not require material to be removed from the pressure and suction faces of the vane. However, some manufacturers will undertake some polishing of these surfaces, but this can be for cosmetic purposes only on large elements with low Reynolds numbers. The vane inlet and discharge edges will require the removal of “flash,” which is the excess volume 727 Brown Boveri, Inc. Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.4.1—An exhaust blade produced from an envelope forging. 728 The Manufacture and Inspection of Rotating Blades of material beyond that required to complete the blade. This material is “squeezed” out from between the dies. Note: With proper die cleaning between the production of individual forgings, the forged surface is normally in a fully acceptable condition. If, however, the dies are not cleaned between the production of each blade, there is the possibility of scabs being formed on, and embedded into the surface, so they do require cleaning and polishing after cooling. This means of production does not require the removal of metal from the vane surfaces as in envelope forgings. However, there is a requirement to machine the root form, and possibly tie wire holes or snubbers. Electric discharge machining. Electric discharge machining is used principally for stationary blade rows, and has been employed for the manufacture of both first stage nozzle boxes and complete diaphragms. However, this method has been used successfully by at least one manufacturer to produce solid machined rotating blades in discrete groups. These are used in control stages, which are machined so they can be attached to the rotors by pinning. Figure 12.4.2 shows the portion of such a row, with blades being removed from the rotor, each group of three blades being held in place with three axial pins. Vane extrusion. A method of vane production used on some older designs, and units that are still in operation in many plants, both utility and industrial, is that of making the blade vane portions from an extruded section. In these blades the vanes are held apart by “spacer pieces.” These spacer pieces are tapered at an angle equal to ψ, where ψ = 360/Zb, and Zb is the number of rotating blades in the row. This is the number of pitches and includes the closing blocks if these are used. The geometry of the spacer piece is shown in Figure 12.4.3. Here the top portion of the piece is produced to a height “H” to control the discharge height of the passage, and the sum of the 729 Turbine Blading Repair, Inc. Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.4.2—Groups of three blades produced by electric discharge machining. These blades are in integral groups of three and are attached to the wheel by three axial pins. Fig. 12.4.3—The extruded vane and the spacer piece placed between the vane segments. 730 The Manufacture and Inspection of Rotating Blades pitches “Pv” and “Pp” is equal to the design pitch between the vanes. The vanes normally have a surface finish consistent with that of the extrusion dies, i.e., with these vanes the manufacturer does not normally make any finishing polish onto the pressure and suction faces. The root forms are selected to allow the form to be machined into the base of the blade by cutting in the tangential direction. The spacer pieces have the same root form machined into them, and are then assembled to the rotor through an access window. On some stages with a large centrifugal load, the vane is bent or “upset” under the spacer piece, as shown in Figure 12.4.4. This helps to secure the vane, and transfer the centrifugal load from the vane “V” to the spacer piece “P” at a load transfer surface “a-a,” which then provides a greater load transfer surface to the rotor. Fig. 12.4.4—The vane “bent foot” under the spacer piece. Pinch rolling. Blades produced by pinch rolling are formed in the hot condition by pinching a billet between rotating rollers. In this method of manufacture the vane is hot formed and (depending upon the rollers) can be rolled to the final shape requiring little or no surface polishing to complete the vane. It may be necessary to dress “flash” from the inlet, and discharge noses after rolling. The vane is then clamped and the root form is machined onto it. 731 Turbine Steam Path Maintenance and Repair—Volume Two The requirements of tolerances are such that while precision forging, pinch rolling, and other non-metal cutting processes can produce vanes, the root portions of all blades are normally produced, or their final dimensional requirements achieved, by metal cutting. This is necessary to ensure the vane has the correct spatial relationship to the root, and the blade vanes achieve the correct radial alignment when mounted to the rotor. Note: In many blades with a large radial height vane, the vane is “tilted” in both the tangential and axial directions to counter a portion of the steam bending stresses induced in it. This tilt is selected so that at a predetermine steam flow the centrifugal bending effect is cancelled in both directions. The extent of tilt is relatively small, which implies that if a blade vane is not aligned correctly to the root for any element there can be large bending stresses introduced into the blades. Therefore it is important that the tip pitch is monitored during manufacture and mounting to minimize this stress. Because blade forging and rolling processes represent “hot” working of the material, there is a need to stress relieve after forming the blade, and such heat treatment can cause the vane to “warp” and/or “twist” to some degree. For this reason, there could be a need to undertake some level of bending or straightening after the basic blade has cooled. This adjustment is a cold working process, and is necessary to ensure the vane lies with the correct radial orientation, and setting angle. If the vane has to be adjusted cold, it is a good practice to subject it to some form of heat treatment, after completion of adjustment. This is necessary to ensure its freedom from plastic deformation, and any residual stresses that may have been induced by the adjustment. 732 The Manufacture and Inspection of Rotating Blades Basic form production by material deformation Material deformation is a means of producing the basic form of the vane. This is normally completed by deforming a portion of the material to form the vane, and leaving sufficient material to produce the root from an integral bulb of material. The material forming processes include forging, extrusion, and pinch rolling. This is a hot deformation process, and heating is necessary to ensure the blade material is in a structural condition that will allow it to be worked into a suitable form, having no residual stresses locked into it. Of these processes, only the envelope forging requires significant amounts of material to be removed from the vane after the basic forming process. The other processes may involve a finishing operation, but this represents only a small amount of material removal, producing a finer finish, and will not represent a significant dimensional change to the vane. Forging of the material The two methods of forging a blade require the use of dies, a male and female portion, which are forced together over a hot billet. Figure 12.4.5 shows forged blades being removed from the heat treat furnace. Cutting metal to form the vane The processes selected to remove the excess stock from the basic material, either bar stock or envelope forgings, for the production of the vane, are a function of its form. The selected processes for any blade depending upon various factors, including: • the complexities of the profile, which include the degree of twist and taper, and the need to produce integrally those discontinuities such as tie wire stubs, or localized thickening in 733 Turbine Steam Path Maintenance and Repair—Volume Two the region of tie wire holes, possibly tip thinning or thickening, and the possible production of an integral coverband • the machine tools and techniques that are available within the manufacturing plant. In fact, if large numbers of a particular blade are to be produced, it may be economical for a manufacturer to invest in new facilities to accommodate the new requirements. Usually the blade is designed so it can be produced with the machine tools and facilities available • the material from which the blades are to be produced. There are some materials used for blading, particularly the titanium alloys, which require a different manufacturing technique, cutting tools, or production sequence from those made from steel. The form of the material, i.e., bar stock or forgings, could also influence the cutting procedure for various materials • the total economics of the production process. The manufacture of blades is expensive when a new blade form is to be produced. The design and manufacturing engineers, together with marketing, will examine the anticipated manufacturing scheduleand predict future requirements, and based on these estimates and the existing facilities, select a manufacturing technique best suited to overall costs (material and labor) consistent with blade quality Until the advent of numerically controlled machining techniques, and the development of multi-spindle copying machines, milling and planing were the major metal cutting processes used for blade vane production. The application of multi-spindle, multi-axis copying machines has allowed far more complex designs to be specified and produced. This has also reduced the per-element cost to the extent that complex vortex profiles can now be produced at competitive prices. This change has allowed the turbine steam path to achieve higher levels of efficiency at reasonable costs. However, there are many turbines still in service that utilize constant section 734 Leibstritz The Manufacture and Inspection of Rotating Blades Fig. 12.4.5—Precision forged rotating blades being removed from the heat treating furnace. blades, and these operate at levels of efficiency that are acceptable in most respects, and which to redesign the form of the vane is uneconomical. In addition, there are stages where the cost of the vortex design cannot be justified. Consider the normal methods of manufacturing the various forms of blade: Cylindrical profiles (vanes of constant cross section). There are various metal cutting techniques used to produce blades with a vane of constant profile; the techniques used for any particular stage depend upon the blade vane length. The selected techniques are also influenced by the stress levels in the blade as determined by the design process. 735 Turbine Steam Path Maintenance and Repair—Volume Two An early form of vane was one in which the pressure face was produced, normally by plunge milling from the tip to the root on a piece of bar stock. Here the vane form is shown in Figure 12.4.6, where the milling cutter is passed radially along the pressure face, the milling cutter being profiled to achieve the correct form of the pressure face. The major disadvantage with this form of design is that the center of gravity of the vane is not coincident with, or near the center of gravity of the root. Therefore, for large radial height blades, high centrifugal bending stresses can be set up in both the vane and root. As a result, this method (while economical in terms of manufacturing costs) was suited only to the smaller blades with lower calculated levels of stress. Fig. 12.4.6—Plunge milling along the axial length of a piece of bar stock to produce one face of the cylindrical profile of the blade vane. Another method of producing the pressure face was to mill across the vane in the direction of the width, as shown in Figure 12.4.7(a), and the suction face as shown in Figure 12.4.7(b). In each case the cutter center locus is shown as “T-T.” Under these circumstances it was normal to position the vane on the root as near coincident with the root platform center of gravity as possible, as shown in Figure 12.4.8, for a rectangular shaped platform, where the vane center of gravity “Gv” and the center of gravity of the root “Gr” are close to coincident. Here the root platform center was at the center of the rectangle at the position “W1,” “T1,” and the vane center of gravity was not coincident by the amount “dx,” “dy.” 736 The Manufacture and Inspection of Rotating Blades Fig. 12.4.7—Milling a cylindrical profile. In (a) cuts are made on the pressure face, and in (b) on the suction surface. T1 T2 W1 dx Gr Gv W2 Gr dy Gv Fig. 12.4.8—Vane placement of the “C of G” from the root platform “C of G”. There were often limitations to the removal of material from the pressure face by across width milling if the radius of curvature of the pressure face was too small. In these cases it was not possible to produce a milling cutter of sufficiently small radius that it could be mounted onto the spindle of the milling machine. In such situations it was necessary to pass along the radial length of the vane, and then make a final cut at the root section to finish the blending radius. 737 Turbine Steam Path Maintenance and Repair—Volume Two Many early designs employed blades that were mounted to the rotor using a vane and spacer piece placed between them. This was a cost-effective method, as the vane material could be produced by milling, or from extruded stock, cut to length, and then possibly a tenon produced on the tip. The required taper in the root portion was machined into the spacer block, as shown in Figure 12.4.3. The spacer block was also required to have the correct form on its two tangential faces to match the shape of the vane. In each case of milling the constant profile suction face in which an integral root (rather than a spacer piece) was used, the milling cutter had to be passed in the width direction across the blade material, in a path that would form the required surface contour. This process is shown in Figure 12.4.7(b). Costs associated with the production of this type of constant profile blade escalate considerably if it is required to produce an angled sidewall at either the inner or outer surface, shown as “α2” and “α1” in Figure 12.4.9, or if it was required that an integral coverband R S S α1 T T α2 R Fig. 12.4.9—A rotating blade with an integral coverband and angled sidewalls at “α1” outer and “α2” inner. 738 The Manufacture and Inspection of Rotating Blades should be produced on the blade, at the outer surface with an angled side wall “α1.” In these cases, it required a separate set-up of the machine tools to remove the material adjacent to the walls, shown as triangles “RST” in Figure 12.4.9, and to produce the remaining vane portion and the inner and outer surface fillet radius. It would also require some hand polish finishing, and radius blending to complete the vane. Fig. 12.4.10—Gang milling a tenon. The milling techniques used with this form of blade were sufficiently controlled and the vanes did not require extensive polishing after the vanes were complete. The only possible exception was the tapered sidewalls, and their point of joining either the root platform or the integral coverband. With this type of profile it was normal to produce a tenon by passing a gang of milling cutters across the tip of the blade, as shown in Figure 12.4.10. Here the milling gang is shown in (a) and the resulting tenon form in (b). An isometric of the 739 Turbine Steam Path Maintenance and Repair—Volume Two tenon is shown in (c). Either one or two tenons could be produced in this manner. The one disadvantage with the production of tenons by gang miller cutting was that there was no platform around the tenon, as is achieved with more modern machining techniques. General Electric Long blades are often produced by cutting the vane in the radial direction, as seen in Figure 12.4.11. This method of production will use either planing or milling cutters to remove material from the envelope forgings. A master is followed for the shape of the vane and can achieve close tolerances. The root fillet radii are produced as a separate operation (shown in Fig. 12.4.12). The blades are finally hand polished to close tolerances (see Fig. 12.4.13). These methods are able to produce high quality blades conforming closely to vortex requirements. However, this complex method of production could only be justified for long exhaust stage blades where the performance of the individual stages was critical to the performance of the turbine. Fig. 12.4.11—Machining the vane portion of long exhaust blades. The cutting direction is radial. 740 General Electric The Manufacture and Inspection of Rotating Blades General Electric Fig. 12.4.12—Machining the vane/root platform fillet radius. Fig. 12.4.13—The final hand polishing operation on a large exhaust stage blade. 741 Turbine Steam Path Maintenance and Repair—Volume Two Vortex profiles (vanes of twisted section). For earlier generation units to produce vanes that accord with the vortex requirements of a row, required the use of compromise profiles constructed from arcs of circles and straight lines. Because of machining limitations, i.e., the need to produce the vanes by milling and other similar procedures, the true vortex requirements could not be met. However, close approximation could be achieved by a combination of axial length milling and planing. But, because of costs, this method of manufacture could only be justified for the longer blades, and other methods such as the “straight generated” profile were used to approximate vortex requirements for other shorter stages. The principle of producing the “straight generated” vane is shown in Figure 12.4.14, where the material billet is placed on the milling machine table, inclined at an angle “ψ,” and then the milling cutter is passed over this block on the line “N-N,” using a different part of the cutter at each radial position, and therefore producing a varying profile, along the radial height. The introduction of the multi-spindle profile copying machine, shown in Figure 12.4.15, made the production of vortex profiles on blades of varying section at economical costs a reality. In this figure, five blades are being produced, following the master profile on the left. Modern high efficiency turbines almost exclusively utilize blades manufactured by such methods if there is enough change of Root Mean Tip T M ψ N R N T M Root Mean Tip Dr Profile Milling Cutter Dt Dm R Fig. 12.4.14—Passing a formed milling cutter over bar stock material at an angle “ψ” to produce the straight generated face on the material. 742 General Electric The Manufacture and Inspection of Rotating Blades Fig. 12.4.15—The multi-axis milling machine producing a number of vortex vanes copied from a master on the far left. section to warrant such manufacture. In some manufacturing facilities blades of even relatively short radial height are produced by this method, since to use a single manufacturing technique has eliminated the need for several production lines, and the cost differences are sufficiently low that with the volume involved the twisted vane can be economically justified. The cutting process develops forces on the vane, which are not (because of the amount of material being removed) equal. Figure 12.4.16 shows a plot of these forces as a function of the position being cut. When the profile cutting is complete, the surface is then given a final polish. This surface requires hand polishing to establish the design specified finish. In fact only a small amount of material is required to be removed, and this can be accomplished relatively easily. The blade is polished to the form of a “guillotine” or “shutter” gauge as shown in Figure 2.13.1 of chapter 2. 743 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.4.16—Showing the variation of cutting force around the vane with profile copy milling. The forging process The forging process uses mechanical pressure to work a hot billet into a form that is either a final form, requiring no material be removed from the vane portion (precision forged), or it requires a small amount of material removal to produce the final form (envelope forging). In either case the process requires the use of dies to produce the final form. Figure 12.4.17 shows a large blade being removed from the press. The dies can be seen on the bed and upper press of the forge. In the forging process, it is normal to undertake some preforming, by hammering to an approximate form, where the initial billet is being extended and partially formed prior to forging. After preforming, the billets are heated to the forging temperature and then pressed into their final correct form. After forming the blade, it is heat treated to achieve the required mechanical properties and relieve stress (see Fig. 12.4.18). 744 Leibstritz The Manufacture and Inspection of Rotating Blades Leibstritz Fig. 12.4.17—A precision forged blade being removed from the press at completion of the forming process. Fig. 12.4.18—Precision forged blades at completion of their initial stress relief. 745 Turbine Steam Path Maintenance and Repair—Volume Two Leibstritz After the initial cooling of the precision forging, the “flash” is removed, as shown in Figure 12.4.19, and the blade is heat-treated. In Figure 12.4.20 the forgings are being loaded into a vacuum furnace for stress relief. Leibstritz Fig. 12.4.19—Removal of the ‘flash’ after forging. Fig. 12.4.20—Blades placed in the vacuum furnace for stress relief. 746 The Manufacture and Inspection of Rotating Blades The extrusion process The extrusion process is not used now to produce high performance blades for large rating units. However, the process can still be used for smaller, industrial type turbines where the costs available from selling these turbines does not support the cost of high performance blades, and where the engineering design exists. In these cases, blades are produced as extruded stock, the extruded pieces having been formed by being drawn through a die that has the airfoil form. Therefore, these blades are of constant section, and require a spacer piece between them to achieve the required pitch, as shown in Figure 12.4.3. The vane pieces are cut to a length that is the sum of the root depth, the vane radial height, and the height of any tenons that must be produced integrally. Pinch rolling Pinch rolling is a forming process, which utilizes two rollers between which is forced a heated billet. The rollers have produced on them the profile of the suction and pressure faces. As the billet is pulled between the rollers, it is formed into the required airfoil form. This process is shown in Figure 12.4.21, where the rollers in vice slots “E-E” capture the root block, and the billet is pulled into the rollers to start the forming process. Once the rolling has started, the remainder of the billet is pulled through, forming the profile. The centerline of a typical profile is shown in the detail, with the centerlines from the root position “R-R” to the tip “T-T” stacked above the center of gravity “G.” The pinch rolling process is shown in Figure 12.4.22. The rollers normally require lubrication, and there is judgment required to ensure the billet does not slip once rolling has started, as this will cause deformation of the profile form. This process, like forging, will often employ an initial hammer preforming to make the rolling easier. The rolling is also sometimes undertaken in two stages, 747 Leibstritz Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.4.21—Pinch rolling billets into profile stock. Fig. 12.4.22—The concept of ‘pinch rolling’, where a preheated billet is pulled between rollers, one with the pressure face profile, and the other with the suction face profile. 