General Disclaimer

One or more of the Following Statements may affect this Document

This document has been reproduced from the best copy furnished by the
organizational source. It is being released in the interest of making available as
much information as possible.

This document may contain data, which exceeds the sheet parameters. It was
furnished in this condition by the organizational source and is the best copy
available.

This document may contain tone-on-tone or color graphs, charts and/or pictures,
which have been reproduced in black and white.

This document is paginated as submitted by the original source.

Portions of this document are not fully legible due to the historical nature of some
of the material. However, it is the best reproduction available from the original
submission.

Produced by the NASA Center for Aerospace Information (CASI)

JPL PUBLICATION 82-99 DOEIET-12548.16
Distribution Category UC-88

Automation of the Longwall

Mining System
W. Zimmerman
R. Aster
J. Harris
J. High

November 1. 1"

Prepared for
U.S. Department of Energy
through an Agreement with
National Aeronautics and Space Administration
by
Jet Propulsion Laboratory
California Institute of Technology
Pasadena. California
Prepared by the 1a Propulsion Laboratory, California Institute of Technology,
for the U.S. Department of Energy through an agreement with the National
Aeronautics and Space Administration.

This report was prepared as an account of work sponsored by an agency of the

United States Government. Neither the United States Government nor any
agency thereof, nor any of their employees, makes any warranty, express or
implied, or assumes any legal liability or responsibility for the accuracy, com-
pleteness, or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by trade
name, trademark, manufacturer, or otherwise, does not necessarily constitute or
imply its endorsement, recommendation, or favoring by the United States
Government or any agency thereof. The views and opinions of authors
expressed herein do not necessarily state or reflect those of the United States
Government or any agency thereof.
 a

ABSTRACT

The longwall automation study presented in this document is the first

phase of a study to evaluate mining automation opportunities. A similar
analysis on the room and pillar system will be conducted in the second phase
of the automation effort. The mining automation evaluation is part of an
overall effort to identify and develop innovative underground coal extraction
systems.

The objective of the study was to identify cost-effective, safe, and

technologically sound applications of automation technology to underground
coal mining. The longwall analysis commenced with a general search for
government and industry experience of mining automation technology. A brief
industry survey was conducted to identify longwall operational, safety, and
design problems. The prime automation candidates resulting from the industry
experience and survey were: (1) the shearer operation, (2) shield and
conveyor pan-line advance, (3) a management information system to allow
improved mine logistics support, and (4) component fault isolation and
diagnostics to reduce untimely maintenance delays. A system network analysis
indicated that a 40 X improvement in productivity was feasible if system
delays associated with all of the above four areas were removed. A technology
assessment and conceptual system design of each of the four automation
candidate areas showed that state-of- the-art digital computer, servomechanism,
and actuatc,r technologies could be applied to automate the longwall system.
The subsequent health and safety evaluation of the automation candidates
projected: (1) a major improvement in reducing expos ire to respirable coal
dust by removing the face crew to a remote location, and (2) a possible
overall reduction in disabling injuries of 23%. The final cost benefit
analysis of all of the automation areas indicated a total net national benefit
(profit) of roughly $200 million to the longwall mining industry if all
automation candidates were installed. This cost benefit represented an
approximate order of magnitude payback on the research and development (RED)
investment.

In conclusion, it is recommended that the shearer operation be automated

first because it provided a large number of other sensor inputs required for
face alignment (i.e., shields and conveyor). Automation of the shield and
conveyor pan-line advance is suggested as the step since both the shearer and
face alignment operations contributed the greatest time delays to the overall
system downtime. Therefore, automation of these areas first would allow a
more rapid payback on the R&D investment. It is finally suggested to complete
the total automation effort with installation of management information and
fault isolation systems.

PRECEDING PAGE BLANK NOT F"ED

iii
 a.

FOREWORD

The JPL Advanced Coal Extraction Systems Project is part of a program to

develop advanced mineral extraction systems. Automation of underground coal
mining equipment is one research task under the Advanced Coal Extraction
Project. This document, the first of two reports resulting from the coal
automation task, provides the results of a longwall automation evaluation. The
second report will present the results of a room and pillar continuous mining
automation evaluation. The research was sponsored by the Division of Coal
Mining, United States Department of Energy, through an interagency agreement,
NASA Contract 7-918, with the National Aeronautics and Space Administration
(DOE Contract No. DE-AI01-76ET12548; Task RE 152, Amendment 90). Ralph
Avellanet, Deputy Director of the Divison of Coal Mining, is the project
officer.