748 The Manufacture and Inspection of Rotating Blades with a final rolling undertaken to achieve a higher quality surface finish. Again this is an older process, and while it can produce highquality blades, it is tending to be replaced by more modern, and ultimately more cost effective methods. At the vane inlet and discharge positions, the rollers do not meet, and there can be a certain amount of “flash,” which will require dressing after the blade has cooled. Those processes that form the blade vane using heat and pressure will normally result in a vane with a degree of twist or warpage. The vane will therefore require straightening after cooling. This straightening will be completed after the blade has been formed at its inlet and discharge edges, and has cooled, but before the root is machined. It is normal to adjust the vane using cold bend and twist methods, using a guillotine gauge to ensure the vane is both of the correct form, and adjusted to the correct setting angle. This is to ensure the expansion passage formed between the vanes is of the correct form. Depending upon the material and the extent of the correction required, the blade should be stress relieved after adjustment to remove the residual stresses that result from plastic deformation. PROFILE AND CASCADE TOLERANCES The vane profile is fundamental to stage performance (efficiency and reliability), and therefore its form must be controlled within close tolerances throughout the manufacturing cycle. During the initial manufacture, repair, and replacement of turbine steam path components there are certain spatial relationships in the axial, tangential and radial directions that must be met, and others that (by preference) should be met to assist in optimizing the performance of the unit. The spatial requirements of what must be achieved in terms 749 Turbine Steam Path Maintenance and Repair—Volume Two of alignment are discussed in chapter 2. To achieve these alignment criteria, it is necessary that the individual components, particularly the blades, are produced within design limits to ensure that adequate adjustment of the components on final assembly is possible. The individual profiles, and their correct position relative to each other are essential to achieve acceptable stage efficiency. During the initial production phase, the turbine supplier will manufacture and assemble the component parts of the steam path to ensure their design requirements are achieved within specified tolerances. When a unit is removed from service for a maintenance outage, it is often necessary to perform some remedial work to correct damage and any deterioration that is found. These remedial actions should aim to return the components, and their arrangement, to as close to the original conditions as possible, consistent with preserving the performance of the unit. Therefore, it is necessary for plant maintenance staff to have sufficient knowledge of the component parts and their assembled requirements, so they can determine the most appropriate course of action in any repair situation. Dimensional conformance during the production phase, for both metal shaping and component assembly, is essential to the satisfactory performance of the unit. Similarly, when repairs are undertaken, safeguards should be employed to retain this dimensional conformance. The turbine stationary and rotating blades are designed, manufactured, and assembled, so they are able to interact with other blades, both within their own row, and with rows that precede and follow them. Engineering tolerances are selected by the unit designer of the component to ensure performance requirements are met and, where appropriate, components can be disassembled for repair or replacement. Engineering tolerances should also be selected so components can be interchanged, both within similar units in a station, and between stations. This engineering requirement also allows for a minimum of replacement parts to be carried in inventory. 750 The Manufacture and Inspection of Rotating Blades Tolerances must be evaluated and defined during the design phase to first achieve the requirements of assembly, and secondly they must support the performance requirements of the unit. It is not, from the manufacturer’s perspective, a good policy to make tolerances tighter than necessary or required to achieve these objectives. To make tolerances tighter adds to the cost of production, and would do nothing to improve the quality of the product. Blade profile and cascade quality tolerances For a blade row to operate with a maximum efficiency, and for those stresses induced in each blade to be shared and equal, it is necessary to achieve to concurrent objectives: • The individual blades, both stationary and rotating, must be identical, or as close as possible within the tolerances defined by design engineering • The individual blades within the rows must be spaced so that the passages formed between them are consistent and form passages of the type defined by design engineering Each of these requirements is fundamental to blade quality, and the manufacturing process selected to meet these requirements must be suitable, and able to meet requirements consistently. The vane profile is fundamental to stage performance. Therefore, its form, finish, and the tolerances applied to its manufacture must be set and controlled within limits, throughout the entire manufacturing cycle, enough to ensure it meets design requirements. While the requirements of the individual profiles and the vanes they form either as a constant section along the radial height, or the rate at which they modify from root to tip, are of considerable importance in establishing the efficiency of the row, the more critical considerations are those associated with the form of the expansion pas- 751 Turbine Steam Path Maintenance and Repair—Volume Two sage formed between them. The requirements of the cascade, and the form of errors were discussed in chapter 2, and will not be considered further. These requirements of positional tolerances, etc. are applicable to both stationary and rotating rows. Blade profile definitions and tolerances Before considering the requirements of the profile, it is necessary to establish a nomenclature to be used in discussing it. Figure 12.5.1 shows a single profile that is labeled to indicate its principal characteristics. A primary concern of any blade is the form of the profiles. If the profile has any of a number of possible errors, it is unlikely it can fulfill its function entirely satisfactorily. There are certain dimensional characteristics of the profile that help to define its quality and ability to meet performance requirements. While these were considered in chapter 2, these requirements are summarized here for completeness. The more significant of these being: Inlet nose β10 Inlet edge Pressure face B Suction face T ξ θο C b Discharge edge β20 Discharge tail Fig. 12.5.1—The nomenclature and definition of a profile. 752 The Manufacture and Inspection of Rotating Blades Inlet nose. The “inlet nose” exists at the point of steam entry to the blade row. This nose is at the inlet, which is the juncture of two surfaces, and defines the form of the inlet to the expansion passage. If the profile has a design specified inlet nose radius “r,” this radius must be maintained; it must also blend without discontinuity to both the pressure or suction faces. The inlet nose of the profile must be shaped to accept the working fluid over a small range of inlet angles while incurring only a minimum energy loss. This nose must guide, or divert the flow into the two passages it helps to form. This must be done without causing excessive flow disturbance, turbulence, or separation of the boundary layer. The suitability of a profile to work over a small range of steam inlet angles is necessary because steam angles could experience minor change of direction under certain conditions, changes in steam properties or quantities, or distortion or damage to the discharge area of the previous row. If there is a distortion of area for steam flow in the previous row, this will influence the pressure at inlet to that row, modifying the enthalpy drop across it, and therefore causing a change in the discharge velocity, which will alter the required inlet angle in the row, into which it is directing its steam flow. The discharge nose (or tail). The discharge nose, as shown in Figure 12.5.1, has a defined thickness “b,” which for the rotating elements must be maintained at the design values, as this represents a region on the profile where stresses both direct and cyclic, are high during operation. It is particularly necessary near the root section, where an undersized discharge tail thickness can be a source of crack initiation in the event of any form of mechanical damage. On stationary blade profiles it is desirable to keep this tail as thin as practicable, to minimize the generation of “wakes.” However, this must be done within the bounds of the bending stresses that are present. For these stationary profiles, this tail region of the profile is 753 Turbine Steam Path Maintenance and Repair—Volume Two that portion that experiences material loss if there is solid-particle erosion. Therefore, there is some justification for making these thicker (increased “b”) in those stages. Discharge tail. The “discharge tail,” or discharge portion of the profile, defines the shape of the throat and controls the exhaust area of the stage. Discharge tail curvature is a characteristic of the profile that is carefully considered by the designer, and the radius of curvature maintained at a large value, which in conjunction with the tail thickness, is used to minimize the width of the “wake region” downstream of the passage discharge. The metal section inlet angle (“β1o”). Figure 12.5.1 shows a vane section in which a mean inlet angle is shown as “β1o.” This angle is a function of the profile skeleton line, and is not influenced by steam flow angles, but is established by the shape of the profile and the angle at which it is mounted on the blade platform, at the setting angle “ξ.” The metal section discharge angle (“β2o”). The metal section discharge angle is shown as “β2o” in Figure 12.5.1, and is the angle between the tangent to the skeleton line at the discharge point on the tail, and a tangential line in the direction of rotation when the profile is at a setting angle “ξ.” The metal section turning angle (“θo”). The metal section has a turning angle “θo,” which in terms of the metal section inlet and discharge angles “β1o” and “β2o” provide a metal section turning angle “θo.” θo = 180° - (β1o + β2o)° The radius of curvature of the profile surfaces. Pressure and suction surface radii of curvature should not change abruptly. On the suction face losses result from sudden reductions in the radius of curvature, as these will promote separation, thereby introducing losses, since once detached the boundary layer will not reattach. Similarly, 754 The Manufacture and Inspection of Rotating Blades on the pressure surface a sudden increase in the radius of curvature could induce losses. For further discussion see chapter 2. Pressure and suction faces. The profile must be formed to produce the pressure face of one steam passage, and a suction face in the adjacent one. It must do this, and at the same time be able to form a passage so the desired degree of reaction or pressure drop occurs. This is achieved by controlling the throat at any radial location, and therefore the discharge area over the entire discharge of the blade row. The “pressure face” is the concave face against which the steam exerts a positive pressure in being deflected through its turning angle “θo.” The “suction face” is the convex face of the profile. Profile chord and thickness. These are the two major characteristics that establish the mechanical strength of the profile, helping establish its ability to carry load and its natural frequency. In Figure 12.5.1 the maximum thickness is shown as “T,” and the chord as “C” (not shown). There are different definitions of profile chord. Profile sectional area and bending modulus. The profile must possess sufficient mechanical strength to be able to withstand the forces and loads that are developed on, and within it. These include those which are the direct, and due to the mass of the vane, and the mass of other stage elements it must support. The profile must also be able to withstand alternating bending stress from vibratory loads and various stimuli induced in the blade row during operation. The blade must continue to operate under these loads, and the stresses that are induced during normal and transient operation. These characteristics are used to define the profile. However, for purposes of manufacture the designer provides an “envelope of tolerances,” as shown in chapter 2, Figure 2.13.2; within this envelope, the profile is considered fully acceptable. 755 Turbine Steam Path Maintenance and Repair—Volume Two Blade cascade definitions and tolerances It is clear from the previous discussion that producing a profile within tolerances is not in itself a guarantee the stage will perform as anticipated. The profile position in relation to adjacent elements must also be examined. The more important considerations affecting cascade acceptability are: Blade pitch, “P.” The pitch “P” between any two profiles (see Fig. 12.5.2) is a function of the stage diameter “D” and the number of blades “Zb” in the row, and can be determined from: Inlet edge β1 θ W ξ O β2 Discharge edge P Fig. 12.5.2—The definition of a bladeFigure cascade. 12.5.2 The definition of a blade cascade. Throat opening “O.” The throat or opening “O” is formed usually at the discharge point of the flow passage (see Fig. 12.5.2), and is the minimum distance between the pressure and suction faces of adjacent profiles. The throat is fundamental to determining the discharge area for any expansion passage, and in total establishes the discharge area 756 The Manufacture and Inspection of Rotating Blades from the row, and therefore the pressure at discharge and the enthalpy drop across the row. For passages of mean throat “Oe,” radial height “H,” and with “Zb” blade elements in the row, the individual throat area is “a” and the total row discharge area is “A,” both defined by the following equation: The steam inlet angle, “β1.” The required steam relative inlet angle “β1,” is determined from the velocity triangles, and is dependent upon the steam and blade velocities. It also requires a blade profile that will admit the steam without excessive incidence. When selecting the profile to use in any stage application, the design process selects, or designs, a profile that has a metal section angle most nearly suited to meet the requirements of “β1.” The inlet nose has, for the more modern profiles, a rounded form, helping the steam to enter the blade passage without undue shock and flow disruption. The steam discharge angle, “β2.” The discharge angle “β2” is a function of the shape of the profile tail, and is affected by the ratio of throat opening to pitch, both of which parameters will vary along the length of the blade vane. Therefore, it is normal for the discharge angle to vary also. Manufacturing tolerances must be applied to this value, and the ratio of throat opening “O,” to pitch “P” to achieve this. The influence of the discharge tail shape can be seen in Figure 12.5.3. Here the throat “O” is formed on the discharge tail, which has a degree of curvature. The throat, or minimum opening, occurs across “g-g.” At that position on the tail, the steam, if it were to separate and flow at the same angle would deviate from the discharge angle of the tail, at its discharge point by an angle “Γ.” Another significant angle at the discharge point is the metal section angle “β2o,” which is set at this angle relative to the tangential direction. 757 Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.5.3—Details of the discharge tail. As discussed in chapter 7, the effective design angle “β2” is: An acceptable angle of deviation “Γ” is dependent upon the designer. This angle is kept as small as possible as its thickness will cause an increase in the “wake” thickness, and introduce efficiency losses in the flow. The larger this angle “Γ” becomes, the greater the chance of boundary layer separation from the discharge edge, and the formation of vortices that could be carried forward into the following row. Many manufacturers attempt to minimize this angle, and set a limit of seven degrees. In doing so, they attempt to maintain a maximum value at the mean diameter. However, it is possible this value will be exceeded at the tip section of constant profile blades, where the pitch has increased, and the throat will have moved to, and be formed on, a smaller radius portion of the profile. The steam turning angle “θ.” The profiles must be able to divert the working fluid through the desired turning angle “θ,” and in changing its momentum have a net thrust or force developed. This force must be able to be transmitted to the rotor and cause it to rotate. An equation was provided earlier for the turning angle of the 758 The Manufacture and Inspection of Rotating Blades metal section “θo.” However, since both the steam inlet angles, due to the possibility of incidence, and the discharge angle, due to the form of the discharge tail will be different, the turning angle of the steam “θ” is given by: θ = 180° - (β1 + β2)° For an impulse stage the value of “θ” is usually in the range 110160 degrees. As the degree of reaction increases, this turning angle reduces, so that for the long last stage blades with a considerable variation of the degree of reaction along the profile, the turning angle requirement is changing also. For a large blade, the turning angle can change from about 160 degrees at the root to 10-20 degrees at the tip section. Therefore there is a considerable range of shapes of profiles that can be encountered in any turbine, or stage of a turbine. For the higher-pressure stage of a unit designed for about 50% reaction in the high and intermediate pressure section, the turning angles will be in the range 60-80 degrees. The profile setting angle, “ξ.” The profile setting angle “ξ,” shown in Figure 12.5.2, is fundamental to the shape of the expansion passage. By being correctly set, it helps ensure the expansion passage is the correct form, and the discharge area of the passage, and therefore the discharge flow from it, is at or near the design value. The expansion passage form. The profile must ensure the passage formed between each pair of surfaces, from adjacent profiles, forms a passage that is of the form required by the pressure ratio across the row. The manufacturer will apply a level of tolerances. With this level being at its extremes, reconvergence should not occur between adjacent elements. (Reconvergence occurs when after convergence, there is a change to divergence in the passage and then a further convergence.) In the pair of profiles shown in Figure 2.14.17, of chapter 2, the passage effective width decreases from “Oe” at entry to the row, to 759 Turbine Steam Path Maintenance and Repair—Volume Two “Oe” at discharge. This shows a normal rate of convergence for a passage designed for subsonic flow. Inlet and discharge edges. The inlet and discharge edge define the theoretical extremities of the profile at “inlet to” and “discharge from” the stage (see Fig. 12.5.2). The manufacturer must specify the extent to which a blade can be “proud of” or “recessed from,” the inlet or discharge edge. For larger blade elements, this value (acceptable at any radial height) could be a function of the blade radial position, or distance from the attachment (root) end of the profile. Cascade width, “W.” This major dimensional characteristics of the vane width “W” are dependent for any profile upon the setting angle. When a vane has been defined as one suitable for application within the steam path, it is not normal for the setting angle “ξ” to be changed. However, there are profiles that can be used at various setting angles, and these have been checked to establish that the shape of the expansion passage is acceptable at each. It is normal for the manufacturer to have established tolerances for each of these characteristics above, and then to monitor the manufacturing and assembly processes to ensure they are achieved. In blade manufacture and assembly, the existence of a nonconforming condition indicates either the dimensional requirements, as outlined in this section, have not been achieved, or there is evidence of structural distress in the form of component distortion. In establishing a maintenance strategy, it is advisable to have available established tolerances to which the components should be returned by any repair/refurbishment procedure that is used. These tolerances should be observed and monitored by the maintenance engineer during the corrective procedures that are selected as the result of an evaluation of the initial nonconforming condition. 760 The Manufacture and Inspection of Rotating Blades Gauging the profile Because the profile is so fundamental to performance, there is a need to be able to gauge the final form to ensure it falls within the design-specified envelope of tolerances. There are various methods available for making these checks. These include: The guillotine gauge. The most common method of gauging the compliance of a blade vane is to use a guillotine gauge. Such a gauge, shown as Chapter 2 Figure 2.13.1, locates the blade in a radial direction from its root, and then a series of shutters are offered to the pressure and suction faces of the vane. These gauges have stops, which allow them to travel to the correct position. Figure 12.5.4 shows a blade in a guillotine gauge before the shutters are closed to measure for dimensional conformance. Leibstritz These gauges are normally produced so that with the guillotines in their fully closed position, there is a gap between the knife-edge of the gauge to both sides of the profile, which is equal to the sum of the plus (+) and minus (-) tolerances, and then a “go/no-go” gauge is used to establish if the profile tolerances have been met. Fig. 12.5.4—The vane prior to profile gauging. 761 Turbine Steam Path Maintenance and Repair—Volume Two Projection methods. Another method of gauging the profile is to measure and project the form onto a screen. To undertake this method, it may be necessary to remove a slice of blade from a vane. This shadowgraph, at a suitable magnification, is then transferred to a sheet of paper or vellum for comparison with the design requirements. Westinghouse Electric Eye lashing. Eye lashing is a method of reproducing the vane profile at any radial location on a piece of paper at a suitable magnification. This consists of preparing a series of arcs from the vane using a follower to monitor the form of the vane. The process of “eye lashing” the output of such an operation is shown in Figure 12.5.5, and the output of such a procedure is shown in Figure 12.5.6 This drawing is then compared with design requirements to establish if compliance exists. This method has been essentially superseded by computer methods. Fig. 12.5.5—Eyelashing a profile. 762 The Manufacture and Inspection of Rotating Blades Fig. 12.5.6—The output of an eyelash examination of a vane section. Computer traces. Modern computer techniques allow a blade profile to be gauged by coordinate measuring machines, as shown in Figure 12.5.7. This method allows a profile to be gauged in three axes, and then compare the manufactured form with the design specification. This can be a relatively slow process, but is of considerable use in setting up and quantifying an initial cut. As shown in Figure 12.5.8, the computer can also provide a plot of the profile shape and its conformance with the design specification. 763 Next Page Turbine Steam Path Maintenance and Repair—Volume Two Fig. 12.5.7—A three axis measuring machine being used to check a blade. Fig. 12.5.8—The output from a computer measurement of a profile showing the deviation from the design profile. This is a graphic comparison between a master and an in-production blade. The actual data are within the tolerance band of 10µm (0.004"). 764 List of Acronyms LIST OF ACRONYMS AA AISI ASME BHN BWR CLA Dcsp EDT EOH EPRI ESV F FEA FofS FOMIS GTAW HAZ HCF LCF HP HP/IP I&TP LCF LF LP NDE NDT NOH NPF OEM ppb psi QA QC RMS SCC SHR SMAW SPE SPI T-G TIG TTD UTS UT arithmetic average American Iron and Steel Institute American Society of Mechanical Engineers Brinell hardness number boiling water reaction centerline average direct current straight polarity enthalpy drop test equivalent operating hours Electric Power Research Institute emergency stop valve Farenheit finite element analysis factor of safety Fossil Operations and Maintenance Information System gas tungsten arc welding heat-affected zone high-cycle fatigue low-cycle fatigue high pressure high pressure / intermediate pressure inspection and test plan low-cycle fatigue load factor low pressure nondestructive examination nondestructive testing normal operating hours nozzle passing frequency original equipment manufacturer parts per billion pounds per square inch quality assurance quality control root mean square stress corrosion cracking station heat rate shielded metal arc welding solid-particle erosion solid-partical impact turbine generator tungsten inert gas terminal temperature difference ultimate tensile stress ultrasonic testing ix Appendix Thermodynamics and the Mollier Enthalpy-Entropy Diagram for Water/Steam INTRODUCTION Thermodynamics is a science that considers those relationships existing between thermal energy and work, and the conversion from one form to another. It does this by considering and evaluating the behavior of gasses and vapors, and their physical variations under the action of changes in environmental pressure and temperature. In engineering terms gas is a substance that has completely evaporated from the liquid state, and is dry, i.e., gas contains none of the liquid phase substance. By definition a vapor is a gas that contains 815 Turbine Steam Path Maintenance and Repair—Volume Two some of the liquid phase in suspension; and as such will not obey the laws of gases, because with the changes in the thermodynamic properties defining the gas phase, some proportion of the liquid phase will change state and cause a flow of heat from the liquid to gaseous phases and vise versa. Therefore vapors obey different laws defining their change of state. This chapter covers the principles required to apply to steam and a steam/water mixture as it is expanded in the steam turbine. Water is a naturally occurring substance. It exists on, below, and above the surface of the earth in vast quantities, and is readily handled. It is also a substance very suited for use in a power cycle, and contains many properties making it suitable for application within these different cycle configurations. Water/steam will accept and reject heat in a manner making it a suitable fluid for such application. At all levels of thermodynamic conditions it is non-toxic, and does not break down easily under the actions of pressure and temperature. However, at high pressures, and temperatures above 1,200°F steam will dissociate into hydrogen and oxygen, and can even support combustion within the steam path. The study of the flow, or expansion of steam is complicated by the condensation process, which exists within many fluids as they give up their thermal energy. This process occurs in the turbine steam path, as the steam expands through the blade rows to produce mechanical power and portions of the working fluid are converted back to the liquid phase. Steam, when heated so that no liquid phase particles are present, can be considered for all practical purposes to behave as a perfect gas, and the perfect gas laws can be applied to its expansion when it has been heated so that no moisture particles remain. However, once water is formed within this steam, the two-phase flow behaves in a somewhat different manner, and it becomes necessary to account for these differences during the design process. This is because of the 816 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam expansion and transportation of the moisture particles that are entrained within the steam, and the effects this water will have upon the steam path components once it is deposited upon, and flows across their internal surfaces. In any discussion of the steam turbine and the need for its maintenance and care, it is essential to recognize that it is a thermal machine, designed specifically for energy conversion from one form to another, i.e., steam from a fossil or nuclear boiler is released into the turbine at high levels of potential energy, and this energy causes the turbine rotor to rotate and either generate electrical power in a generator, or drive some piece of mechanical equipment. The internals of the unit are arranged so the energy of the steam is released in a controlled manner that maximizes the efficiency of the energy conversion process, and therefore produces a maximum of power from the minimum amount of steam energy. For power generation installations, the driven machine is a generator that produces electric power, which can then be utilized in a number of different applications. However, even within power generating facilities there can be other smaller units used for driving feed pumps. These are variable speed units and will require certain considerations that are similar to the main constant speed units, but present other conditions that must be evaluated. The process of designing the turbine is complex. The engineer responsible for this work is required to select various blade and other elements for the unit so that while they can be arranged to optimize the energy conversion process, the elements can be arranged and aligned, so the conversion process is as efficient as possible, and the unit will operate with a high degree of reliability. The amount of coal, oil, gas, or other fuel consumed per unit of electrical power generated must be kept to a minimum. The design process therefore will utilize thermodynamic and aerodynamic principles in order to ensure that the total energy conversion is achieved as economically as possible. 817 Turbine Steam Path Maintenance and Repair—Volume Two THE PHYSICAL PROPERTIES OF WATER/STEAM There are certain properties or physical characteristics that define the energy level of steam and/or a mixture of water as it exists within the turbine. It is important to be aware of their significance, as they are fundamental to the design process. There are other properties that define important characteristics, as they will apply to steam flowing through a turbine steam path. Pressure—”P or p” Pressure is a measure of the force that is exerted on any surface by the kinetic action of the steam against that surface. In imperial units this pressure is measured in pounds per square inch (psi). An important pressure is that of 1 standard atmosphere, which is the pressure exerted by the atmosphere on a surface at sea level. This pressure is equal to 14.7 (14.696) psi. It is also convenient to relate this atmospheric pressure to the column of mercury (in a Bourdon gauge). This means of representing pressure is that 14.7 psi will support a column of mercury, which is 30 (29.9) inches at an ambient temperature of 62°F. For all practical purposes, it is sufficiently accurate to convert these quantities, and say that 1 psi is equal to 2 inches of mercury, which is a convenient means of defining sub-atmospheric pressure in the power cycle. Specific volume—”Vs or v” The specific volume of steam is a measure of the volume that will be occupied by a given weight of the steam at the local environmental conditions. In imperial units this is in cubic feet per pound (cu ft/lb). 818 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam The specific volumes of water and saturated steam are determined by experimentation. These results can then be reduced to empirical formulae, which allow their computation for all values of pressure and temperature. The normal units of specific density are lb/cu ft, and density is equal to the reciprocal of specific volume. For water and saturated steam the specific volume is a known quantity at any pressure and temperature level. For partially evaporated water, or partially condensed steam, the total volume can be found from the sum of the volume of the liquid (water) and the gaseous phases at that pressure. If the mixture has a dryness fraction “q,” or moisture content of “x.” q = 1-x then the total volume of the mixture “Vm” can be found from: Vm = q. Vs + (1 - q) . Vw or Vm = (1 - x) . Vs + x . Vw where: Vw is the specific volume of 1 lb of water Vs is the specific volume of 1 lb of dry saturated steam It can be seen that at “q = 1.0” the water term disappears, and the specific volume is then equal to “Vs.” In fact, for all practical engineering purposes the water term “(1 - q). Vw” can be ignored, as the specific volume of water is particularly small compared to the saturated steam term, so its inclusion will add very little to the accuracy of the value determined in the form of steam engines encountered. Figure A2.1 shows a curve of the specific volume of dry, saturated steam from 1,100 psia to 100 psia shown as a function of saturation pressure and temperature. A similar curve for water covering the same pressure range is shown in Figure A2.2. The pressure and temperature values are the saturation values. Similar curves for a pressure range from 100 psia to 10 psia are shown in Figures A2.3 and A2.4. 819 Turbine Steam Path Maintenance and Repair—Volume Two Fig. A2.1—The specific volume of steam from 100 psia to 1,100 psia. Fig. A2.2—The specific volume of water from 100 psia to 1,100 psia. 820 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Fig. A2.3—The specific volume of steam from 10 psia to 100 psia. Fig. A2.4—The specific volume of steam from 10 psia to 100 psia. 821 Turbine Steam Path Maintenance and Repair—Volume Two For superheated steam specifically, the best method of establishing an approximate value of the specific volume for any particular degree of superheat and at a pressure “p” is to apply the general gas equation. where: p Vs Ts Vt Tt is the pressure of the steam is the specific volume at saturated conditions is the saturation temperature is the specific volume after superheating is the superheated final temperature At constant pressure, the previous equation reduces to: This is a convenient and reasonably accurate method of computing these specific volumes. Temperature—”T” Steam temperature is a measure of the degree of hotness or energy level of the steam. In imperial units this is measured in degrees Fahrenheit (F). In order to quantify the degree of heat, it is necessary to establish accurate and repeatable scales to which comparisons with bodies of unknown temperature can be compared. Therefore, a temperature scale can be constructed by establishing these two limits, one upper and one lower. This range is known as the fundamental interval. This fundamental interval is then divided into a number of equal subdivisions, each division then being known as one degree. 822 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam In engineering work there are two basic scales available and in current use: • Centigrade scale—this scale takes as its upper and lower intervals the boiling and freezing point of water, when measured at a pressure of one standard atmosphere. This scale is divided into 100 subdivisions or degrees • Fahrenheit scale—this scale establishes its lower and upper limits as the freezing point of alcohol, and the temperature of pig’s blood, both at a pressure of one standard atmosphere. This scale is divided into 100 subdivisions or degrees. A more general, and convenient definition of the Fahrenheit scale is that the fundamental interval between the freezing and boiling point of water is divided into 180 degrees, with the freezing point being set at 32°F Because of the difference in the fundamental intervals, these two scales are not equal, and 100 degrees Centigrade is equal to 180 Fahrenheit degrees. In addition the zero point of the Centigrade scale is equivalent to the 32-degree point on the Fahrenheit scale. Therefore: Degree F = (Degrees C x 180/100) + 32 Degree C = (Degrees F - 32) x 100/180 Absolute temperatures. Considering a gas that conforms to Charles Law, the equation for volumetric change for change in temperature is: Vt = Vo [1+ αt] where: Vt is the gas volume at temperature “t” Vo is the gas volume at absolute zero temperature αt is the coefficient of increase in volume at constant pressure 823 Turbine Steam Path Maintenance and Repair—Volume Two The implication of this equation is that as temperature reduces, the specific volume decreases until at absolute temperature the volume of the mass of gas is zero. It also implies that the lowest temperature possible is the absolute temperature, which is equal to -273.12 degrees Centigrade, or -460 degrees Fahrenheit. Temperatures quoted in the absolute scale are normally referred to as the Kelvin scale in degrees Centigrade, and are designated as degrees “K,” or degrees Rankine “R” in the Fahrenheit scale. The internal energy—”E” The internal energy of a substance is the total energy stored in that substance by virtue of its thermal and other energy forms. It is impossible to quantify the total internal energy level in any gas or vapor. However, from engineering considerations the significance of internal energy is the change in energy levels that occurs as a consequence of changes to the substance as a result of its giving out or taking in energy, and doing or having work done upon it. Should a substance take in energy and do no work, then its level of internal energy will increase. Similarly if a substance does work and takes in no thermal energy, then this work is done at the expense of its internal energy, and its level will decrease. The internal energy of a given quantity of gas is dependent upon its temperature only, and unless this gas does work in expanding, or takes in or gives out heat, its internal energy will remain unchanged. This statement is normally referred to as Joules Law. This internal energy represents the energy stored in the gas, and can be made available to perform external work. The relationships existing between thermal energy, work, and internal energy are discussed below: 824 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam The enthalpy of a Gas—”H” Enthalpy is a relatively simple concept used to define the total energy of a substance including gasses and vapors, as they exist at some specific set of thermal conditions. The total heat (enthalpy) of any substance is the sum of the internal energy “E” and the pressure energy that exists as a consequence of the potential to do work or change state. The quantity enthalpy is normally designated by the symbol “H.” “J” is the Joules mechanical equivalent of work. In imperial units, the equivalent of “J” is 1 Btu is equal to 778 ft lbs of work. When a substance is expanded at constant pressure “p,” and there are temperature reductions from “T1” to “T2,” its condition or internal energy is reduced from “E1” to “E2” by the removal of a quantity of heat. If during this expansion the volume of the gas changes from “V1” to “V2,” then the work done “W = p (V2-V1),” which is equal to the quantity of heat removed. Heat removed = E2 – E1 + p(V2-V1) This equation indicates that the quality of heat removed is equal to the pressure work done by the gas in expanding. If the quantity of heat converted to work during the expansion is given the symbol “∆H,” then: That is: Giving: 825 Turbine Steam Path Maintenance and Repair—Volume Two This equation indicates that for any expansion, the work done (neglecting losses that might occur as a consequence of the expansion) is equal to the change in total heat or enthalpy of the substance from initial to final conditions. This heat energy is quite distinct from the temperature. Heat is a measure of the energy that is contained in the steam. The unit of heat energy is a measure of the amount of heat in any given weight quantity. In imperial units this is British thermal unit per pound (Btu/lb) where a British thermal unit is the amount of heat energy required to raise one pound of water one Fahrenheit degree. The total heat in a body is a quantitative measure of the amount of thermal energy contained in that body. There are various definitions of heat used in engineering work. Among these, the most commonly used in engineering work are the following: • Calorie (Cal)—the calorie is the amount of heat required to raise one gram (gm) of water one Centigrade degree at one standard atmosphere • British thermal unit (Btu)—The Btu is the amount of heat required to raise one pound of water one Fahrenheit degree at one standard atmosphere. A normal manner of seeing this unit quoted in relation to the discussions of power plants is in Btu/pound, or abbreviated to Btu/lb, indicating that each pound of steam at any location within the cycle contains a certain number of Btus of thermal energy • Centigrade heat unit (CHU)—this unit has now become largely discarded. It is the amount of heat required to raise one pound of water one Centigrade degree. This unit will not be considered further In fact, in these units the amount of heat required is also dependent upon the initial temperature. An international agreement has been reached in defining both the calorie and the Btu. For the calo- 826 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam rie, it is defined as the amount of heat required to raise 1 gram of water from a temperature of 14.5 to 15.5°C. This is known as the 15°C calorie. Similarly, the Btu is defined as the amount of heat required to raise one pound of water from a temperature of 59.5 to 60.5°F. The total heat of steam is the sum of those heats required to raise its temperature or internal energy to the condition at which it exists. The most convenient manner of defining total heat is to establish the amount of thermal energy above some arbitrary level. The level chosen for engineering work is the point at which one pound of ice has been converted to one pound of water at a temperature of 32°F (492°K). Therefore, this total heat or enthalpy of the steam includes the sensible heat required to heat the water to 212°F, the latent heat of evaporation required to convert the water to dry saturated steam, plus any superheat used to raise the temperature of the steam to a condition above the saturation point. These heat quantities all assume the pressure at which the heat is transferred is held constant at all times. Therefore, the enthalpy “H” of steam at some pressure “p” can be found from: H = h + L + Hs where: H h L Hs is the total heat of the steam is the sensible heat required to convert 0°F water to boiling water at pressure “p” is the latent heat required to evaporate the water to dry saturated steam is the superheat The total heat “H” of saturated steam at some pressure “p” can be found from: 827 Turbine Steam Path Maintenance and Repair—Volume Two H=h+L For a quantity of water in which only a portion of the total latent heat has been added, and therefore is only partially evaporated, the total heat is less than that given by the above equation. If only a percent of the total latent heat required to evaporate the water has been supplied, then the total heat of the two-phase mixture will be: H=h+q.L H = h + (1-x) . L where “q” is the dryness fraction, and is defined as the ratio weight of the dry steam contained in the mixture to the total weight of the mixture. If “Ws” is the weight of the dry steam, and “Ww” is the weight of the water, then: In most work related to the steam turbine, interest is usually related to the amount of water that has been formed in the steam as a consequence of its expansion through the steam path, and the reduced pressure at various locations that result from this expansion. Such expansion is normally from an initial superheat condition, and its conversion of thermal potential energy to mechanical work. It is common to refer to and define this presence of water as the moisture content of the working fluid and the presence of water is defined and given the symbol “x” representing a percentage content. Therefore, if 10% of the saturated steam has condensed, and exists either in the steam as transported droplets, or has been deposited upon the internal surfaces of the steam path components, then the steam is said to have a moisture content of 10%, and the dryness fraction is 90%. 828 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam The viscosity of water and steam The viscosity of any fluid is a measure of the shearing stresses that exist within that fluid when it is flowing. These stresses will be present in the fluid whenever there is a difference in velocity from one stream element to another. Such stresses will also reduce in magnitude as the differences in velocity diminish. Consider the viscous flow velocity profile shown in Figure A2.5. Here the velocity of the fluid (gas or liquid) increases from zero or stationary at the solid wall to some mean velocity. This velocity is constant at a value of “C” along any stream tube. The stream filaments are shown as lines parallel to the “x” axis. The derivative “dV/dy” is a measure of the rate of change of velocity at a distance from the solid surface at which the velocity is “0.” y Velocity profile x Fig. A2.5—Viscous flow velocity profile, adjacent to a surface. The shearing stress “τ” that exists from element to element can be found from: Where “µ” is defined as the coefficient of viscosity. The values of kinematic viscosity of water and steam are shown in Figure A.2.6. 829 Turbine Steam Path Maintenance and Repair—Volume Two Fig. A2.56—The ‘kinematic viscosity’ of steam and of water. 830 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam THE GAS EQUATIONS Superheated steam can be considered a prefect gas, and one that obeys the laws of such gases. There are also certain physical parameters or characteristics of any gas that must be understood. These are well defined in engineering terms: Boyles law Boyles law states that the volume of a fixed quantity of gas, maintained at a constant temperature, depends on the pressure to which it is subjected. What this means is that for a given quantity of gas (or superheated steam), the pressure it exerts on its surrounding surfaces is dependent upon the volume that the gas occupies if the temperature remains constant. Pressure multiplied by volume is a constant, or: P x Vs = k where: P is the steam pressure Vs is the gas volume k is a constant Charles law Charles law states that for a fixed quantity of gas heated at a constant pressure, the volume increases by a constant fraction of the volume at 0º for each degree increase in temperature. This means that for a given quantity of gas (or superheated steam), the volume occupied at constant pressure is inversely proportional to the temperature. Volume divided by temperature is a constant, or: 831 Turbine Steam Path Maintenance and Repair—Volume Two V/T = k where: V T k is the gas volume is the gas temperature is a constant The general gas equation The equations of Boyle and Charles can be combined into a single equation known as the general gas equation. This equation relates these properties of pressure, temperature, and specific volume, and can be applied to a gas, and to steam when it is in the superheated condition. If the suffix “1” applies to an initial condition, and suffix “2” applies to a second set of conditions at completion of some thermal process, then the properties of the gas or steam before and after their change of state, assuming at both conditions the gases are still a perfect gas, are related by the equation: This equation can also be written in the general form: or: P .V = m . R . T The mass of gas employed is “m,” and the constant “R” is known as the general gas constant, and will remain the same for all conditions. Note: In applying this general gas equation to any set of conditions, the units must be compatible and the temperature used must be in the absolute scale. 