ACKNOW:..*_DGMENTS

The authors would like to acknowledge the assistance provided by the

many individuals associated with the coal industry. Cooperating organizations
included Joy Manufacturing, U. S. Steel Research, Lee Engineering, Foster
Miller Associates, Skelly and Loy Inc., Carbon Fuel Coal Company, and Utah
Poker and Light (the Wilberg and Deer Creek Coal Companies).

iv
 CONTENTS

I. INTRODUCTION AND SUMMARY. . . . . . . . . . . . 1-1

A. OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . 1-1

B. ASSUMPTIONS AND CONSTRAINTS . . . . . . . . . . . . . . . . . 1-1

C. SUMMARY OF PREVIOUS WORK ON LONGWALL AUTOMATION . . . . . . . 1-2

D. SUMMARY Or EVALUATION APPROACHES AND RESULTS OF ANALYSIS . . 1-4

E. STRUCTURE OF DOCUMENT . . . . . . . . . . . . . . . . . . . . 1-6

II. DESCRIPTION OF THE CONTEMPORARY LONGWALL SYSTEMS. . . . . . . . . . 2-1

A. OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1

B. SHEARER OPERATION . . . . . . . . . . . . . . . . . . . . . . 2-2

C. ROOF SUPPORT OPERATION . . . . . . . . . . . . . . . . . . . 2-3

D. FACE CONVEYOR OPERATION . . . . . . . . . . . . . . . . . . . 2-4

E. OTHER COMPONENTS . . . . . . . . . . . . . . . . . . . . . . 2-4

III. IDENTIFICATION OF AUTOMATION OPPORTUNITIES . . . . . . . . . . . . . 3-1

A. OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1

B. RESULTS OF THE MARSHALL SPACE FLIGHT CENTER (MSFC)

DESIGNEFFORT . . . . . . . . . . . . . . . . . . . . . . . . 3-1

C. RESULTS OF THE INDUSTRY SURVEY . . . . . . . . . . . . . . . 3-2

D. SYSTEM COMPLEXITY, POTENTIAL COSTS, AND PRACTICALITY . . . . 3-5

E. POTENTIAL HEALTH AND SAFETY IMPACTS . . . . . . . . . . . . . 3-E

F. SUMMARY OF LONGWALL AUTOMATION OPPORTUNITIES . . . . . . . . 3-7

IV. DESCRIPTION OF EVALUATION "TOOLS FOR RANKING

AUTOMATIONOPPORTUNITIES . . . . . . . . . . . . . . . . . . . . . . 4-1

A. OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1

B. NETWORK ANALYSIS . . . . . . . . . . . . . . . . . . . . . . 4--2

C. AUTOMATION FEASIBILITY AND DEVELOPMENT COST INDICATORS . . . 4-3

D. MINER ACCEPTANCE . . . . . . . . . . . . . . . . . . . . . . 4-5

v
 E. HEALTH AND SAFETY INDICATORS . . . . . . . . . . . . . . . . 4-5

F. COST-BENEFIT INDICATORS . . . . . . . . . . 4-6

V. RESULTS OF THE LONT4ALL AUTOMATION STUDY . . . . . . . . . . . . . . 5-1

A. OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . 5-1

B. RESULTS OF NETWORK ANALYSIS . . . . . . . . . . . . . . . . . 5-1

C. AUTOMATION TECHNOLOGY ASSESSMENT . . . . . . . . . . . . . . 5-16

D. MINER ACCEPTANCE AND HEALTH AND SAFETY IMPACTS . . . . . . . 5-23

E. THE COSTS AND BENEFITS OF AUTOMATION . . . . . . . . . . . . 5-32

VI. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . 6-1

A. OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1

B. SUMMARY OF RESULTS . . . . . . . . . . . . . . . . . . . . . 6-1

C. IMPLICATIONS OF RESULTS . . . . . . . . . . . . . . . . . . . 6-2

D. RECOMMENDATIONR . . . . . . . . . . . . . . . . . . . . . . . 6-4

RE: ERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1

APPENDICES

A. DETAILED AUTOMATION DEVELOPMENT AND PRODUCTION COSTS . . . . A-1

B. DETAILED DESCRIPTION OF THE IPEG (IMPROVED PRICE

ESTIMATION GUIDELI T ES) MODEL . . . . . . . . . . . . . . . . B-1

C. DETAILED ANALYSIS OF FUTURE LONGWALL MARKET PENETRATION . . . C-1

Figures