832 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Normal temperature and pressure The properties of gases as they are considered to exist in a standard atmosphere are always given at standard conditions that in Fahrenheit units are found from substituting into the General Gas Equation for one pound of gas: Temperature Pressure Specific volume 32 + 460 = 492 degrees R 14.669 psia 12.39 cu ft/lb Therefore, applying the general gas equation to one lb of steam at atmospheric pressure gives: THE HEATING AND EXPANSION OF STEAM As heat is added to or removed from steam there is a change in its physical characteristics. It is necessary to consider these energy levels changes, and how they will influence the resulting steam conditions at the end of the process. There are a number of interchange processes that can affect the steam turbine, and it is necessary to develop an understanding of each. The addition of heat to water/steam In order to understand the factors that affect the design and dimensioning of the steam path, and also those factors that influence the unit after it goes into operation, it is necessary to have an under- 833 Turbine Steam Path Maintenance and Repair—Volume Two standing of the working fluid, which is either steam or a mixture of steam and water. The chemical compound H2O can exist in three basic forms or states—solid, liquid, and gaseous. The solid phase is commonly known as ice, the liquid phase as water, and the gaseous phase as steam. This compound will readily change from one of these states to the other by the addition or removal of heat. The amount of heat required to achieve these state changes will vary, and is dependent upon the pressure at which the heat transfer occurs. The addition of heat and the change of states. If a block of one pound of ice is at, or near absolute zero, and subjected to atmospheric pressure that is heated, it will expand by some small amount under the influence of this heat addition. The internal temperature of this ice will also increase. This process of heating, expansion, and temperature rise can continue until a temperature of 32°F is reached. Up to this temperature the ice will remain in the solid form. Then with the addition of a small amount of heat, the ice will begin to convert to water at the same temperature. If further heat is added the ice will melt, converting completely to the liquid phase. During this addition of heat there will be no increase in the temperature of the ice/water, which will remain at 32°F (492°K). The amount of heat required to complete this phase transformation from the solid to liquid states is termed the latent heat of fusion, and the quantity of heat required per pound of ice to complete this phase change is dependent upon the pressure at which the process is being undertaken. The formation of steam (heat addition to water). If one pound of water, which was obtained from the addition of the heat of fusion to the one pound of ice, existing at a temperature of 32°F and a pressure of one atmosphere continues to receive heat, its temperature will again begin to rise. Therefore, the temperature of the liquid phase substance (water) will increase, but the state or form of the 834 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam water will not change. The heat being added is called sensible heat. Its effect can be measured and quantified by measurement of the liquid phase temperature. Again a point will be reached when the temperature has risen at atmospheric pressure, by 180°F to 212°F (672°K). At this point a portion of the substance will be driven off from the surface of the water as steam. From this heat, or energy level, the addition of further heat will not increase the water temperature, but will continue to convert the water (liquid phase) to the gaseous phase steam, at a constant temperature of 212°F. The heat that is added to the water, increasing its energy level but introducing no temperature change, is known as the latent heat of evaporation. This latent heat is defined as the amount of heat required to evaporate the water completely forming one pound of steam at the same temperature and pressure. If the pressure at which the heat is added changes, so will the quantity of heat required to complete this phase change. The latent heat of evaporation. The quantity of heat required to completely evaporate the one pound of water and convert it to one pound of steam at the same temperature and pressure is known as the latent heat of evaporation. The quantity of heat required to evaporate this one pound of water is again dependent upon the pressure at which the substances exist, which is the pressure of evaporation. When the one pound of water has been completely evaporated, and before any further addition of heat to increase its temperature, the steam is said to be saturated. Saturated steam is by definition dry, i.e., no moisture exists within it. Moisture content and dryness fraction. An important parameter of water, particularly when applied to heat engines such as the steam turbine, is the dryness fraction, or moisture content of the working 835 Turbine Steam Path Maintenance and Repair—Volume Two fluid. When water is evaporated, it is conceivable that until high levels of evaporation have occurred, there will be moisture present in the two distinctly recognizable states of water and steam (liquid and gaseous). However, when steam expands from a higher to a lower pressure, giving up its internal energy, the temperature falls until a point is reached when the enthalpy of the steam is such that moisture must exist in the working fluid. Under these circumstances, water droplets are normally formed by nucleation, and exist within the steam, and are transported by it through the steam path. As the expansion continues and the internal energy level of the two-phase flow reduces, the steam (gaseous phase) gives up further energy, and a further portion of the working fluid condenses and converts to the water (liquid phase). For the engineer a definition of the moisture content is essential as this helps define what portion of the latent heat exists in the steam and is available for conversion to some other form of energy. Superheated steam. Once the ice has been converted to dry saturated steam, if heat is continued to be added to the one pound of saturated steam, this further heat cannot change the state, but will cause an increase in the temperature of the steam. The steam is then defined as superheated. The degree of superheat is dependent upon the temperature to which the steam is raised. Again, the amount of heat required to achieve a certain degree of superheat is dependent upon the pressure at which the steam exists. The conversion of ice to superheated steam. The various processes and the effect on phase changes for the heating of ice to superheated steam have been discussed earlier. These processes are shown diagrammatically in Figure A4.1, where the changes of state and the increases in temperature are shown. The addition of heat from “A” to “B” raises the temperature of ice to the point where the 836 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam further addition of heat from “B” to “C” will cause the ice to melt forming the liquid phase (water). This water will be at the same temperature as the ice at condition “B,” i.e., the latent heat of fusion has been added causing no increase in temperature. Fig. A4.1—The effect on phase changes and conditions of adding heat to a quantity of water. From “C” to “D” the sensible heat has been added to the water, raising its temperature to the point where the addition of further heat will cause evaporation. From “D” to “E” the water is fully evaporated converting the liquid/gaseous phase to dry saturated steam at “E,” during which process there has been no temperature increase in either of the phases. This represents the addition of the latent heat of evaporation. The addition of further heat from “E” to “F” will cause the steam to become superheated. The heating and expansion relationships If steam is heated and expands at constant pressure “P,” the volume of the gas will increase from “V1” to “V2” during this heating process, then the work done “W” by the steam in expanding is: W = P . (V2 - V1) 837 Turbine Steam Path Maintenance and Repair—Volume Two However, if the pressure does not remain constant during the heating process, but changes, then the work done by the expanding steam can be found from: From this definition it can be seen the work done by the steam during such an expansion, or done on it as it is being compressed, is represented by the area bounded by the “p-v” expansion curve, and the upper and lower pressure limits, i.e., the area “ABCD” of the curve shown as Figure A4.2. This form of diagram is commonly known as an indicator diagram. Fig. A4.2—The ‘indicator diagram’. If the expansion of the steam is at constant temperature, it is an isothermal expansion, then “p1.V1 =p2.V2,” and the work done in expanding is equal to: 838 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam The specific heats of steam The specific heat of steam is defined as the amount of heat energy required to raise the temperature of unit weight of the substance through a temperature rise of one degree. The heat that is supplied to the steam is used to increase its intrinsic energy. When heat is added to steam and it expands, the work that is done is given by the previous equations. However, the amount of work that is done is subject to considerable variation, depending upon the amount of expansion that occurs, and the consequent change in pressure levels. In the case of an ideal gas (including superheated steam), it is most appropriate to consider two specific heats for different modes of parameter change. These are those occurring when the gas is heated at constant pressure, “Cp” and those when the gas is heated at constant volume, “Cv.” Consider a weight of steam “m” and that its temperature is raised by the addition of heat from some external source by an amount “dT;” the heat supplied to it, if its volume is maintained constant is “m.Cv.dT,” and no work is done. Then if this same steam is heated at constant pressure, heat equal to “p.δv/J” is added where “δv” is the increase in volume due to the addition of the heat at constant pressure “P,” causing a further temperature increase of “dT,” and the heat supplied is equal to “m.Cp.dT.” Therefore, a consequence of this combined heating process in two stages is: 839 Turbine Steam Path Maintenance and Repair—Volume Two This reduces to: But P.V = m. R.T, therefore: Therefore: The heating of steam Another method of considering this relationship that exists between the specific heats of steam at constant pressure, and then at constant volume can be developed as follows. Consider steam with initial conditions of “p, V1, and T1” and with final conditions after heating at constant pressure to “p, V2, and T2.” If one pound of this gas is heated at this constant pressure the work done can be found from: W = P.(V1 - V2) But P.V2 = R.T2, and P.V1 = R.T1 Therefore, substituting these values in the equation for W gives W = R (T2 - T1) = R/J[T2 - T1] thermal units Also H = Cp [T2 – T1] And E = Cv [T2 – T1) 840 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam From the law of conservation of energy the following general equation can be stated: Total Heat Supplied = External Work Done + Increase in Internal Energy That is: H = W + E From which it can be seen that In its more familiar form: The ratio of the specific heats When steam is heated at constant volume it does no work as there is no expansion, and therefore no change in volume. With this form of heating the specific heat is known as the specific heat at constant volume, and is given the symbol “Cv.” When gas is heated at constant pressure it will increase in volume, there is expansion, and therefore, work will be done. This method of heating gives a higher value of the specific heat at constant pressure “Cp.” The specific heat of a gas at constant pressure will be numerically greater than the specific heat at constant volume. This difference is the consequence of the gas having to do work on the internal pressure in expanding. The ratio of these specific heats is an important consideration in the study of thermodynamic processes and is given the symbol “γ.” 841 Turbine Steam Path Maintenance and Repair—Volume Two The expansion of steam Steam enters the turbine at a high pressure and temperature, and expands through the steam path to some lower set of conditions. It is normal for this to occur in a number of series stages, arranged along the axial length of the unit. In order to examine the requirements of the steam path it is necessary to consider certain thermal concepts. Using the general gas equation, curves can be drawn for the variation of steam condition as the steam expands giving up thermal potential energy; both the pressure and temperature reduce with such expansion. The thermodynamic properties accompanying an adiabatic expansion are shown in Figure A4.3, which illustrates the change in pressure, temperature, and specific volume from inlet to discharge conditions. These curves are plotted to a base of total heat, the initial steam condition being 900 psia, and 900°F and the exhaust pressure is 2.0" Hga (1psia). At a pressure of about 78 psia, and an enthalpy of about 1,183 Btu/lb, the total superheat has been expended and the conditions pass into the saturated region. Moisture then begins to form in the parent steam. A curve of moisture content for continued expansion to the exhaust pressure is shown. The specific volume curve is corrected for the reduction in effective steam quantity. From these curves it can be seen that with this expansion the pressure and temperature fall relatively even. However, if the condition curve for the specific volume is examined, there is a dramatic increase in the volume as the steam approaches exhaust conditions. From these curves, consider one pound of steam (1 lb) at a pressure of 900 psia and a temperature of 900°F. This 1 lb will occupy a volume of about 0.85 cubic feet. Then after this 1 lb of steam has expanded through the steam path to a condenser pressure of 1 psia (2.0" Hga), the 1 lb will now occupy a volume of 334.1 cu ft/lb. Had there been no condensation, and with about 20% converted to moisture particles, this volume of steam would be about 267 cu ft/lb. At this pressure the specific volume of water is 0.0161 cu ft/lb. 842 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Fig. A4.3—The variation of steam conditions as a function of enthalpy for an adiabatic expansion. However, the moisture content at completion of this adiabatic expansion is at about 20%, giving it a dryness fraction of 0.790. Therefore, the total volume (assuming none of the moisture has been deposited on the sidewalls and steam path components) would be Vm, given by: Vm = 334 x 0.80 + 0.0161 x (1.00 - 0.20) = 267.40 cu ft/lb From these numbers it can be seen that the steam phase occupies 267 cu ft/lb, and the water portion 0.01 cu ft/lb. This demonstrates that there is (in most cases) sufficient justification in ignoring the contribution of the water to the total volume of the mixture. In Figure A4.3 an adiabatic expansion that is at constant entropy was assumed. If steam at this same condition, but with an expansion efficiency in the range 80% is assumed, then the curves in Figure A4.4 843 Turbine Steam Path Maintenance and Repair—Volume Two could be drawn for the expansion of the working fluid in the turbine steam path—these conditions recognizing the inefficiencies, or losses, which occur in the steam path expansion. Again, from these curves it can be seen for this non-reheat unit that the pressure falls relatively evenly as does the temperature. However, if the condition curve for the specific volume is examined, it can be seen there is again the dramatic increase in volume as the steam nears exhaust conditions. Fig. A4.4—The variation of steam conditions as a function of enthalpy for an 80% efficient expansion. This increase in volumetric flow in the lower pressure regions of an expansion accounts for the dramatic increase in blade length of a steam turbine as the steam expands from stage to stage. It also emphasizes the need to often provide multiple flows in the low-pressure section, so there can be made available sufficient annulus area 844 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam to allow the steam to exit to the condenser at realistic velocities. If instead of 2.0" Hg the condenser had produced an exhaust pressure of 1.0" Hg, then the steam volume would have been 656 cu ft/lb. From these volumetric values it can be seen that at the exhaust from the low-pressure section, the volume flow of the steam is very sensitive to pressure. In fact, throughout the low-pressure sections relatively small changes in steam pressure will have a significant impact on the volume flow, and therefore exhaust velocities. The work done by steam in expanding Steam can expand from a higher to a lower pressure in a variety of ways with different combinations or variations in the defining parameters of pressure, volume, and temperature. The amount of work done during these different expansions depends upon the variation of the heat content of the steam from initiation to completion of the expansion. The expansion of superheated steam follows closely the laws of a perfect gas. However, once moisture begins to form, there is a need to modify the predictions of the final conditions that will be achieved to account for the effects of this moisture. Expansion at constant volume. The case of steam expanding at constant volume is the same as for a perfect gas, but in fact has little or no application in heat engines, as the purpose of an engine is to do work, and for any constant volume expansion the work done will be zero. Expansion at constant pressure. Heat engines convert thermal potential energy to work by allowing the pressure and enthalpy level of the steam to degrade in a controlled manner sufficient to generate mechanical work. For this reason the constant pressure expansion has little or no application in heat engines. 845 Turbine Steam Path Maintenance and Repair—Volume Two General case, an expansion according to the law, pVn = C (the polytropic process). Consider the expansion of steam from condition “p1,V1” to conditions “p2,V2,” as shown in Figure A4.5. The total work done “W” during this expansion is equal to the area under the “pv curve” ABCD, from the following equation: But P . Vn = P1 . V1n, and Therefore: Which can be simplified to: 1 Pressure - ’p’ A p.v n = C 2 D B Volume - ’v’ C Fig. A4.5—The indicator diagram for an expansion according to the law pvn = C. 846 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam An adiabatic expansion, (pVγ = constant). An adiabatic expansion is one that is carried out reversibly, and in which no heat is allowed to enter or leave the system during the process, i.e., at completion of expansion (or compression) the total heat of the gas is unchanged. This implies there is no loss in the gas due to turbulence, eddies, or frictional heating (no internal energy losses), and as a result, the process must be carried out slowly. Therefore, any work done during an adiabatic expansion is done at the expense of the internal energy of the gas. Similarly, during an adiabatic compression all the work that is done in compressing the gas is converted to internal energy within that gas. Applying the law of conservation of energy to an adiabatic expansion, in which no heat is accepted to or rejected from that process, gives: But dH = 0 Therefore, for the adiabatic expansion For the adiabatic expansion of a perfect gas therefore: In a perfect gas “dE = Cv.dT,” therefore the expression for the adiabatic expansion of a perfect gas can be written: But P = RTV 847 Turbine Steam Path Maintenance and Repair—Volume Two Therefore : Which on integration gives: Substituting “Cp - Cv = R/J,” and dividing by “Cv” gives: Writing: γ = Cp/Cv γ Loge V - Loge V + Loge T = A constant But: P . V/T = A constant Therefore: Loge P + Loge V - Loge T = A constant Adding these last two equations gives: Loge P + γ . Loge V = A constant or P . Vγ = A constant Therefore, the expression for the work done in an adiabatic expansion is similar to the general expansion considered previously, except that the index of expansion is defined as “γ,” which is equal to “Cp/Cv.” The change of temperature during an adiabatic expansion or compression can be found from consideration of the two relationships: 848 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Combining these two equations gives: Therefore, the work done “W” in an adiabatic expansion can be shown to be equal to: An isothermal expansion (at constant temperature) An isothermal expansion is one that occurs at constant temperature, and during expansion the gas satisfies the relationship: P.V = R.T The work done in such an expansion again can be found from: In this equation “T” is the temperature at which the expansion takes place. 849 Turbine Steam Path Maintenance and Repair—Volume Two During an isothermal expansion or compression there is no change in the internal energy of the gas since there is no change in temperature, (the internal energy of the gas is a function only of temperature). Therefore, the heat taken in or given out during an isothermal process is equal to the work done or expended. Throttling. Throttling or wire drawing is the name given to any process in which steam passes, or escapes through a constricted opening, causing a reduction in pressure. In this case, any energy expended by the movement or expansion of the steam is converted to velocity energy, which is reconverted to heat energy as the velocity is destroyed. Consider the leakage through a small gap between two surfaces shown as Figure A4.6. Here the steam is accelerated through the opening, and achieves a velocity “Ac” at discharge. Small clearance Ea A pa Vsa Eb pb Cx Vsb B Fig. A4.6—The small gap for throttling. Figure 6A4. The small gap for throttlin g If the steam conditions on the higher pressure side “A” are “pa” and “Vsa,” and the steam has an internal energy “Ea,” and on the lower pressure side “B,” the steam conditions have changed to “pb,” and “Vsb” with internal energy to “Eb.” Now work is done on the steam in forcing it to flow from side “A” to side “B,” and this work is equal to “pa.Vsa/J;” after the steam has passed through the constriction it does work on the steam on that side by an amount equal to “pb.Vsb/J.” 850 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Equating the energy on the two sides, after the velocity energy has reconverted to heat energy gives: Because the sum of the work “P.Vs” term and the internal energy at any point is equal to the enthalpy, then: Ha = Hb and therefore the work done in a throttling expansion “W = 0.” Metastable conditions during expansion (supersaturation) In discussing the expansion of a gas, it has been assumed that, at every condition during this expansion, equilibrium conditions are maintained, and that as conditions are reached at which moisture is to form, there is an immediate heat transfer from the gaseous to the liquid phase and that moisture particles appear, and are suspended in and then transported by the steam as it continues its expansion. In fact, this heat transfer will take a finite period of time to occur, and there is therefore a delay in the formation of these moisture particles or droplets. Under these conditions the rapidly expanding steam becomes temporarily supercooled or supersaturated. For steam expanding in a turbine within the saturated region, the total expansion (because of the velocity with which the steam is moving through the blade rows) is always unstable, and there is some degree of delay in transferring heat from the gaseous to the liquid phase. Therefore, there is always some supersaturation present when steam is expanding in the saturated region. 851 Turbine Steam Path Maintenance and Repair—Volume Two When in the supersaturated state, the density of the vapor is higher than the equilibrium density of the saturated vapor at the local environmental pressure. Therefore, this supersaturated condition is not stable, i.e., it cannot correct itself and disappear completely as soon as there has been time for the heat transfer from the gaseous to liquid phase to occur, because during this time there has been a further expansion and a further portion of the vapor has not had time to condense. There is, in fact, a limit to the degree of supersaturation that can be achieved in any expansion of steam. The degree of supersaturation depends upon a number of factors, including the chemical purity of the steam itself. THE ENTROPY OF STEAM An important thermodynamic concept or property in understanding and quantifying thermodynamic processes and energy conversion is entropy. Entropy is best, and most simply defined as a relationship that exists between heat and temperature. To provide a suitable definition of entropy, consider a quantity of steam that is heated reversibly, or simply has heat added to it without its doing any work. By definition, in adding a small quantity of thermal heat “δH” to a quantity of steam, when its temperature is “T,” will cause an increase in its entropy of “δH/T.” Starting from some suitable zero point if a quantity of heat “δH” is added to a substance at a temperature “T,” then the change in entropy “δs” is: 852 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam If any substance is subjected to changes in state in a reversible manner from condition “1” to condition “2” then the change in entropy associated with this change is: Therefore: It is of value to consider the level of entropy changes that occur when steam is changing its basic conditions. The following examples illustrate these entropy level changes: Entropy changes for a reversible cycle in which the temperature is continually changing. Shown in Figure A5.1 is the “p-V” diagram for a quantity of steam cycling through a reversible process, in which the temperature is continually changing, and returning or recycling to the original condition. To determine the entropy change during each complete cycle, consider this cycle to consist of a large number of individual, and small, Carnot cycles, each individual cycle comprising two adiabatic and two isothermal processes. Each adiabatic shown (“a-a,” “b-b,” “c-c” etc.) and the isothermals (shown as the small lines between the adiabatic expansions and compressions at the upper and lower conditions “a-b,” “b-c,” “c-d,” etc.) is traversed twice, once in each direction, canceling each other. It can be seen that in following the elementary cycles that these represent, in total, the change of conditions in the base cycle. If “dH” is the amount of heat transferred in an elementary cycle, at temperature “T,” it can be seen that the entropy of each individual elementary cycle is “dH/T,” and for the complete set of elementary cycles: 853 Pressure Turbine Steam Path Maintenance and Repair—Volume Two a c e a b e c d f b f d Volume Fig. A5.1—A reversible process with constantly changing temperature. Therefore, it can be seen that for any cycle comprising only reversible processes in which the original conditions are repeated, the change in entropy “ds” is zero. Entropy changes with an adiabatic and isothermal expansion The adiabatic expansion occurs without change in the heat content of the steam. Therefore, “dH = 0.” dh = 0 Therefore: ds = 0 During an adiabatic expansion or compression there is no change in the entropy level of the gas, and the expansion is “isen- 854 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam tropic.” An isentropic expansion occurs at constant temperature “T.” On a temperature-entropy diagram this expansion is represented as a horizontal line. Entropy changes when expanding a gas at constant volume When a gas is heated and maintained at a constant volume, having an initial temperature “T1,” and a final temperature “T2,” then the quantity of heat “H” added can be found from: H = Cv . (T2 - T1) For a small change in the total heat “dH” there will be a change in entropy of “Cv.dT.” Therefore, dH/T = Cv.(dT/T), and ds = Cv. (dT/T). If during this change of state the temperature changes from “T1” to “T2” then: Giving: Entropy changes when expanding a gas at constant pressure When a gas is heated at constant pressure, and has an initial temperature “T1,” and a final temperature “T2,” then the quantity of heat “H” added can be found from: H = Cp.(T2 – T1) 855 Turbine Steam Path Maintenance and Repair—Volume Two The change in entropy when heating at constant pressure can be found in the same manner as used for constant volume. The equation of state reduces to: dT ds = Cp . T Which integrating between the temperature limits “T1” and “T2”gives: Reducing to: T2 s2 - s1 = Cp .loge T1 The cases of steam expanding at constant volume and constant pressure have no practical application in the steam turbine, as all expansions within the steam path occur with a change of all three states—pressure, volume, and temperature. Therefore, it is necessary to consider the application of all three property changes. General expression for change in entropy When steam is heated, or allowed to expand in the steam turbine, there will be a change in the physical conditions, changing them from an initial state of “P1,” “V1,” and “T1” to a final state of “P2,” “V2,” and “T2.” Therefore, in the general case the change of steam conditions can be found from the general equation “H = W + E” represented by: dH = dW + dE = T.ds = P.dV/J + m . Cv . dT By recognizing that steam obeys the law “P.V = R . m . T,” and that “Cv” remains sensibly constant, the processes can be considered reversible (see next section). Therefore, these variables can be 856 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam eliminated in turn, and the changes in entropy can be determined in terms of “V,T,” “T,P,” and “V,P.” For 1 pound of gas, that is “m = 1.0.” By differentiating and substituting “dT = (P.dV + V.dP)/R” REVERSIBILITY The concept of reversibility is that the change of state that occurs during a thermodynamic process of expanding or contracting can be reversed. During reversal the substance will pass back through all the phases (conditions) that it passed through during its expansion or 857 Turbine Steam Path Maintenance and Repair—Volume Two contraction, and at completion of the reversal process the gases will have the same final state they started with. The implication of reversibility is that the expansion or compression must occur without any energy being lost due to friction, and without causing (or introducing) any aerodynamic losses. For reversibility to exist, the process must normally be slow. During such a process the gas must remain in mechanical and thermodynamic equilibrium with its surroundings. The heat engine. A heat engine is designed to convert thermal potential energy to work. This conversion is achieved by lowering the energy level of the supplied fluid, and converting the extracted energy. To achieve this conversion the inlet energy level is raised to as high a level as is economically achievable by the combustion of fuel, or the disintegration of nuclear energy. The lower level is achieved using whatever methods are available. In the majority of steam applications, this is by the use of a condenser that produces a sub-atmospheric pressure to increase the range of available energy. The efficiency of the heat engine is defined as: Work Done Energy Used Work Done Engine Efficiency = Energy Supplied - Energy Rejected Engine Efficiency = Reversibility of heat engines. For an engine to be reversible, each process in its operating cycle must be reversible, which is obviously not possible, as the combustion of fuel is itself an irreversible process. However, the comparison of a cycle or piece of equipment to one that is reversible (such as the Carnot cycle) provides a measure against which other engines and cycles can be compared. 858 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam STEAM PROPERTIES AND DIAGRAMMATIC REPRESENTATION There are a number of diagrams or charts that can be constructed showing the relationships existing between various parameters, defining the thermodynamic properties of water and steam. Two of these, the temperature-entropy (T-s) and the total heat-entropy (H-s) are the most useful, and therefore the ones most frequently used in engineering work. However, others can be just as important in the design process, and are referred to by the engineer to ensure the best possible results and accuracy are obtained from their calculations. These two most frequently used diagrams are capable of allowing accurate predictions to be made of the results of many processes, and will allow a diagnosis of many situations that can occur within the cycle and the components that it comprises. The steam tables The steam tables are a listing of the various properties of steam at a range of conditions. They were determined by experimental formulations established from a considerable amount of research into the individual properties and their variation under certain conditions. The physical properties of water and steam are determined experimentally, and have been published as steam tables. There has been a history of the production of these tables; each generation represents a greater accuracy based on improved experimentation, as testing equipment and instrumentation becomes more accurate. Associated with this experimentation is the determination or definition of formulae that describe these steam parameters with greater accuracy for any specific set of conditions. The most recent, and 859 Turbine Steam Path Maintenance and Repair—Volume Two those referred to in this work, are those produced by the American Society of Mechanical Engineers, published in 1967. Table A7.1 shows a portion of the steam property tables. Table A7.1—A portion of the Steam Tables, (ref. A.1) showing the properties of saturated steam and water 860 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Fig. A7.1—The “temperature - entropy’ diagram for steam/water. The temperature-entropy (T-s) diagram This diagram shows the relationship between the gaseous and liquid phases of water and their variation for different temperatures 861 Turbine Steam Path Maintenance and Repair—Volume Two to a base of entropy. Locus curves are constructed to represent the variation or change of entropy from water to steam, from the partially wet to saturated and the superheated condition. The outline of this diagram for water/steam (from ASME 1967 tables) is shown in Figure A7.1. On this diagram are certain characteristics that define the relationships existing between the properties of water, a water/steam mixture, saturated and superheated steam. Consider the following characteristics as drawn on the diagrammatic representation of the “T-s” diagram in Figure A7.2: c m1 m2 T Pq q u r t v n1 n2 a d s Fig. A7.2—The ‘T-s’ diagram for water/steam. The saturated water line. The saturated water line defines the entropy at the boiling point of water at various pressure levels. If the entropy of water at its boiling point, when all the “sensible heat” has been added, were plotted for a number of pressures, a line known as the saturated water line would be obtained. This saturated water line is shown in Figure A7.2, as existing from points “a” to “c.” This water has an enthalpy content equivalent to condition “D” of Figure A4.1. The water contains the total 862 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam heat required to convert the ice from absolute zero condition “A” of Figure A4.1 to water at its boiling point condition “D.” However, as previously stated, for engineering work the reference (or point of zero enthalpy) is shown as energy level “C” of Figure A4.1. The saturated steam line. Similarly, if the locus of dry saturated steam were constructed for a number of pressures, the line “c” to “d” would be obtained. This line then defines the entropy level of the steam at completion of the addition of latent heat, sufficient to convert all the water to dry saturated steam. This is represented by condition “E” of Figure A4.1, and condition “r” of Figure A7.2 for a steam pressure “Pq.” The latent heat. From the discussion of the saturated water and saturated steam lines, it is clear the horizontal distance between these two curves represents the change in entropy that occurs as the latent heat of evaporation is added to the water. This is represented by the heat addition from “D” to “E” of Figure A4.1, and “q” to “r” of Figure A7.2. Consider the line “q - r” on Figure A7.2. This line represents the heating of water from condition “q” to condition “r,” and the evaporative effect, which occurs during this process. It is clear the addition of latent heat at constant pressure represents or provides no change in the temperature of the mixture. The critical point. From an examination of the “T-s” diagram of Figure A7.2, it is clear there is a convergence of the water and steam lines, and that these would meet at some condition of temperature, pressure, and entropy. This meeting or convergence point is shown as “c.” At this condition it would not require the addition of any latent heat to convert the saturated water to saturated steam. The conditions at which this “no latent heat requirement” occurs, is known as the critical point. At this critical point the steam 863 Turbine Steam Path Maintenance and Repair—Volume Two has a pressure of 3,208 psia, and the corresponding temperature is 705.4°F. At all fluid pressures lower than the critical pressure the fluid can exist as a liquid. Above this pressure the fluid can exist in the gaseous form only. If water exists below the critical temperature and has a higher pressure than the saturation pressure (corresponding to that temperature), it is considered to be compressed or sub-cooled water. The moisture content. In considering the addition of heat at a constant pressure, raising the enthalpy levels from a condition at some pressure “Pq” from the saturated water condition “q” to dry saturated steam condition “r” (Fig. A7.2), the entropy level was increased and the water was fully evaporated. The increase in enthalpy due to the addition of heat is the distance “q - r.” When only a portion of this latent heat had been added, the entropy level will have been raised to some intermediate condition represented by point “t.” At this condition, represented by point “t,” the fluid is a two-phase mixture containing (q - t)% dry saturated steam, and (t - r)% water. This quantity still requires evaporation to be converted to dry, saturated steam by the addition of sufficient heat to increase the entropy level by an amount “t - r.” Therefore, the moisture content “x” of any partially evaporated mixture can be found from the ratio: (q - t) x = (q - r) . 100 % If a constant moisture level were established for different pressure levels, then the lines connecting this constant content at various pressure levels would look like the line “c-t-v” in Figure A7.2. The lines of constant superheat. At completion of the evaporative process at constant pressure, if further heat is added (again at 864 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam constant pressure), the entropy level will increase. Also, since the internal energy of the gas will increase by the addition of this heat, the gas temperature will increase also. This temperature rise is known as the process of adding superheat. If a curve is drawn from point “r,” at pressure “Pq” in Figure A7.2, the consequential change in entropy and temperature (heating at this constant pressure), is the locus “r - u.” This line indicates the line of superheating at a constant pressure. If lines of constant degrees of superheat are constructed for various pressures, these will appear as a series of lines shown as “m n” in Figure A7.2. The actual “T-s” diagram (Fig. A7.1), is prepared from data contained in the 1967 ASME steam tables. The “T-s” diagram has considerable potential for examining the effects of various processes and condition changes in a thermal power cycle. However, it has limited application to the actual steam turbine engine, and the Mollier enthalpy-entropy diagram allows for a more complete analysis of the effects and influences within the turbine unit itself. The enthalpy-entropy (Mollier) diagram To begin to examine operating and other problems arising in the steam turbine, any investigation is aided considerably by a working understanding of the Mollier “enthalpy-entropy” diagram for water/steam. This diagram (with a little interpretation) provides a clear assessment of what is occurring within the unit as there are changes in steam conditions, or as various forms of expansion occur. The Mollier diagram allows an understanding of the phase change from superheated to saturated steam, and allows a prediction of where water will form and the possible effects of its presence within the unit. Steam as a working fluid is ideal for the turbine; in fact had it not occurred naturally, it is ideal and would probably 865 Turbine Steam Path Maintenance and Repair—Volume Two Fig. A7.3—The Mollier ‘Enthalpy-Entropy’ diagram for steam. 866 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam have been invented for this use. Steam lends itself to expansion, can be easily transported through the steam path, and in its “pure” form causes no chemical reaction with the materials of construction. The one unfortunate characteristic of steam is that it forms water, which can (under the influence of high pressures and at high velocities) cause various forms of damage and deterioration within the unit. Within the steam path of the turbine, steam expands, i.e., it passes through the various blade rows, reducing or giving up its pressure, increasing the volumetric flow, but in the process driving the blades and generating power to drive the rotor, which in turn can be applied to any number of mechanical or electrical devices. The Mollier “enthalpy-entropy” diagram provides the locus of the various steam parameters as a function of the steam enthalpy to a base of the entropy of the water/steam fluid. This diagram is of considerable use in identifying the effect of various actions on and within the steam turbine, and it allows analysis of the expansion, leakage, and other factors and limitations that occur within the unit itself. Figure A7.3 shows the Mollier diagram for steam as determined from the ASME steam tables. The Mollier diagram is a complex graph of the variation of main steam parameters, shown as a function of total heat (enthalpy) and entropy (a function of the heat and temperature). Adiabatic expansion. If the steam were to expand through the steam path without incurring any losses due to friction, leakage etc., the expansion would be said to be adiabatic or isentropic, i.e., there would be no change of entropy, and the total process would be reversible. Under these conditions the constant entropy expansion would, on the Mollier diagram, be a vertical line. While the pressure and temperature would fall, and the specific volume would increase, the expansion would be as efficient as possible. Such an expansion would be as shown on the representative Mollier diagram in Figure A7.4. 867 Turbine Steam Path Maintenance and Repair—Volume Two H Pi H1 Pd H2 s1 s Fig. A7.4—A constant entropy (isothermal) expansion between pressures ‘Pi’ and ‘Pd’. Here the inlet pressure is “Pi,” at an enthalpy of “Hi” and the discharge pressure would be “Pd,” at an enthalpy of “Hd.” The expansion would be at constant entropy “s1.” Throttling or a non-expansive expansion. If steam expands and does no work, such as a leakage flow, then the total heat of the steam would be unchanged as there is no expenditure of the steam’s internal energy. In this case, the expansion on the Mollier diagram would be a horizontal line; i.e., a line at constant enthalpy. A non-expansive, or “throttling” expansion is shown in Figure A7.5, where the steam pressure falls from “Pi” at inlet to “Pd” at discharge. During this expansion the enthalpy remains constant at “H1,” but the entropy increases to “s2” by an amount “ds.” This assumes that the enthalpy that is converted to velocity is recovered at the end of the expansion. 868 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam A steam path expansion. In the practical steam path the expansion is not without loss, and the line therefore is not vertical. In fact, the line is displaced to the right on the Mollier diagram, indicating that as energy is released in the steam, there are frictional and other losses, and these tend to increase the entropy of the expanding steam. H H1 Pi A s1 Pd ds B s s2 Fig. A7.5—A throttling or constant entropy expansion between pressures ‘Pi’ and ‘Pd’. A typical expansion line of steam moving through a steam path is shown as Figure A7.6. The actual steam conditions of the steam in the path can be determined from this diagram, as the expansion line represents a locus of the steam parameter variations. At inlet to the steam path, the steam is at a pressure “Pi,” enthalpy “H1” and entropy “s1.” At discharge the pressure has reduced to “Pd.” There will also be a reduction in the enthalpy to “H2,” and an increase in the entropy to “s2” as shown by “ds.” This type of expansion and the form of the expansion locus introduced by these losses will be considered in greater detail later in this appendix. 869 Turbine Steam Path Maintenance and Repair—Volume Two H H1 A Pi dH H2 B C Pd ds s1 s2 Fig. A7.6—The expansion of steam in the turbine steam path, showing the enthalpy drop, from pressure ‘Pi’ accompanied by an increase in entropy, at the same discharge pressure ‘Pd’. The representation of basic power cycles on the thermodynamic charts The steam power cycle is an arrangement of processes that allow the steam or water/steam mixture to be continually cycled through a number of series processes in converting energy from one form to another. The selection and arrangement of these processes are considered in detail later. Also, these processes are conveniently represented on both the “T-s” and “H-s” diagrams. The Carnot cycle The principle of Carnot states that no cycle can be more efficient than one comprising only reversible processes. The Carnot cycle consists of two adiabatic and two isothermal processes working between an upper temperature level of “T1” and a lower temperature of “T3.” This cycle is shown on the “pv” diagram Figure A7.7(a) and on the “T-s” diagram in Figure A7.7(b). 870 Pressure Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Isothermals at constant temperature p1,v1,T1 1 2 p2,v2, T1 γ pv = c p3,v3, T3 4 p4,v4,T3 3 Volume Fig. A7.7(a)—The ‘Carnot cycle’ on the ‘pv’ diagram. T 1 2 4 3 Temperature T1 T2 s1 Entropy s2 s Fig. A7.7(b)—The ‘Carnot cycle’ on the ‘T-s’ diagram. If “P1, V1, and T1” represent the conditions at the beginning of the cycle, condition “1” (the highest or inlet conditions), and the gas is then allowed to expand isothermally to condition “2” (P2, V2, and T1), the expansion is then completed adiabatically to condition “3” 871 Turbine Steam Path Maintenance and Repair—Volume Two (P3, V3, and T3). At this condition the gas is then compressed isothermally to condition “4” (P4, V4, and T3), the compression is then completed adiabatically to return the condition to “1” (P1, V1, and T1), and the cycle is complete. The highest gas conditions are P1, V1, and T1; the lowest conditions are P3, V3, and T3. The ratio of pressure expansion during the isothermal expansion 1-2 is “re,” and equals P1/P2. Similarly, the ratio of compression “rc” during compression is P4/P3. For the cycle to close, these two ratios must be equal, and will be designated as “r.” The heat supplied = P1.V1 loge r = R.T1. loge r The heat rejected = P2.V2 loge r = R.T3. loge r Work Done = Heat Supplied - Heat Rejected = R.T1. loge r - R. T3. loge r = R.loge r . (T1 - Te) Cycle Efficiency = = Work Done Heat Supplied R.loger. (T1 - T3) R.logeR . T1 = T1 -T3 T1 The Rankine cycle In the Rankine cycle the steam is completely condensed from condition “3” to condition “4a” of Figure A7.8. From this condition the water is heated along the saturated water line “4a” to “1.” The cycle is then repeated. 