4-1. Example Network . . . . . . . . . . . . . . . . . . . . . . . 4-2

4-2. Servomechanism Data Flow Diagram . . . . . . . . . . . . . . . 4-4

5-1. Longwall System Functioaal Network . . . . . . . . . . . . . . 5-3

5-2. Flow Chart to Project Automation Impacts on

Operating Time and Productivity . . . . . . . . . . . . . . . 5-7

5-3. Shearer Ranging Arm Articulation Flow Diagram . . . . . . . . 5-18

5-4. Shearer Cowl Articulation Package Flow Diagram. . . . . . . . 5-19

W

vi
a
 5-5. Shearer Face Alignment Flow Diagram . . . . . . 5-19

5-6. Face Conveyor Alignment Flow Diagram. . . . . . 5-20

5-7. Shield Advance Control Flow Diagram . . . . . . . 5-20
5-8. Monitoring/Control Flow Diagram . . . . . . . . . 5-22
5-9. Aut• , Fault Isolation Flow Diagram . . . . 5-22
5-10. Steps in Coat-Benefit Analysis. . . . . . 5-35
5-11 Seam Heights in USA Longwall Panels . . . . . . . . . . . 5-40

C-1. Longwall tartups in the USA . . . . . . . . . . . . . . . . . C -5

C-2. Market Curve Fit Variance as a Function of Market
Penetration by 1995 . . . . . . . . . . . .. . . . . . . C-6

Tables

1-1. A Summary of Total Longwall Automation Benefits . . . . . . . 1-7

5-1. Longwall System Functional Network Key. . . . . . . . . . . . 5-4
5 -2. Initial Projection of Production Time Savings . . . . . . . . 5-8
5-3. Expert Adjustment of Projected Automation Impacts on
Production Delays . . . . . . . . . . . . . . . . . . . . . . 5-10
5-4. Estimated Longwall Automation Development and
Production Costs . . . . . . . . . . . . . . . . . . . . . . . 5-24

5-5. Average Yearly Longwall Disabling Injuries (DI)

by Task/Hazard and Accident Class . . . . . . . . . . . . . . 5-27
5-6. Hazards Potentially Affected by Automation Incorporation. . . 5-28
5-7. Results of the Longwall Automation Health Analysis. . . . . . 5-28
5-8. Longwall Disabling Injury Projection Considering
Automation Opportunities . . . . . . . . . . . . . . . . . . . 5-30
5-9. Steps and Average Timeframes for MSHA lipproval. . . . . . . . 5-33
5-10. FCR For Equipment . . . . . . . . . . . . . . . . . . . . . . 5-36
5-11. Underground, Longwall, and Potential Longwall
Market Tonnage Projections . . . . . . . . . . . . . . . . . . 5-38
5-12. Underground Longwall Annual Markets . . . . . . . . . . . . . 5-39
5-13. Baseline Annual Section Production Rates. . . . . . . . . . . 5-39

Vii
W,

5-14. Summary of Productivi,,:y Increases . . . . . . . . . . . . . . 5-42

5-15. Maximum Possible Benefits . . . . . . . . . . . . . . . . . . 5-42
5-16. Costs of Automation . . . . . . . . . . . . . . . . . . . . . 5-43
5 -17. Net Benefits Per Section . . . . . . . . . . . . . . . . . . . 5-43
5-18. Combiliations of Automation Opportunities. . . . . . . . . . . 5-44
5-19. Time Delays for Market Penetration . . . . . . . . . . . . . . 5-45
5-20. Diffusion of Automation Opportunities Into the
Longwall Coal Mining Industry . . . . . . . . . . . . . . . . 5-46
5-21. Present Value National Benefits . . . . . . . . . . . . . . 5 -47
5 -22. Annual Injuries in Longwall Coal Mines. . . . . . . . . . . . 5-48
5-23. Annual and Total Safety Benefits from Longwall
Automation . . . . . . . . . . . . . . . . . . . . . . . . . . 5-49
5 -24. Summary of Total Benefits . . . . . . . . . . . . . . . . . . 5-50
6-1. Ranking of Automation Options by Impact on
Coal Production . . . . . . . . . . . . . . . . . . . . . . . 6-2