872 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam 1 2 Temperature T1 T3 4a s1 3 Entropy s2 s Fig. A7.8—The ‘Rankine cycle’ on the ‘T-s’ diagram. THE BASIC POWER CYCLES The “T-s” diagram is the most suitable means for analyzing power cycles, while the “Mollier diagram” is the most convenient way of examining the expansion of steam in the turbine. Figure A7.8 shows the basic Rankine cycle in which steam was heated to the saturation line, and then allowed to expand in the turbine. This is close to representing the conditions that occur in the nuclear and geothermal cycles, but is not characteristic of the cycles employed in most units where superheat is added, and then possibly reheated after partial expansion. Figure A8.1 shows the more common basic cycles represented on the “T-s” diagram. The representation of the power cycles Figure A8.1(a) shows this simple “Rankine cycle” as described in Figure A7.8, where steam has its temperature raised from vacuum 873 Turbine Steam Path Maintenance and Repair—Volume Two saturation temperature “Tc,” condition “A,” to temperature “Ts” (the saturation temperature at which the steam is to be delivered to the turbine). At this condition “B,” the steam then has heat added to it at constant temperature “Ts,” raising its energy level to condition “C.” The steam is then expanded in the turbine to a pressure with a saturation temperature “Tc,” condition “D” where it is condensed to condition “A.” This basic cycle is complete. Figure A8.1(b) shows the “T-s” diagram for a cycle, but in this application, while having the same initial and exhaust pressures, there is an addition of further heat after the saturated condition “C” is reached. This heat increases the temperature of the steam, from “C” to “S.” The steam is superheated to a temperature “Tr.” The steam is then expanded in the turbine again to condition “D.” However, the effect of superheating the steam before admission to the turbine results in a lower quantity of water in the exhaust. As the steam expands it crosses the saturation line at condition “N,” at which condition the steam has a temperature “Tn.” Moisture will begin to form soon after crossing this line (see chapter 3). Figure A8.1(c) shows the effect of a single reheat. After being heated to temperature “Tr” in the boiler superheater, and then partially expanded in the turbine to condition “E,” the steam is removed from the turbine, returned to the boiler, and has its temperature raised again to “Tr,” condition “R.” The steam is then returned to the turbine and allowed to expand to condenser pressure and temperature condition “D.” Again, the effect of the reheat can be seen in the final moisture content at condition “D.” In the final cycle is shown the effect of double reheat [Fig. A8.1(d)], where the steam is removed from the turbine for a second stage of reheating, again completing its expansion to condition “D.” In this cycle it has been shown that higher initial steam temperatures and pressures were used. 874 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam S Tr B Ts C B Ts C N Tc A G D s1 Tc A s2 (b) (a) S R1 Tr S Tr B Ts Tc C R B Ts C R2 E F E A D s1 s2 (c) G D s1 s2 Tn G Tc G A s1 D s2 (d) Fig. A8.1—The basic ‘power cycles’ on the ‘T-s’ diagram. In (a) is shown the basic cycle; in (b) superheat has been added to the steam; in (c) there is steam reheat; and in (d) there is double reheat. When considering the steam conditions within the turbine, the Mollier diagram is concerned only with the variation of steam parameters from conditions “C” to “D.” Figure A8.2 shows the same cycles illustrated in Figure A8.1. The enthalpy is shown on these diagrams. In Figure A8.2(a) is the simple cycle, with the expansion line in the steam path represented by the expansion line “A-Da.” The increase in entropy from “s2” to “s2” represents the loss of available 875 Turbine Steam Path Maintenance and Repair—Volume Two H H S Hi Pi Hi Tr Ts Pi Tn Hn Ts C N G Tc G Tc Pe Pe He Da He D Da D s S2’ S2 S2 (a) s S2’ (b) H Hr 2 H R Hr Hi S Pi Ts Tr = Ti R2 R1 Hr1 Hi Hc2 Hc1 Tr = Ti S Pi E E F Ts C C Tn Tn N N G Pe He D He Tc G D Da Da S2 (c) Tc S2’ S2 s S2’ s (d) Figure A8. on the Mollier ‘enthalpyFig. A8.2—The cycles shown in figure A8.1 but2drawn The cycles entropy’ diagram.shown in figure A8.1 but drawn on the Mollier ‘Enthalpy- energy due to inefficiencies in the steam path. In each of the following figures the entropy “s2” is shown as the condition at the end of steam expansion. The effect of superheat and reheat is shown in Figures A8.2(b) and (c), and the effects of reduction in the final moisture content in the cycles (a), (b), and (c), with the same initial pressures can be seen, as the final steam/water mixture is passed to the condenser from condition “Da.” 876 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam The actual expansion and mixing in the steam path In previous sections three various expansions (as they would be represented on the Mollier diagram) were considered, and in Figure A8.2 various cycle configurations were examined. However, the actual expansion that occurs within the steam turbine is not a simple process, and there are a number of locations where there are unequal expansions between the same pressures, and other situations where there is a mixing of steam, or removal of water at different energy levels. A brief description of what occurs at these points will help the engineer responsible for turbines, to make maintenance and replacement decision a little easier to justify. Consider the following conditions: The leakage of steam under a stationary blade row. Steam that leaks past seals and between the stationary blade row and the rotor will re-enter the main steam flow. This leakage of steam upon reentering the main steam flow has two deteriorating effects on performance. These are: • The steam will have throttled past the blade row and have a higher energy level. Figure A8.3 shows this total effect (refer to the Fanno curve, chapter 10, Fig. 10.4.4) • The steam in reentering the main steam flow will disrupt the orderly flow of the steam from the stationary to rotating row causing some level of turbulence Figure A8.3 shows the expansion through a single stage. Steam at inlet to the stage has conditions indicated by subscript “i,” in the axial gap between the stationary and rotating rows by subscript “q,” and at discharge from the stage by subscript “e.” If the quantity of steam leaking past the stationary blades is “ms” #/sec, and the main steam flow through the stationary row is “Ms” #/sec, then a quantity of “ms” #/sec, with an energy content “Hm” Btu/# will mix with a quantity “Ms” #/sec with an energy content “Hq” Btu/#, in the axial gap between the rows. 877 Turbine Steam Path Maintenance and Repair—Volume Two Ti Hi ms Pi Hm M Hq dHs Tq Pq Te He Pe Fig. A8.3—The effect on the expansion line of stationary blade row leakage. This mixture “Ms+ms” of the two flows will have an energy level between “Hq” and “Hm.” The actual energy level is determined by the ratio of “ms/Ms.” Normally the quantity “ms” is small compared to “Ms,” but as leakage quantities increase, so the mean energy level between the stages worsens. Therefore this mixing of the two quantities causes a reduction in the energy that is available to be converted in the stationary blade row. This now unavailable energy is represented by the enthalpy “dHs.” The steam then entering the rotating blade row will have an inlet enthalpy “Hq+dHs.” The leakage of steam over a rotating blade row. A similar situation exists with leakage over the rotating blade tip. Figure A8.4 shows the conditions around a typical rotating blade row. By a similar logic, the leakage quantity over the rotating blades is “mr” and the main steam flow is “Mr.” This will again cause a reduction in the energy available for conversion in the blade row and an enthalpy reduction of “dHr.” 878 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Mr dHs Hq Tq Pq Mt mr He Te Pe dHr s Fig. A8.4—The effects of rotating blade row tip leakage. dHs Hq Pe ∆Ho ∆Ha Pq He ds s Sa Sb Fig. A8.5—Showing the effect of increasing the entropy of the steam entering the rotating blade row. 879 Turbine Steam Path Maintenance and Repair—Volume Two Note: If the Mollier diagram in Figure A8.4 is examined, it will be seen that the apparent energy on the rotating blade has not been reduced significantly by the reducing in energy levels “dHs” in the stationary row. There has been a small increase in entropy, from “dsa” to “dsb,” as shown in Figure A8.5. In this figure, the row still operates between the pressures “Pq” and “Pe,” with an original available energy of “∆Ho.” When the effect of stationary blade leakage is taken into account, there is an increase in the entropy level of the steam from “sa” to “sb” causing an increase of “ds.” However, the available energy is now “∆Ha.” In fact, because of the reheating effect this is slightly larger than “∆Ho,” because some of the thermal energy from the leaking steam has been returned to the working main steam flow making a minor increase in the energy that is available. Steam leakage at an intermediate seal location (N2). If turbine sections (high pressure and reheat) are combined in a single shell, there are situations where extensive losses can be experienced, because steam will leak from one section to another, bypassing complete sections of the steam path. Mh Mr v v mi (b) (a) (c) (d) Fig. A8.6—A combined high pressure and reheat section. In this design the ‘hot’ sections of both expansions are at the center of the casing. 880 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam Consider the form of unit shown in Figure A8.6, which has a combined high pressure and reheat section in a single casing. With this design the hot sections are located at the center of the rotor, with the high-pressure steam entering the unit on plane “a,” and exiting on plane “b.” The steam reenters the unit, after being reheated on plane “c” and exits on plane “d.” Therefore, while the hot sections are located at the center of the rotor, there is a considerable pressure difference at that location, and steam quantity “mi” will leak along the seals that exist there. The effect of this leakage can be seen on the Mollier diagram, Figure A8.7. Here high-pressure section steam, after partial expansion (normally through the first stage nozzle block) will leak along the rotor, and enter the reheat section at condition “K,” and at a pressure “Pr” and with an energy level “Hl,” which is lower than the reheat section steam enthalpy “Hr” with which it is mixing, either directly or after the reheat steam has expanded through the first stage stationary row. The point of reentry depends upon the detail of design. This leakage quantity “mi” mixes with the reheat section steam quantity “Mr” and lowers the mean reheat section enthalpy. This leakage causes a reduction of unit output and heat rate for two reasons: • The leakage steam quantity “mi” bypasses the high-pressure section blading, and does no work there • The leakage steam quantity “mi” causes some level of turbulence or disturbance in the reheat section, again causing losses On the positive side, it can be seen that the energy of the leakage steam “mi” does become available, and does some work in the reheat section, but this is not sufficient to replace that lost in the high-pressure section. The removal of moisture. The removal of deposited moisture from the steam path is discussed in chapter 3, and the effect on the Mollier diagram shown in Figure 3.6.8. Here again the removal of 881 Turbine Steam Path Maintenance and Repair—Volume Two moisture causes a small increase in the entropy, and it would appear from first examination that the energy available would cause a reduction in output. However, because of the reheat effect as discussed earlier, and the water level is reduced, there is less drag on the steam from alternately accelerating and retarding the transported water particles, that the output is increased. Hr Pr Pi T Hi Hl K HP Rht Hch Hcr LP He Pe Fig. A8.7—The expansion lines for the unit shown in figure A8.6. The effect of internal leakage can be seen with steam leaking from the high pressure section mixing with lower pressure steam in the reheat section. 882 Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam REFERENCES A1. ASME Steam Tables, published by the American Society of Mechanical Engineers, New York, NY, 1967 A2. Yellot, J.I., and C.K. Holland. The Condensation of Flowing Steam, Condensation in Diverging Nozzles, Engineering, 1937 A3. Gyamarthy, G. Grundlagen einer Theorie der Nassdampfturbine, Juris Verlag, Zurich, 1960 883 INDEX Index Terms Links A Accept-as-is decision 703-704 Acid washing (boiler) 261 Acronyms Adiabatic expansion ix 847-850 entropy 854-855 Mollier diagram 867-868 Adjustment computation 96-121 adjustment effects 106-110 horizontal-joint blades 111-115 mismatch errors 115-120 raw data vane projection across half joint Adjustment effects final discharge area opening opening-pitch ratio re-measurement of passages steam-discharge angle Adjustment (stage-discharge area/angle) 854-855 96-106 120-121 106-110 109-110 107 107-108 110 108-109 93-96 material removal 95-96 vane bending 95-96 weld buildup 95-96 Administration (quality-assurance program) 683-687 Advanced planning (reverse engineering) 670-671 This page has been reformatted by Knovel to provide easier navigation. 867-868 Index Terms Aging Alignment rotating blades specification tie wires Alignment (rotating blades) Links 292 411-412 665 411-412 712-723 722-723 blade-vane tilt 719-723 center of gravity shift 721-722 incorrect 714-718 Alloy-steel materials 643 Assembly/alignment specification 665 Axial-entry blade roots 502-503 Axial error 767-770 Axial-gap increase 263 Axial-pressure deformation 122 Axial rubs 370 corrective actions 462-463 coverband 391-392 rotor 459-463 783-788 391-392 459-463 corrective actions 462-463 location 460-461 B Base-material buildup 224-230 Base-metal buildup 234-236 Basic-form production 712-723 712-723 blade-tilt effect on tenons Axial rubs (rotor) 665 733 This page has been reformatted by Knovel to provide easier navigation. 459-463 Index Terms Links Basic power cycles 873-882 actual expansion and mixing (steam path) 877-882 representation 873-877 Batching/connection (coverband) 390 Batch stress 413 discontinuity 413 twist 413 Below-shield erosion 239-240 Bend-damage classification (rotor) 469-470 Bending 275 blade 275 rotor 464-479 Bending (blade) 275 Bending (rotor) 464-479 causes 464-466 damage classification 469-470 hardness checks 474-475 permanent 467-468 run-out checks 470-474 straightening options 475-479 stress relief temporary 479 466-467 Bend straightening (rotor) 476 Between-shield erosion 240 Beyond-shield erosion 239 Blade access Blade bending Blade-cascade definitions blade pitch expansion-passage form 464-479 295-296 275 756-760 756 759-760 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Blade-cascade definitions (Cont.) cascade width 760 inlet/discharge edges 760 profile-setting angle 759 steam-discharge angle steam-inlet angle 757-758 757 steam turning angle 758-759 throat opening 756-757 Blade dressing 275 Blade-fillet radii 517-519 Blade inlet-edge erosion damage/refurbishment 206-241 braze-attached shield 230-234 braze-attached shield/base-metal buildup 234-236 braze-attached shield detachment 208 braze-material cracking 211-212 dressing an eroded surface 238-239 local erosion penetration 208 off-shield erosion repair 239-241 raw-weld deposit 236-238 thermally-hardened inlet edges 212-213 weld-attached inlet edge/base-material buildup 224-230 weld-attached shield cracking 209-210 weld-attaching solid-bar inlet nose 213-224 Blade listing Blade-manufacturing processes basic-form production 247 726-749 733 cutting metal 733-744 electric-discharge machining 729-730 envelope forging 727 extrusion process 747 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Blade-manufacturing processes (Cont.) forging of material 733 forging process 744-746 metal cutting 726-727 pinch rolling 731-732 precision forging 727-729 vane extrusion 729-731 Blade modifying steam-flow distribution 260 Blade moment 243 Blade mounting 725 Blade numbering 243-246 blade moment 243 blade weighing 243 data analysis Blade opening/throat (O) 243-246 29-30 Blade pitch 756 errors 803-811 Blade-pitch errors (factors) 803-811 803-811 blade-root platform pitch 803-805 blade-vane inclination 806-807 root-wedge angle error 808-811 vane placement on root platform 805-806 vane twist/shape variations 747-749 808 Blade placement 248 Blade-production techniques 724 Blade-profile definitions 752-755 discharge nose/tail 753-754 inlet nose 753 metal-section discharge angle 754 metal-section inlet angle 754 766-770 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Blade-profile definitions (Cont.) metal-section turning angle 754 pressure/suction faces 755 profile chord/thickness 755 profile sectional area/bending modulus 755 radius of curvature (profile surfaces) Blade-profile/cascade tolerances 754-755 749-764 blade-cascade definitions 756-760 blade-profile/cascade quality 751-752 blade-profile definitions 752-755 gauging the profile 761-764 Blade-profile/placement errors 766-770 Blade removal 275-277 access 295-296 Blade replacement 501-502 Blade reuse 499-501 Blade-root corrosion 483-487 295-296 fretting 486-487 pitting 483-486 Blade-root entry 502-503 783-796 axial 502-503 783-788 tangential 788-796 Blade-root platform pitch 803-805 Blade-root region (rotor) 516-517 Blade-root steeples/wheel rim 480-497 closing-window damage 496 corrosion 483-487 fretting 486-487 high-cycle fatigue 481-482 impact damage 496-497 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Blade-root steeples/wheel rim (Cont.) low-cycle fatigue 493-494 pitting 483-486 side-grip damage 488-492 stress-concentration centers 481-482 washing erosion wheel gouging Blade-root tolerances 497 494-496 775-802 axial-entry direction 783-788 blade-vane positioning (root platform) 776-783 load-bearing surface radius 796-797 radial-entry direction 797-801 root surface-finish requirements tangential-entry direction 802 788-796 Blade selection 247-248 Blade-setting error 768-770 Blade trailing-edge erosion 255-257 Blade vanes 213-216 384-385 422-424 719-723 776-783 806-807 crack (tie wires) 422-424 inclination 806-807 positioning (root platform) 776-783 preparation 213-216 shortening 384-385 tilt 719-723 Blade-vane tilt 719-723 effect on tenons 722-723 Blade weighing 243 Blast cleaning 205 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Boiler tube 261-262 materials 261 temperature 262 Borasonic examination (rotor forgings) Borehole machining 342 542-543 Boyles law 831 Braze-attached shield 208 base-metal buildup 234-236 cleaning 231-232 detachment joint geometry post-braze inspection post-brazing activities preheating 230-236 208 232-233 234 233-234 232 Braze attachment 208 230-236 shield 208 230-236 Braze-material cracking Brazing 427-431 211-212 208 211-212 230-236 283-284 388-389 427-431 attachment 208 230-236 427-431 coverband 388-389 material cracking 211-212 post-brazing activities 233-234 Built-up rotor 331-332 Burns (rotor) 459-463 corrective actions 462-463 location 460-461 Butt clearances (gland ring) Buttering (weld material) 620-621 539 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links C Carbon rings (gland ring) 619-620 Carnot cycle 870-872 Carrier geometry (gland ring) 622-623 Cascade and profile tolerances (blade) 749-764 blade-cascade definitions 756-760 blade-profile/cascade quality 751-752 blade-profile definitions 752-755 gauging the profile 762-764 width Casing components 760 13-15 casing-exhaust geometries 171-173 explosion/relief diaphragm 168-169 high-pressure packing heads 170-171 high-pressure/high-temperature sections 162-166 low-pressure/low-temperature sections 166-168 Casing-creep cracking Casing definition 162-173 184-186 13-15 diffuser at exhaust 15 explosion diaphragm 15 inlet section 15 shaft-end packing head 15 shells 15 Casing distortion 590-592 Casing-exhaust geometries 171-173 Casing operating problems/repair methods 174-198 casing-creep cracking 184-186 casing ovality 186-193 low-cycle fatigue cracking 176-184 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Casing operating problems/repair methods (Cont.) horizontal joint leakage 197-198 humping/hogging 193-196 predictable phenomena repair methods unpredictable phenomena Casing ovality 175 180-184 176 186-193 correction 190-193 inward and binding 187-188 outward 188-189 Casing repair methods Casings components creep cracking definition diffuser at exhaust 190-196 174-198 2 4-5 162-198 590-592 13-15 162-173 184-186 13-15 15 distortion 590-592 exhaust geometries 171-173 explosion diaphragm 15 inlet section 15 operating problems/repair methods 174-198 ovality 186-193 shaft-end packing head 15 shells 13 15 13 15 Casing shells Cast construction 60-61 Caulked strips (gland/seal strip) 636-637 Center of gravity shift 721-722 Centrifugal stress 641 411 This page has been reformatted by Knovel to provide easier navigation. 13-15 Index Terms Centrifuged-moisture erosion Charles law Chemical deposition Chemical-vapor deposition Classes of error (blade profile/placement) Links 54-56 831-832 36-39 266 766-770 axial 767-770 blade setting 768-770 pitch 767-768 770 205-206 231-232 Cleaning blast 205 hand 206 shield Closing-window damage Coating processes 231-232 496 264-270 chemical vapor 266 diffusion/alloying 265 gas phase 268 overlay 265 pack cementation 266-269 plasma 266-270 physical vapor (electron beam) Coating protection (corrosion) Company organization (quality assurance) Component evaluation/testing (reverse engineering) 266 289-291 684 676-677 Component installation (reverse engineering) 677 Component operating modes (changes) 693 Components (rotating) 309-362 coverbands 313-317 rotor 327-342 single-tie connections 323-326 326 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Components (rotating) (Cont.) tie wires 317-326 turbine-rotor discs 342-362 Components (stationary) casing definition diaphragms stationary-blade definition Compressive stress Computer traces (gauging) Concentricity/balance machining Constant-pressure gas expansion (entropy) Constant-volume gas expansion (entropy) Continuous connection repair (tie wires) blade-vane crack fretting mid-span crack Continuous connection (tie wires) repair Control-stage massive-particle damage 1-15 13-15 5-12 5-6 411 763-764 542 855-856 855 418-424 422-424 424 419-421 414-416 418-424 418-424 277-280 Copper plugs/backing strips 70-73 Correction/repair 59-87 124-128 132-148 153-162 174-198 239-241 291-303 418-426 462-463 59-87 124-128 133-148 erosion 239-241 291-292 horizontal joint 153-154 702-703 axial rubs (rotor) 462-463 casing 174-198 decision 702-703 diaphragm This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Correction/repair (Cont.) stationary blade 154-162 tie wires 418-426 Corrective options (diaphragm) 124-128 132-133 141-148 281-291 304 398-399 424 439-454 483-487 diametral change 132-133 dishing 124-128 heat and pressure 124-125 inner web 141-148 machining 125-128 permanence Corrodents 133 450-454 hydrogen sulphide 453-454 oxygen 452-453 sodium hydroxide 451-452 Corrosion blade root 483-487 corrodents 450-454 coverband 283-287 cracking 398-399 287 disc/spindle interface 445-450 effects 281-291 fatigue 289 fretting 304 398-399 424 486-487 pitting 287-289 483-486 rotating blades 281-291 304 rotor 439-454 steam throttling 442-443 stress cracking 287 This page has been reformatted by Knovel to provide easier navigation. 442 Index Terms Links Corrosion (Cont.) susceptibility 281-287 tenons 283-287 tie wires Corrosion (blade root) 424 483-487 fretting 486-487 pitting 483-486 Corrosion (coverband) corrosive attack 283-287 398-399 tenons 283-287 Corrosion effects 398-399 398 fretting Corrosion cracking 398-399 398-399 287 281-291 coating protection 289-291 regions susceptible to damage 281-287 rotating-blade corrosion forms 287-289 Corrosion fatigue 289 Corrosion (fretting) 304 398-399 486-487 tie wires 424 Corrosion pitting 287-289 Corrosion (rotor) 439-454 blade-tip seal leakage 483-486 444 corrodents 450-454 diaphragm leakage 443-444 disc/spindle interface 445-450 hydrogen sulphide 453-454 oxygen 452-453 rotating-blade rows 442 shaft-end sealing positions 445 This page has been reformatted by Knovel to provide easier navigation. 