6-2. Percentage Improvement Over the Baseline Productivity . . . . 6 -3

6-3. Suggested Automation Technology Development Sequence

Based on Impact on Ii:vestment Payback . . . . . . . . . . . . 6 -4
6-4. Ranking of Automation Options Based on Health and
Safety Impacts . . . . . . . . . . . . . . . . . . . . . . . . 6-5

6-5. Ranking of Automation Options Based on the Net Total

Cost Benefits (Nationwide) for the Years 1989 to 2000 . . . . 6-5
A-1. Total Common Element Costs . . . . . . . . . . . . . . . . . . A-3
A-2. Add-on Ranging Arm Articulation Cos*s . . . . . . . . . . . . A-3

A -3. Cowl Articulation Costs . . . . . . . . . . . . . . . . . . . A-4

A-4. Combined Costs of Ranging Arm and Cowl Articulation . . . . . A-4

A-5. Add-on Face Alignment Measurement Costs . . . . . . . . . . . A-5

A-6. Add-on Face Alignment Control Costs . . . . . . . . . . . . . A -5

A-7. Add-on Shield Advance Automation Costs. . . . . . . . . . . . A-6

A-8. Add-on Face Alignment/Shield Advance Costs. . . . . . . . . . A-6

viii
 ^-^-,• ate..
a-.:..-^rO-F^.... .=°b'F,'f''"l:-'z-;=

A-9. Add-on Auto Fault Isolation Costs . . . . . . . . . . . . A-7

A-10. Add-on Monitoring/Control System Costs. . . . . . . . . . . . A-7

B-1. IPEG Coal Mine Coefficients, 15% ROB. . . . . . . . . . . 3-5

B-2. IPEG Coal Mine Coefficientsp 25% ROE. . . . . . . . . . . . . B-6

B-3. IVEG4 Engineering, Financial, and Operational Inputs. . . . . 3-7

C-1. Historical Longwall Market Share * . . . . . . . . . . . . . . C-4

ix
 SECTION I

INTRODUCTION AND SUMMARY

A. OBJECTIVES

The Jet Propuls:.on Laboratory (JPL), sponsored by the U.S. Department of

Energy (DOE), Division of Coal Mining, was asked to identify and develop cost-
effective and technologically sound applications of automation technology to
underground coal mining. Ultimately, plans are to be developed that provide a
step-by-step approach for automating longwall and room and pillar continous
mining systems that are practical and useful to the mining industry. This
report evaluates and ranks several automated functions and provides an estimate
of benefits and costs of technology development.

To meet the objectives, the study used: (1) the longwall mining
guidance and control system work conducted by the NASA Marshall Space Flight
Center (MSFC) for DOE; (2) existing automation work done by the mining
industry on longwall, continuous miner, roof bolter, and shuttle car systems;
(3) the system modeling and evaluation tools developed by JPL under the
Advanced Coal Extraction Systems Project; (4) automation technology advances
developed at JPL; and (5) an assessment of useful automation and remote
control technology advances available in other industries. This document, the
first o" two external reports, identifies the n..>st likely automation areas for
a longwall mining system.

The following sections discuss the major assumptions which form the
foundation of the study, the constraints under which the study was performed,
and a summary of the rEsults of the analysis. Finally, a brief statement of
the general structure of the document is provided to assist the reader.

B. ASSUMPTIONS AND CONSTRAINTS

Although several di:ferent approaches were used to establish the best

opportunities for automating the longwall system, the final ranking of the
automation areas and measure of benefit to industry was bared on an economic
analysis. The cost benefits were calculated in 1982 dollars. It was assumed
that all supporting reference material for the cost benefit calculations
obtained from the Department of Energy, other government agencies, and
industry was reasonably accurate. In addition, it was assumed that once the
automation candidates were selected, a successful R&D effort would follow so
that eventually all the automation opportunities would be incorporated by
industry. This was a necessary assumption in order to: (1) realize the total
cost and cost benefit fir automating the complete system, and (2) retain
continuity of the analysis and not attempt tc second-guess each mine
operator's equipment configuration preferences.

In addition, several ground rules evolved from sponsor requests and time
and budget considerations. First, the DOE Division of Coal Mining requested
that the longwall study utilize the production projections provided by the
Energy Information Agency in 1981. Second, the sponsor indicated a desire •o
limit the market impact analysis to the U.S. coal. industry. Third, although

1-1
s
 a

the user requirements revealed a need for automation concepts which encom-
passes state-of-, the-art developments as well rs evolving technology, the
sponfsor requested that the study only consider candidates which utilize
exiEting technology and could be incorporated in the near term (i.e.,
1984-1986 time frame).

The user requirements developed to identify the desired automation areas

depended partly on a survey which included equipment manufacturers and
designers, mine managirs, equipment operators, and maintenance personnel.
Although it would have been desirable to have interviewed more individuals in
each group, project funding and milestone commitments constrained the scope of
this survey. Although the number of individuals interviewed was limited, the
consistency of the answers increased our confidence in :he results.

The final major constraint revolved around the longwall network analysis
developed es a tool to assess potential automation impacts on productivity.
Again, project funding level and milestone commitments prevented the develop-
ment of a longwall probabilistic PERT network similar in sophistication to the
current KETRON Room and Pillar Critical Path Model. Because of the lack of an
equivalent longwall critical path model, a deterministic network was developed
using KETRON longwall industrial engineering data. The resulting network
provided adequate detail to establish the critical path and amxess the impacts
each automation candidate would have on productivity.

C. SUMMARY OF PREVIOUS WORK ON LONGWALL AUTOMATION

The potential for automating longwall mining components has been widely
examined from both a micro and macro system viewpoint. For example, at a
micro level research has been conducted toward the development of proper
motion and spatial sensors for guiding and controlling the longwall shearer.
At the macro level, several studies have bee% completed which examine the
feasibility and productivity benefits of automating the longwall mining
system. This section provides a brief review of some of the well-known
technical developments and studies at both the micro and macro system levels.

Mining industry experience has indicated several areas where automation

could enhance longwall machine operating efficiency. One major area of
concern has been the probiem of controlling the amount of coil remaining on
the roof after the shearer has made a cut. Of equal concern is the opposite
problem of cutting rock along with the coal. The Bendix and A.D. Little
studies of longwall system operation suggested that these areas were major
contributors to low face production (1, 2). Industry has responded to these
problems through several developments. One technological development that has
appeared promising is the natural radiation background sensor. This system
allows the coal thickness above the shearer to be measured (and maintained as
the shearer cuts) using the natural background radiation of the above strata
(3, 4). Other similar applications of technology include the use of acoustics
and the "sensitized pick" for seam thickness measurement and coal/rock
interface detection (S, 6). The whole thrust of the above technologies, has
been to provide more accurate information about seam characteristics to the
shearer operator so that he can cut the maximum amount of coal with each
pass. One important aspect of plotting and controlling the path of the

1-2
shearer as it traverses the fake is locating its position as the seam
undulates. Lasers are presently being employed to allow the last cut to be
duplicated based on locating the shearer position relative to seam dips (7. 8).

Industry experience has also indicated that face alignment problems have
contributed to many conveyor and panline failures. One example of the mining
industry's response t, this problem is the yaw measurement sensor daveloped by
the Benton Corporation. This sensor measures angular deviations in the panline
and transmits information about the straightness of the face to the operator.
As an adjunct to all of the above research, the NASA Marshail Space Flight
Center Long% vall Program tested the performance of several shearer and conveyor
sensors, and then examined design problems associated with retrofitting the
shearer and conveyor with the most promising sensors.

In the area of longwall shield operation it appears that low coal

conditions inhibit the support operators from keeping pace with the shearer.
The remote shield operation system designed by Eickhoff is another good example
of applications of automation to the longwall system (9). A microprocessor
system, which stores sensor inputs on the position of the shearer loader,
transmits pulses to the shield control device to serially activate each set of
shields an the shea •:er passes (9). Up to 200 support units are controlled by
one microprocessor.