424 Index Terms Links Corrosion (rotor) (Cont.) sodium hydroxide 451-452 steam throttling 442-443 Coverband damage/repair/refurbishment axial rubs batching/connection of groups 283-287 391-392 390 blade-vane shortening 384-385 brazing 388-389 corrosion 283-287 impact damage 363-371 over-riveting of tenons 395-396 refurbishment techniques 396-397 riveting 402-410 screw attachment 382-383 surface rubs 385-388 tenon and hole requirements 400-401 tenon failures 362-410 398-399 395 tenon-hole form 393-394 400-401 tenons 371-382 393-396 400-410 Coverbands 283-287 313-317 323-326 362-410 attached 314 batching with tie wires 323-326 cracks at hole 370-371 damage/repair/refurbishment 283-287 forms 314-317 functions 313-314 integral 314 multi-layer 315 types/shapes 362-410 315-317 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Crack filling 529-530 Crack grinding 504-505 Cracking (corrosion) Cracking (cross-section) Creep deflection Critical point (temperature-entropy) Cross-section irregularities (vane) Cutting metal cutting forces 287 535-537 122 863-864 32-35 726-727 735-741 vortex profiles 742-743 Cylindrical profiles 733-744 743 cylindrical profiles Cutting process (machining) 145-147 706 735-741 D Damage and refurbishment (rotating blades) 201-307 SEE ALSO Rotating blades (damage and refurbishment) Damage and refurbishment (rotating components) 309-548 SEE ALSO Rotating components (damage and refurbishment) Damage and refurbishment (stationary components) 1-199 SEE ALSO Stationary components (damage and refurbishment) Damage examination (weld repair) 531-538 axial-surface rebuild 531 cross-section cracks 535-537 stub shaft 537-538 wheel rebuild 531-535 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Damage, deterioration, and failure mechanisms (overview) x-xxiv aging xiv external factors xv failure definition xvi leakage (seal/gland) xviii maintenance actions xxi-xxiv maintenance options xx-xxiv maintenance problems xiv operating changes xiv quality and inspection repair/refurbishment options xviii xix-xx steam environment xv 283-287 root-fastening area 281-282 tie-wire hole in vane 282-283 Data analysis (blade numbering) 243-246 individual opening-discharge areas opening (O) pitch (P) xvii-xviii 281-287 coverband at tenons Data (raw) xvii xviii-xvix situation evaluation Damage susceptibility (corrosion) xvii 96-106 103-106 97-99 98 100 106 radial height 100-103 ratio O/P 103-104 Deep-weld repair (rotor cracks) cracks 106 508-527 509 522-527 stub attachment crack 508-509 522 surfaces 508-514 This page has been reformatted by Knovel to provide easier navigation. 103-104 Index Terms Links Deep-weld repair (rotor cracks) (Cont.) weld attachment of forgings 522 524-527 wheel 508 514-522 Deflection 121-122 145-147 creep 122 145-147 elastic 121 Design and delivery assumptions xiii Design-review meeting minutes 693 Design specification xiii assembly/alignment 665 material 664 nondestructive tests 665 physical dimensions 664 special processes 665 surface finish 665 Deterioration (tie wires) 426 Diametral change (diaphragm) 129-133 corrective options 132-133 inner web 147-148 Diaphragm inner-web correction 145-147 diametral changes 147-148 dishing 142-145 Diaphragm leakage 443-444 diaphragm vane 147-148 141-148 creep deflection Diaphragm repair 663-665 59-87 124-128 59-87 inner web 141-148 sidewall 133-141 thermal distortion 124-128 This page has been reformatted by Knovel to provide easier navigation. 133-148 Index Terms Diaphragms explosion inner ring/web leakage nozzle plate/box outer ring repair stationary blades thermal distortion vane Diaphragm-sidewall repair methods Links 2-12 15 121-148 443-444 15 8-9 9-12 5-7 59-87 121-133 7-8 133-141 welding (cast iron) 139-141 weld rebuild 135-137 Diaphragm thermal distortion 121-133 corrective options 124-128 diametral change 129-133 dishing 121-128 59-87 60-63 metallic inserts 84-86 suction-face finishing 86-87 vane repair methods 64-87 weld buildup 64-84 Diffuser at exhaust 15 Dimensional requirements (reverse engineering) 132-133 133 manufacturing tolerances Diffusion coatings/alloying 124-128 9 137-138 Diaphragm-vane repair (methods) 141-148 443-444 metallic inserts permanence of correction 59-87 265 671-672 This page has been reformatted by Knovel to provide easier navigation. 133-148 Index Terms Dimensional requirements/specification before adjustment reverse engineering Links 28-35 83-84 664 671-672 697-698 724 56-58 753-754 83-84 671-672 stationary-blade row 28-35 surface conformance 697-698 Dimensional requirements (stationary-blade row) 28-35 blade opening/throat (O) 29-30 ratio O/P 29-30 setback (inlet/trailing edge) 31-32 sidewall-discharge diameters 32 vane cross-section irregularities 32-35 vane pitch (P) 29-30 vane-setting angle 30-31 vane-tilt angle 31 Discharge area (adjustment) 109-110 Discharge edge (size) 118-120 Discharge/inlet edges 760 Discharge nose 753-754 Discharge tail 27 cracking 56-58 Discs 342-362 corrosion 445-450 keyways and securing 358-362 removal 356-359 rim 348-349 spindle interface 445-450 turbine rotor 342-362 Disc-spindle assembly 352-356 Disc/spindle interface 445-450 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Discs (turbine rotor) 342-362 disc removal 356-359 disc rim 348-349 disc-spindle assembly 352-356 disc-to-hub fillet radii 350 forms functions 343-350 343 interstage seals 346-348 keyways and securing 358-362 machining for shrink fits 350-351 pressure-balance holes 349-350 shrink-fit design 351-352 spindle interface 445-450 Disc-to-hub fillet radii 350 Dishing correction 124-128 heat and pressure 124-125 machining 125-128 Dishing (diaphragm) 121-128 axial pressure deformation corrective options 124-128 122 elastic deflection 121 rubbing Documentation and shipment Dressing blade eroded surface 142-145 122 creep deflection inner web 445-450 142-145 122 699 238-239 275 275 238-239 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links E Elastic deflection (dishing) Electric-discharge machining Electron-beam physical-vapor deposition Engineering review 121 729-730 266 680-682 access to sub-supplier plants 682 drawings 681 hold points 682 inspection and test plan 681 manufacturing processes 681 nonconforming components 681 quality-assurance program quality records Enthalpy and entropy 680-681 682 813-883 enthalpy of gas 825-828 entropy changes of steam 852-857 Mollier diagram 865-870 Enthalpy-entropy (Mollier) diagram adiabatic expansion steam-path expansion throttling/non-expansive expansion Enthalpy of gas 865-870 867-868 869 868-869 825-828 British thermal unit 824 calorie 824 Centigrade heat unit 824 Entropy changes of steam 852-857 adiabatic/isothermal expansion 854-855 constant-pressure gas expansion 855-856 constant-volume gas expansion 855 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Entropy changes of steam (Cont.) general expression for change 856-857 temperature-changing cycle 853-854 Erosion damage/repair/control repair 46-50 52-56 153 206-241 254-280 291-292 365-366 454-456 497 239-241 291-292 259 263-270 SEE ALSO Solid-particle erosion (SPE) Solid-particle impact (SPI), AND Water-induced damage Erosion-resistant coatings chemical-vapor deposition 266 diffusion coatings/alloying 265 gas-phase coatings 268 overlay coatings 265 pack cementation physical-vapor deposition (electron beam) plasma-coating process Erosion-shield damage cracks Error types (blade profile/placement) 266-269 266 266-270 208-210 254-255 254-255 766-770 axial 767-770 blade setting 768-770 pitch 767-768 770 Examples 246-247 249-252 597-603 kW loss prediction 600-603 Martin method 585-589 moment weighing 246-247 seal maintenance economics 597-599 249-252 This page has been reformatted by Knovel to provide easier navigation. 585-589 Index Terms Links Expansion at constant pressure 845 855-856 Expansion at constant volume 845 855 Expansion-passage form 759-760 Experience with supplier 693 Explosion diaphragm Extended vanes 15 168-169 33-34 Extrusion (vane) 729-731 Eye lashing (gauging) 762-763 747 F Fabricated construction 60-61 Failure (definition) xvi Fatigue (corrosion) 289 Ferrules loss 424 Forging 337-342 522 524-527 727-729 733 744-746 524-527 envelope 727 material 733 precision 727-729 process 744-746 rotor 337-342 522 522 524-527 337-342 522 rotor attachment Forgings (rotor) basic production borasonic examination 337-338 342 central-inspection boreholes 340-341 inspection during manufacture 338-340 rotor attachment 522 524-527 This page has been reformatted by Knovel to provide easier navigation. 524-527 Index Terms Links Forms of seal knife-edge discharge 604-618 knife-edge form 605-607 multi-strip seal configurations 612-618 seal-location geometry 608-612 Fretting corrosion 304 398-399 486-487 tie wires 424 Full-arc admission mode 260-262 Fusion techniques (rotating blades/hardware) 427-433 braze attachment 427-431 weld attachment 431-433 G Gas equations 831-833 Boyles law 831 Charles law 831-832 general gas equation 832 normal temperature and pressure 833 Gas-phase coatings Gauging (profile) 268 761-764 computer traces 763-764 eye lashing 762-763 guillotine gauge 761 projection methods 762 General case (steam expansion work) General expression (entropy) 846 856-857 General gas equation 832 Geometry-change machining 542 This page has been reformatted by Knovel to provide easier navigation. 424 Index Terms Links Gland-ring form 618-627 butt clearances and tangential location 620-621 carbon rings 619-620 carrier geometry 622-623 seal and radial/axial steam forces 624-627 spring loading 621 steam seal 623 Gland-ring operating problems 643-650 Gland-ring/seal-strip materials 641-643 alloy steel sand cast 643 642-643 Gland/seal-strip assembly 634-643 caulked strips 636-637 inserted segments 637-638 materials 641-643 staked strips 635-636 Grinding repair 180 Guillotine gauge 761 641 504-505 H Hand cleaning 206 Hardness checks (rotor bending) 474-475 Heat addition to water/steam 833-837 change of states 834 conversion of ice to superheated steam 836-837 formation of steam 834-835 latent heat of evaporation moisture content/dryness fraction superheated steam Heat-affected zone (HAZ) 835 835-836 836 541 This page has been reformatted by Knovel to provide easier navigation. Index Terms Heat engine Heating and expansion of steam Links 858 833-852 addition of heat to water/steam 833-837 expansion of steam 842-845 heating and expansion relationships 837-839 heating of steam 840-841 metastable conditions during expansion (supersaturation) ratio of specific heats 851-852 841 specific heats of steam 839-840 work done by steam expanding 845-851 Heating and expansion relationships (steam) Heat of evaporation Heavy damage (SPI) 837-839 835 44-46 High-cycle fatigue 481-482 High-pressure packing heads 170-171 High-pressure/high-temperature sections 162-166 Hi-lo staggered seal configuration History (steam turbine) 615 x Hogging. SEE Humping/hogging Hold or witness points Hole plugging/re-drilling Horizontal half-joint Horizontal joint 696 296-299 589 111-115 153-154 589 adjustment blade adjustment; half joint weld repair Humping/hogging correction 111-115 589 153-154 193-196 194-196 This page has been reformatted by Knovel to provide easier navigation. 197-198 Index Terms Links Hybrid rotor 336-337 Hydrogen sulphide 453-454 I Ice conversion to superheated steam 836-837 Impact damage 363-371 coverband 363-371 Impact damage (coverband) axial rubs 363-371 370 cracks at coverband hole 370-371 excessive overspeed 368-369 heavy radial rubs 368-369 increased thickness 368 reduced thickness 366-368 solid-particle erosion 365-366 solid-particle impact 363-364 water-impact damage 364-365 Impulse-unit seals 578-581 Induced bends (rotor) 464-479 bend-damage classification 469-470 causes 464-466 hardness checks 474-475 permanent 467-468 run-out checks 470-474 straightening options 475-479 stress relief temporary Inlet/discharge edges 496-497 479 466-467 760 Inlet-edge erosion (blade) 206-241 Inlet-edge profile finishing 222-223 This page has been reformatted by Knovel to provide easier navigation. Index Terms Inlet nose definition Inlet portion size Inlet section Links 26 753 753 117-118 15 Inlet-stage nozzle geometry 260 263 Inner ring/web 8-9 141-148 correction 141-148 Inserted segments (gland/seal strip) 637-638 Inspection and test plan (I&TP) 688-689 692 Inspection (components) 233-234 338-342 blades brazing rotor forgings welding 711-813 234 338-342 233 Inspection department 686 Inspection-instrument calibration 695 Inspection (quality control) 694-695 instrument calibration 695 records 694-695 Inspection/surveillance (activities and responsibilities) dimensional/surface conformance xviii-xvix 697-698 documentation and shipment 699 hold or witness points 696 inspection-instrument calibration 695 inspection records material compliance 694-695 698 nonconforming items 696-697 nondestructive testing 698 quality-program review 694-699 694-695 This page has been reformatted by Knovel to provide easier navigation. 711-813 Index Terms Links Inspection/surveillance (Cont.) special processes supplier-purchaser control 697 698-699 Integral connection repair (tie wires) 425 Integral snubbers 299 Internal energy 824 Interstage seals (disc) Irregularities (vane cross-section) 346-348 32-35 J Joint geometry 232-233 K Keyways and securing (disc) 358-362 Knife-edge form 605-607 KW loss prediction 599-604 examples 600-603 L Labyrinth leakage 559-594 calculation 565-567 gland-ring form 618-627 gland-ring operating problems 643-650 Martin method 567-594 references 651 seal knife-edge discharge forms 604-618 seal-maintenance economics 595-604 seal-strip forms 627-630 seal-strip/gland-ring materials 641-643 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Labyrinth leakage (Cont.) seal-strip insertion and securing 630-641 steam seal 559-567 Last-stage massive-particle damage Latent heat evaporation Layout/measurements (welding) Leakage 277-278 863 835 65-66 69 559-594 877-882 labyrinth 559-594 loss calculation 567-594 seal 559-594 steam 877-882 Leakage loss (Martin method) 567-594 example 585-589 incremental loss 593-594 radial-seal clearance 589-592 reaction-turbine dummy pistons 571-575 seal axial clearance 592-593 shaft-end glands 568-571 steam-path seals 575-585 Light damage (SPI) 835 41-42 Lines of constant superheat 864-865 Load-bearing surface radius (blade root) 796-797 Long-shank blades 505-507 Low-cycle fatigue 176-184 457-458 522-524 cracking 176-184 rotor 457-458 This page has been reformatted by Knovel to provide easier navigation. 493-494 Index Terms Links Low-cycle fatigue cracking 176-184 grinding 180 stitching 181-184 welding 180-181 Low-pressure/low-temperature sections 166-168 M Machined from solid (construction) 61 Machining 61 350-351 477-478 541-543 705-709 729-730 components 705-709 finish 541-543 from solid 61 remachining rotor 477-478 shrink fit 350-351 Machining (components) cutting process 705-709 706 surface finish 708-709 surface integrity 707-708 Machining finish boreholes 541-543 707-709 542-543 concentricity/balance 542 geometry changes 542 pressure-balance holes 543 surface 542 Machining (shrink fit) 707-709 350-351 Maintenance xiv actions xxi-xxiv economics 595-604 xx-xxiv This page has been reformatted by Knovel to provide easier navigation. 595-604 Index Terms Links Maintenance (Cont.) options problems xx-xxiv xiv Manufacture/inspection requirements (steam-turbine blades) 711-813 blade-manufacturing processes 726-749 blade-pitch errors (factors) 803-811 blade-root tolerances 775-802 manufacturing techniques 723-726 passage swallowing capacity 770-774 profile and cascade tolerances 749-764 profile and placement errors 765-770 radial alignment (rotating blades) 712-723 references 813 special processes (vane) 774 stage-hardware requirements 812 Manufacturing processes (blade) basic-form production 726-749 733 cutting metal 733-744 electric-discharge machining 729-730 envelope forging 727 extrusion process 747 forging of material 733 forging process 744-746 metal cutting 726-727 pinch rolling 739 precision forging 727-729 vane extrusion 729-731 747-749 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Manufacturing techniques (blade) 723-725 blade mounting 725 blade production 724 dimensional requirements 724 material specifications 724 nondestructive examination 725 post-assembly requirements 725 spatial relationships 724 special processes 725 surface-finish requirements 724 tolerances 724 blade 724 Manufacturing tolerances 60-63 cast construction 60-61 fabricated construction 60-61 machined from solid 61 pinned construction 60-61 welded construction 60-61 vane adjustment 62-63 Martin method (leakage quantification) 724 567-594 blade tip 586-587 diaphragm packing 586-588 examples 585-589 incremental leakage loss 593-594 radial seal clearance 589-592 reaction turbine dummy pistons 571-575 seal axial clearance 592-593 shaft-end glands 568-571 steam-path seals 575-585 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Massive-particle damage 273-280 all stages bending and dressing blade removal 278-280 275 275-277 continue to operate 275 control stages 277 last-stage blades Materials 277-278 69 95-96 539 664 672-675 698 724 733 compliance 698 deformation (form) 733 forging 733 removal 69 reverse engineering requirements/substitution specification 95-96 672-675 664 724 Measurements required 65-66 69 Medium damage (SPI) 43-44 Metal section angles (blade) Metallic inserts 754 84-86 Metastable conditions (steam expansion) 851-852 Mid-span crack (tie wires) 419-421 Misalignment 411-412 rotating blades 712-723 tie wires 411-412 Mismatch errors (vane) 137-138 712-723 115-120 small discharge edge 118-120 small inlet portion 117-118 vane mismatch 115-117 Moisture content/dryness fraction 539 835-836 864 This page has been reformatted by Knovel to provide easier navigation. 152-153 Index Terms Moisture-impact damage Links 54-56 364-365 SEE ALSO Water-induced damage. Moisture removal (steam) 881-882 Mollier diagram (enthalpy-entropy) 865-870 adiabatic expansion steam-path expansion throttling/non-expansive expansion Moment weighing alternate means 867-868 869 868-869 224 253 blade numbering 243-246 examples 246-247 moment-weighing process tangential-position determination 242-253 249-252 243 247-248 Monobloc rotor 330-331 Multi-strip seal configurations 612-618 alternative stationary/rotating 614 hi-lo staggered 615 multi-high strip 615-618 multi-high strip staggered 615-617 straight-through 613-614 N New units (refurbishment) 292 Nonconforming items 696-697 Nonconforming situations 699-704 accept-as-is 703-704 repair 702-703 rework scrap and replace Nondestructive examination (NDE) 703 700-702 725 This page has been reformatted by Knovel to provide easier navigation. 454-456 Index Terms Nondestructive testing (NDT) Non-expansive expansion (throttling) Normal temperature and pressure (gas) Nozzle plate/box Nuclear unit steam-path seals Links 665 698 868-869 833 9-12 32-33 557 O Off-shield erosion repair 239-241 below shield 239-240 between shield 240 beyond shield 239 pressure surface 240 tenon 241 Older units (refurbishment) Opening pitch-ratio adjustment 292 107-108 Operate/shutdown decision 275 Operating changes xiv xvii Operating damage mechanisms (rotating components) 309-548 SEE ALSO Rotating components (damage and refurbishment) Operating damage mechanisms (stationary components) 1-199 SEE ALSO Stationary components (damage and refurbishment) Operating experience (supplier) Operating phenomena (stationary components) 692 36-59 chemical deposition 36-39 solid-particle erosion 46-50 solid-particle impact 40-46 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Operating phenomena (stationary components) (Cont.) thermal transients 58-59 vane-discharge tail cracking 56-58 vane-discharge thumbnail crack water-induced damage Outer ring 58 50-56 5-7 damage 148-153 Outer-ring damage 148-153 solid-particle erosion steam-seal face damage Overlay coatings 153 149-153 265 Over-riveting (tenons) 395-396 Overspeed damage 368-369 coverband tie wires Oxide scale erosion preventive/corrective actions 426 257-264 259-264 257 rotating-blade tip-section suction face 258 Oxygen (corrodent) 426 368-369 rotating-blade inlet edge tenons attaching coverband 148-153 258-259 452-453 P Pack cementation 266-269 Particle hardness 263 Particle size 263 Passage swallowing capacity 770-774 Peening 271-273 Rotor 273-280 477 477 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Performance definitions 660-663 performance factors 661-663 Performance factors (quality) 661-663 availability efficiency Permanence of correction Physical properties (water and steam) 662 662-663 133 816-830 enthalpy of gas 825-828 internal energy 824 pressure 816 specific volume 816-822 temperature 822-824 viscosity 829-830 Physical-vapor deposition Pinch rolling 266 731-732 Pinned construction 60-61 Pitch (vane) 29-30 747-749 767-768 770 767-768 770 803-811 Pitting (corrosion) 287-289 483-486 Plasma-coating process 266-270 803-811 errors Post-assembly requirements Post-brazing activities inspection 725 233-234 234 Post-weld heat treatment 220-222 Power cycles 873-882 actual expansion and mixing (steam path) 877-882 representation 873-877 Predictable phenomena (casing) 175 Preheating 232 540 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Pressure-balance holes 349-350 disc rotor 349-350 517 Pressure property (water and steam) 816 Pressure/suction faces (blade) 755 Pressure surface erosion 240 Preventive/corrective actions (oxide scale erosion) 259-264 acid washing of boiler 261 axial-gap increase 263 blades to modify steam-flow distribution 260 boiler-tube materials 261 boiler-tube temperature 262 erosion-resistant coatings 259 full-arc admission mode 263 260-262 inlet-stage nozzle geometry 260 scale-particle hardness 263 scale-particle size 263 vane material 262 Product surveillance (quality) 517 685-686 263 690-699 inspection/surveillance activities and responsibilities 694-699 responsibility 685-686 supplier facility inspection (preparation) 691-693 Profile and cascade tolerances (blade) 749-764 blade-cascade definitions 756-760 blade-profile/cascade quality 751-752 blade-profile definitions 752-755 gauging the profile 762-764 Profile and placement errors (blades) classes of error 765-770 766-770 This page has been reformatted by Knovel to provide easier navigation. 543 Index Terms Profile chord/thickness (blade) Profile gauging Links 755 761-764 computer traces 763-764 eye lashing 762-763 guillotine gauge 761 projection methods 762 Profile sectional area/bending modulus 755 Profile setting angle 759 Profile (vane) blade chord/thickness 20-23 755 blade placement errors 765-770 blade tolerances 749-764 constant-but-reducing 21-22 constant section 20-21 gauging 761-764 sectional area/bending modulus 755 setting angle 759 twisted 25-28 22-23 Program preparation/implementation (quality assurance) 683-684 Projection methods (gauging) 762 Purchaser assurance (quality) 689-690 Purchasing/requisitioning engineer 692 Q Quality assurance (QA) 653-710 available quality-assurance program 704-705 definition of quality 658-659 definitions of performance 660-663 design specification 663-665 This page has been reformatted by Knovel to provide easier navigation. 749-770 Index Terms Links Quality assurance (QA) (Cont.) engineering review 680-682 inspection and test plan 688-689 machining of components 705-709 nonconforming situations 699-704 product surveillance 690-699 purchaser assurance of quality 689-690 quality-assurance manual 679-680 quality-assurance program 677-679 683-687 704-705 683-687 694-695 quality-assurance program responsibility/administration references 683-687 710 responsibility for quality 657-658 reverse engineering 666-677 Quality-assurance manual 679-680 Quality-assurance program 677-679 704-705 availability company organization final quality inspection department product quality responsibility 704-705 684 686-687 686 685-686 program implementation 684 program preparation 683 responsibility/administration 683-687 review/inspection 694-695 Quality (definition) 658-659 Quality-program review/inspection 694-695 instrument calibration records 695 694-695 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links R Radial alignment (rotating blades) 712-723 blade-tilt effect on tenons 722-723 blade-vane tilt 719-723 center of gravity shift 721-722 incorrect alignment 714-718 Radial rubs 368-369 Radial-entry direction (blade root) 797-801 Radial-seal clearance 589-592 casing distortion 590-592 horizontal half joint 589 Radius of curvature (profile surfaces) 754-755 Rankine cycle 872-873 Ratio O/P 29-30 Raw data 96-106 individual opening-discharge areas opening (O) pitch (P) 103-104 106 100 103-104 103-106 97-99 98 106 radial height 100-103 ratio O/P 103-104 Raw-weld deposit 235-238 Reaction-turbine dummy pistons 571-575 Reaction-unit seals 581-585 References 106 199 305-307 544-548 651 710 813 883 This page has been reformatted by Knovel to provide easier navigation. Index Terms Reforming Links 371 crack grinding 504-505 root 503-504 rotor rim 503-505 tenons 503-505 371 Refurbishment techniques (coverbands) 396-397 Refurbishment techniques (rotating blades) 201-307 SEE ALSO Rotating blades (damage and refurbishment) Refurbishment techniques (rotating components) 309-548 SEE ALSO Rotating components (damage and refurbishment) Refurbishment techniques (stationary components) 1-199 SEE ALSO Stationary components (damage and refurbishment) Relief diaphragm 168-169 Remachining (rotor) 477-478 Re-measurement of passages Repair (casing) Repair (diaphragm) 110 174-198 59-87 inner web 141-148 sidewall 133-141 thermal distortion 124-128 vane Repair (tie wires) continuous connection 124-128 59-87 418-426 418-424 deterioration 426 ferrules loss 424 overspeed 426 snubber/integral connection 425 This page has been reformatted by Knovel to provide easier navigation. 133-148 Index Terms Repair/correction Links 59-87 124-128 133-148 153-162 174-198 239-241 291-303 418-426 462-463 702-703 axial rubs 462-463 casing 174-198 decision 702-703 erosion 239-241 291-292 59-87 124-128 diaphragm horizontal joint 153-154 stationary blade 154-162 tie wires 418-426 Responsibility for quality 657-658 product 685-686 quality-assurance program 683-687 Reverse engineering 666-677 advanced planning 670-671 component evaluation/testing 676-677 component installation 677 concept 668-670 dimensional requirements 671-672 material requirements 672-673 material substitution 673-675 special processes 675-676 Reversibility heat engine Review (quality-assurance program) Rework decision Reworking tenons 683-687 857-858 858 694-695 703 371-372 This page has been reformatted by Knovel to provide easier navigation. 133-148 Index Terms Links Riveting process (tenon) 402-410 hand 405 pneumatic 405 rolled-rivet head 405 409-410 395-396 402-410 Riveting (tenon) over-riveting 395-396 process 404-410 rolled Rolled rivets rivet head Root 405 409-410 405 409-410 409-410 281-282 503-504 808-811 fastening area 281-282 reforming 503-504 surface finish wedge angle 802 808-811 Rotating blades (damage and refurbishment) 201-307 blade inlet-edge erosion damage 206-241 blade trailing-edge erosion 255-257 corrosion effects 281-291 erosion-resistant coatings 264-270 erosion-shield cracks 254-255 fretting corrosion 304 massive-particle damage 273-280 moment weighing (refurbished blades) 242-253 references 305-307 rotating-blade refurbishment 291-303 solid-particle erosion by oxide scale 257-264 This page has been reformatted by Knovel to provide easier navigation. 802 Index Terms Links Rotating blades (damage and refurbishment) (Cont.) solid-particle peening 271-273 steam-path cleaning 204-206 water induction 303-304 Rotating-blade corrosion blade row 281-291 442 corrosion effects 281-291 corrosion fatigue 289 corrosion pitting 287-289 fretting 304 stress-corrosion cracking 287 Rotating-blade erosion 257-258 inlet edge 257 tip-section 258 Rotating-blade refurbishment erosion damage/repair/control 291-303 291-292 forming new tenons 303 new units 292 older units 292 vane-tip cracks 299-303 vane weld repair (tie-wire hole) 293-299 Rotating components (damage and refurbishment) 304 309-548 blade-root steeples/wheel rim 480-497 components 309-362 coverband damage/repair/refurbishment 362-410 fusion techniques 427-433 induced bends (rotor) 464-479 references 544-548 rotor-damage mechanisms 433-463 rotor-rim damage 498-507 This page has been reformatted by Knovel to provide easier navigation. 442 Index Terms Links Rotating components (damage and refurbishment) (Cont.) rotor weld repair 508-527 tie-wires damage/repair/refurbishment 410-426 weld-repair process 527-543 Rotor 327-362 433-479 508-527 539 bending 464-479 construction 328-337 corrosion 439-454 damage mechanisms 433-463 discs 342-362 forgings 337-342 functions 327-328 rim damage 498-507 surface 508-514 turbine 342-362 welded 332-336 weld repair 508-527 weld material 539 wheel 508 Rotor and weld material Rotor bending 539 464-479 bend-damage classification 469-470 causes 464-466 hardness checks 474-475 permanent 467-468 run-out checks 470-474 straightening options 475-479 stress relief 514-522 479 This page has been reformatted by Knovel to provide easier navigation. 498-527 Index Terms Links Rotor bending (Cont.) temporary 466-467 Rotor construction 328-337 built-up 331-332 hybrid 336-337 monobloc 330-331 welded 332-336 Rotor corrosion blade-tip seal leakage 439-454 444 corrodents 450-454 diaphragm leakage 443-444 disc/spindle interface 445-450 hydrogen sulphide 453-454 oxygen 452-453 rotating-blade rows 442 shaft-end sealing positions 445 sodium hydroxide 451-452 steam throttling 442-443 Rotor-damage mechanisms 433-463 axial rubs 459-463 burns 459-463 corrosion 439-454 low-cycle fatigue 457-458 rough/out-of-phase synchronization 437-438 short circuit of generator 433-437 stage wetness/moisture 454-456 Rotor forgings basic production borasonic examination 337-342 337-338 342 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Rotor forgings (Cont.) central-inspection boreholes 340-341 inspection during manufacture 338-340 Rotor-rim damage 498-507 load-bearing surface skimming 498-503 long-shank blades 505-507 rebuild by welding 498 reforming 503-505 Rotor surface 508-514 Rotor weld repair 508-527 attachment of forgings 522 524-527 cracks 509 522-527 stub attachment crack 508-509 522 surfaces 508-514 wheel Rotor wheel 508 514-522 508 514-522 blade-root region 516-517 fillet radii 517-519 pressure-balance holes 517 wheel-forging attachment 522 Rubbing (coverband) 385-388 391-392 122 368-370 385-388 391-392 459-463 370 391-392 coverband 368-370 385-388 coverband surface 385-388 axial 368-370 370 coverband surface 385-388 radial 368-369 Rubbing/rubs axial This page has been reformatted by Knovel to provide easier navigation. 391-392 Index Terms Links Rubbing/rubs (Cont.) radial 368-369 rotor 459-463 Run-out checks (rotor bending) 470-474 S Sand-cast materials 642-643 Saturated-steam line 863 Saturated-water line 862-863 Scale particle 263 hardness 263 size 263 Scrap and replace decision 700-702 Screw attachment (coverband) 382-383 Seal knife-edge discharge forms 604-618 knife-edge form 605-607 multi-strip seal configurations 612-618 seal-location geometry 608-612 Seal-location geometry 608-612 Seal-maintenance economics 595-604 examples 597-603 kW loss prediction 599-604 Seals, glands, and sealing systems 549-651 labyrinth leakage 559-594 maintenance economics 595-604 steam-path seals 551-557 steam-sealing system functions 557-559 steam-seal leakage 559-567 Seal-strip forms 627-630 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Seal-strip insertion/securing 630-641 gland and seal-strip assembly Seal-strip materials alloy steel sand cast 634-641 641-643 643 642-643 Seal-strip/gland assembly 634-641 caulked strips 636-637 inserted segments 637-638 staked strips 635-636 Setback (inlet/trailing edge) 31-32 Setting angle (vane) 30-31 Set-up (welding) 66-68 Shaft end glands packing head sealing 15 445 Shield cleaning 231-232 Shield damage 208-210 433-437 Shrink fit (disc) 350-352 design 351-352 machining 350-351 dressing discharge diameters washing erosion 15 699 Short circuit (generator) Sidewall 568-471 15 13 Side-grip damage (blade root) 445 568-571 Shells (casing) Shipment of products 641 488-492 32 52-54 78-81 32 52-54 This page has been reformatted by Knovel to provide easier navigation. 78-81 Index Terms Links Snubber/integral connection (tie wires) 414-418 repair Snubber repair (tie wires) 425 425 Sodium hydroxide 451-452 Solid-particle erosion (oxide scale) 257-264 preventive/corrective actions 259-264 rotating-blade inlet edge 257 rotating-blade tip-section suction face 258 tenons attaching coverband Solid-particle erosion (SPE) oxide scale Solid-particle impact (SPI) 425-426 258-259 46-50 153 273-280 365-366 257-264 40-46 heavy damage 44-46 light damage 41-42 medium damage 43-44 Solid-particle peening 271-273 273-280 363-364 697 Spatial relationships 724 Special processes 665 675-676 725 774 reverse engineering 675-676 specification 665 vane 774 Specification (design) 257-264 663-665 assembly/alignment 665 material 664 nondestructive tests 665 physical dimensions 664 special processes 665 surface finish 665 724 724 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Specific heats (steam) 839-841 ratio of specific heats Specific volume (water and steam) Spring loading (gland ring) Stage-discharge area/angle (determination) 841 816-822 621 88-96 adjustment methods 93-96 material removal 95-96 vane bending 95-96 weld buildup 95-96 Stage-hardware requirements 812 Stage wetness/moisture damage 454-456 Staggered seal configuration 615-617 Staked strips (gland/seal strip) 635-636 State change (heat addition) Stationary-blade damage repair Stationary-blade repair Stationary-blade row components casings 834 154-162 154-162 154-162 1-15 2 definitions 5-15 diaphragms 2-12 Stationary-blade row geometry 15-35 dimensional requirements 28-35 three-dimensional 20-23 two-dimensional 16-20 vane profile 25-28 vane tilt 23-25 Stationary blades damage definition 1-35 4-5 154-162 154-162 5-6 This page has been reformatted by Knovel to provide easier navigation. 13-15 Index Terms Links Stationary blades (Cont.) repair row components row geometry Stationary components (damage and refurbishment) adjustment computation 154-162 1-15 15-35 1-199 96-121 casing components 162-173 casing operating problems/repair methods 174-198 components 1-15 diaphragm inner web correction 141-148 diaphragm sidewall repair methods 133-141 diaphragm thermal distortion 121-133 diaphragm-vane repair horizontal-joint weld repair operating phenomena outer-ring damage references 59-87 153-154 36-59 148-153 199 stage-discharge area/angle 88-96 stationary-blade row geometry 15-35 stationary-blade damage Stationary/rotating seal configuration 154-162 614 Steam discharge angle 30 Steam-discharge angle 108-109 adjustment 108-109 Steam expansion 842-852 adiabatic 869 847-850 constant pressure 845 constant volume 845 general case 846 throttling 757-758 850-851 This page has been reformatted by Knovel to provide easier navigation. Index Terms Steam-flow distribution Links 260 Steam forces (steam seal) 624-627 Steam formation 834-835 Steam heat 833-852 Steam-inlet angle Steam leakage 757 559-594 enthalpy-entropy (Mollier) diagram) 865-870 intermediate-seal location 880-882 rotating-blade row 878-880 seal 559-594 stationary-blade row 877-878 Steam-path cleaning 204-206 blast cleaning 205 hand cleaning 206 water washing 205-206 Steam-path expansion/mixing 842-852 expansion 842-852 moisture removal 881-882 steam leakage (intermediate-seal location) 880-882 steam leakage (rotating-blade row) 878-880 steam leakage (stationary-blade row) 877-878 Steam-path seals 877-882 149-153 face damage 149-153 general requirements 554-556 impulse unit 578-581 leakage 559-594 nuclear units 557 reaction unit 581-585 sealing system 557-559 869 551-594 This page has been reformatted by Knovel to provide easier navigation. 877-882 Index Terms Links Steam properties 859-873 Carnot cycle 870-872 Rankine cycle 872-873 steam tables 859-860 temperature-entropy diagram 861-865 Steam-seal face damage metallic inserts 149-153 152-153 Steam seal (gland ring) 623-627 Steam-seal leakage 559-594 calculation 565-567 Martin method 567-594 Steam-sealing system 557-559 Steam tables 859-860 Steam throttling (corrosion) 442-443 Steam-turbine blades (manufacture/inspection requirements) 711-813 blade-manufacturing processes 726-749 blade-pitch errors (factors) 803-811 blade-root tolerances 775-802 manufacturing techniques 723-726 passage swallowing capacity 770-774 profile and cascade tolerances 749-764 profile and placement errors 765-770 radial alignment (rotating blades) 712-723 references 813 special processes (vane) 774 stage-hardware requirements 812 Steam-turbine components (quality assurance) 653-710 available quality-assurance program 704-705 definition of quality 658-659 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Steam-turbine components (quality assurance) (Cont.) definitions of performance 660-663 design specification 663-665 engineering review 680-682 inspection and test plan 688-689 machining of components 705-709 nonconforming situations 699-704 product surveillance 690-699 purchaser assurance of quality 689-690 quality-assurance manual 679-680 quality-assurance program 677-679 683-687 quality-assurance program responsibility/administration references 683-687 710 responsibility for quality 657-658 reverse engineering 666-677 Steam-turbine technology xii Steam turning angle 758-759 Stitching repair 181-184 Straightening (rotor bend) 475-479 bending 476 peening 477 remachining 477-478 stress relief 476 thermal 476 Straight-through seal configuration 613-614 Stress-concentration centers 481-482 Stress-corrosion cracking 287 This page has been reformatted by Knovel to provide easier navigation. 704-705 Index Terms Stress relief Links 82-83 540-541 479 heat-affected zone 541 rotor straightening 476 tempering Stress (tie wires) 82-83 411-413 batch discontinuity 413 batch twist 413 centrifugal 411 compressive 411 misalignment 411-412 thermal 412 vibration 411 Stub-attachment crack Stub shaft attachment Suction-face finishing Superheated steam ice conversion 508-509 522 530 537-538 530 86-87 836-837 864-865 836-837 Supersaturation (steam) 851-852 Supplier facility inspection (preparation) 691-693 component operating modes (changes) 693 design-review meeting minutes 693 experience with supplier 693 inspection and test plan 692 operating experience 692 purchasing/requisitioning engineer 692 technical-purchase specification 692 Supplier-purchaser control 479 698-699 This page has been reformatted by Knovel to provide easier navigation. 476 Index Terms Surface curvature Surface finish coverband curvature integrity Links 28 665 724 802 28 385-388 508-514 529 531 665 707-709 724 802 28 665 707-709 724 802 385-388 707-708 machining 542 707-709 rebuild 529 531 requirements 724 rotor specification Surface finish (machining) cutting fluids integrity machining rates Surface integrity built-up edge retention 508-514 665 542 707-709 709 707-708 709 707-708 707-708 surface burning 707 tears and gouges 707 tool chatter 707 Surface rebuild 707-709 529 Surface rubs (coverband) 385-388 Surface skimming 498-503 axial-entry roots 502-503 blade replacement 501-502 blade reuse 499-501 Synchronization (rotor) 437-438 531 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links T Tangential-entry direction (blade root) 788-796 Tangential position determination (blade) 247-248 blade listing 247 blade placement 248 blade selection Technical-purchase specification 247-248 692 Temperature-changing cycle (entropy) 853-854 Temperature-entropy diagram 861-865 critical point latent heat lines of constant superheat 863-864 863 864-865 moisture content 864 saturated-steam line 863 saturated-water line 862-863 Temperature (water and steam) 822-824 absolute temperatures 821-822 Tempering/stress relief 82-83 Tenon and hole requirements (coverband) 400-401 Tenon corrosion 283-287 398-399 Tenon erosion 241 258-259 Tenon failures 395 Tenon forming 303 new 303 reforming 371 Tenon-hole form Tenons 371 393-394 400-401 241 258-259 283-287 303 371-381 393-396 398-410 722-723 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Tenons (Cont.) corrosion 283-287 398-399 erosion 241 258-259 failures 395 forming 303 hole form/requirements 393-394 over-riveting 395-396 reforming 371 reworking 371-372 riveting 395-396 weld deposit 372-373 weld rebuild 373-381 Thermally-hardened inlet edges Thermal straightening (rotor) stress relief Thermal stress Thermal transients Thermodynamics of water and steam 476 412 58-59 813-883 entropy of steam 852-857 gas equations 831-833 heating and expansion of steam 833-852 physical properties 816-830 883 reversibility 857-858 steam properties 859-873 increased reduced 479 479 873-882 Thickness (coverband) 402-410 212-213 basic power cycles references 400-401 366-368 368 366-368 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Three-dimensional considerations (stationary-blade row) 20-23 vanes (constant-but-reducing profile) 21-22 vanes (constant section) 20-21 vanes (twisted profile) 22-23 Throat opening Throttle-controlled units Throttling (steam expansion) non-expansive expansion 756-757 34-35 850-851 868-869 Thumbnail crack (vane-discharge tail) 58 Tie-wire damage/repair/refurbishment 410-426 forming/connection methods 414-418 repair considerations 418-426 stresses 411-413 Tie-wire forming/connection 414-418 continuous connection 414-416 snubber/integral connection 414-418 Tie-wire hole hole plugging/re-drilling Tie wires 868-869 282-283 293-299 296-299 282-283 293-299 317-326 410-426 batching with coverbands 323-326 cross sections 320-323 damage/repair/refurbishment 410-426 functions 317-318 forming/connection 414-418 forms 318-320 tie-wire hole 282-283 293-299 Tilt (vane) 23-25 31 angle 31 axial 23-25 This page has been reformatted by Knovel to provide easier navigation. 719-723 Index Terms Links Tilt (vane) (Cont.) compound radial tangential Turbine-rotor discs 24 23-24 342-362 disc removal 356-359 disc rim 348-349 disc-spindle assembly 352-356 disc-to-hub fillet radii 350 functions 343 forms 343-350 interstage seals 346-348 keyways and securing 358-362 machining for shrink fits 350-351 pressure-balance holes 349-350 shrink-fit design 351-352 Twisted profiles 22-23 Two-dimensional considerations (stationary-blade row) 16-20 742-743 U Unpredictable phenomena (casing) 176 V Vane adjustment 62-63 Vane bending 95-96 Vane cross-section irregularities 32-35 extended vanes 33-34 nozzle box 32-33 throttle-controlled units 34-35 This page has been reformatted by Knovel to provide easier navigation. Index Terms Vane-discharge tail cracking thumbnail crack Links 56-58 58 Vane dressing 78-81 Vane extension 33-34 Vane extrusion 729-731 Vane inclination 806-807 Vane material Vane mismatch errors 262 115-120 small discharge edge 118-120 small inlet portion 117-118 Vane pitch (P) Vane placement (root platform) Vane profile 29-30 805-806 20-23 constant-but-reducing 21-22 constant section 20-21 discharge tail 27 inlet nose 26 surface curvature/finish 28 twisted Vane projection across half joint 25-28 22-23 120-121 Vane repair methods 64-87 metallic inserts 84-86 suction-face finishing 86-87 weld buildup 64-84 Vanes 747 293-299 7-8 20-35 56-58 62-87 95-96 115-121 262 293-303 719-723 729-731 747 805-808 adjustment 62-63 bending 95-96 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Vanes (Cont.) cross-section irregularities 32-35 discharge-tail cracking 56-58 extension 33-34 extrusion 729-731 inclination 806-807 material mismatch errors pitch placement profile projection across half joint 262 115-120 29-30 805-806 20-23 64-87 setting angle 30-31 sidewall dressing 78-81 tilt 23-25 twist/shape variations weld repair 31 719-723 31 719-723 808 293-299 30-31 Vane tilt 23-25 angle 31 axial 23-25 24 tangential 23-24 Vane-tip cracks 299-303 Vane twist/shape variations 293-299 299-303 Vane setting angle compound radial 25-28 120-121 repair methods tip cracks 747 808 Vane weld repair (tie-wire hole) 293-299 blade removal/access 295-296 hole plugging/re-drilling 296-299 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Vane weld repair (tie-wire hole) (Cont.) welding in situ 295 welding integral snubbers 299 Vapor deposition 266 Vibration stress 411 Viscosity (water and steam) 829-830 Vortex profiles 742-743 W Washing erosion 454-456 Water and steam thermodynamics 813-883 basic power cycles 873-882 entropy of steam 852-857 gas equations 831-833 heating and expansion of steam 833-852 physical properties 816-830 references 883 reversibility 857-858 steam properties 859-873 Water-impact damage 364-365 Water-induced damage 50-56 205-206 454-456 497 centrifuged moisture 54-56 sidewall washing 52-54 washing erosion 454-456 water-impact damage 364-365 water washing 205-206 worming/wire drawing Water washing erosion 497 364-365 497 454-456 497 205-206 454-456 497 454-456 497 51-52 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Weld attachment 209-210 213-230 522 524-527 inlet edge/base material buildup 224-230 shield 209-210 forgings to rotor solid-bar inlet nose Weld-attaching inlet nose 522 213-224 213-216 final moment weighing 224 inlet-edge profile finishing 222-223 post-weld heat treatment 220-222 weld inspection 223 welding process 220-221 welding process preparation 217-219 64-84 copper plugs/backing strips 70-73 dimensional requirements before adjustment 83-84 diaphragm 66-68 initial weld deposit 74-77 layout/measurements 95-96 65 initial set-up inlet edge 524-527 213-224 blade vane preparation Weld buildup 431-433 224-230 65-66 material removal 69 required measurements 69 stress relief/tempering 82-83 vane/sidewall dressing 78-81 weld metal/filler materials 77-78 weld preheat 73-74 69 This page has been reformatted by Knovel to provide easier navigation. 224-230 Index Terms Weld deposit Links 74-77 235-238 372-373 540 tenons 372-373 Welding process 217-221 preparation 217-219 rotating components 527-543 Welding repair 527-543 180-181 508-543 attachment of forgings 522 524-527 deep cracks 509 522-527 options 529-530 process 527-543 rotor surfaces 508-514 stub attachment crack 508-509 522 508 514-522 64-84 95-96 135-137 139-141 180-181 209-210 213-230 233 235-238 293-299 372-381 431-433 498 508-543 209-210 213-230 522 524-527 64-84 95-96 224-230 235-238 372-373 wheel Welding attachment buildup cast iron 139-141 construction 60-61 deposit 74-77 540 in situ 295 inspection 233 integral snubbers 299 metal/filler materials 431-433 77-78 This page has been reformatted by Knovel to provide easier navigation. Index Terms Links Welding (Cont.) preheat 73-74 process 217-221 527-543 rebuild 135-137 498 repair 180-181 508-543 repair options 529-530 rotating components 527-543 set-up 66-68 tenons 373-381 Weld inspection 233 Weld metal/filler materials 77-78 Weld preheat 73-74 Weld rebuild 135-137 diaphragm sidewall rotor rim tenons Weld-repair options crack filling 135-137 498 373-381 529-530 529-530 rebuilding surface 529 rebuilding wheel 529 stub-shaft attachment 530 Weld-repair process (rotating components) 527-543 damage examination 531-538 finish machining 541-543 material removal 539 preheating 540 procedure evaluation 530 repair options rotor and weld material 373-381 529-530 539 This page has been reformatted by Knovel to provide easier navigation. 498 Index Terms Links Weld-repair process (rotating components) (Cont.) stress relief weld deposit Wheel gouging Wheel rebuild 540-541 540 494-496 529 Work done (steam expansion) 845-851 adiabatic expansion 847-850 expansion at constant pressure 845 expansion at constant volume 845 general case 846 throttling Worming/wire drawing 531-535 850-851 51-52 This page has been reformatted by Knovel to provide easier navigation.