The above discussion on automation research at the micro (component)

level is not intended to cover all the developments in sensor, guidance, and
control technology. For the purpose of this study, it sufficed to simply
understand state-of-the-art component automation well enough in order to
establish conceptual designs for various automation options. Using tested
technology in the conceptual designs confirmed the feasibility of incorporating
varir ► s automation options and also allowed the research, development, and
capital investments to be approximated.

At the macro (system) level, several studies have been done on the
potential impact of longwall automation on productivity. Two major system
studies were performed by Skelly and Lev and CONINEC (10, 11). The results of
the Skelly and Loy study suggested that a 60% increase in productivity was
feasible if both the shearer and face advance systems were automated. The
productivity increase stemmed from: (1) a projected improvement in the
shearer *reverse speed (caused by not having the shearer and conveyor speed
paced by operator motility); (2) a reduction in face personnel, therefore
allowing the prodvictivity per man to increase; and (3) alleviation of trim cuts
caused by poor face alignment (10). The Skelly and Loy study also suggested an
environmental and equipment monitoring system be employed since there would be
no face personnel present to monitor methane emission and equipment failures
(10).

The COMINEC study similarly projected a 60% improvement in productivity

(11). As with the Skelly and Loy study, the COMTNFC analysis suggested major
increases in productivity thr.ugh shearer and shield automation. The CONIKEC
analysis also indicated that shearer efficiency could be increased by
incorporating a coal-rock interface detecting system (11). T ►, e COMINEC study
examined other sources of operational problems revolving around operators and
mine management. The resulting recommendations included: (1) an educational
program on equipment operation., and (2) development of a computer simulation to
optimize system design and predict productivity improvements (11).

1-3
 a,

It is interesting to note that the results provided in this document,

although derived independently, are essentially the same as both the Skelly
and Loy and COMINEC studies. The major differences between the results
provided in this document and the other studies stem from the depth of the
cost benefit and safety analyses. For example, this document considers:
(1) time of market penetration, ( 2) long - term national coal demands, and
(3) expected longwall growth as key inputs to the cost - benefit calculations.
Similarly, the safety analysis draws on detailed historial injury and worker
activity data to project the safety impacts. Overall, it appears that the
results provided in this document serve to enhance and extend the findings of
previous longwall automation studies.

D. SUMMARY OF EVALU:!TION APPROACHES AND RESULTS OF ANALYSIS

1. Summary of Evaluation Approaches

Several steps were required to identify potential areas for automa-

tion, establish the appropriate sensor, guidance, and control technology, and
determine productivity impacts and cost benefits, namely:

(1) Identification of automation opportunities.

(2) Network analysis.

(3) Automation technology assessment.

(4) Cost-benefit projections.

Zhe following discussion summarizes the approach used in each step of the
analysis.

a. Identification of Automation Opportunities. The

state-of-Lhe-art assessment consisted of developing functional descriptions
for each major component in the longwall system, followed by conducting an
industry survey of equipment manufacturers and designers, mine managers, and
miners. The functional description provided information on mining operations,
the present degree of system mechanization, and operations potentially
amenable to automation. The survey confirmed the system operational and
functional descriptions, and provided valuable information on problem areas
where automation could enhance productivity or improve miner health and
safety. These findings formed the basis for deterriining the automation
opportunities. The other considerations were: (1) the previous experience
obtained from the NASA Marshall Space Flight Center longwall sensor program
and automation experience; (2) consideration of p,^: p ntial costs, increases in
systc.,i complexity, and decrease: in reliability; and (3) health and safety
impacts based on the major histarical longwall hazards.

b. Network Analysis. A detailed 'longwall operational network

was assembled based on the functional analysis, industry inputs, and addi-
t t onal industrial engineering data obtained from the KETRON longwall study.
The industrial engineering data were crucial to understanding system delays

1-4
wh-ch detracted from production time, but had potential to be streamlined with
automation. Once the automation opportunities (developed from the considera-
tions summarized above) were reflected in the form of appropriate delays in
the network, an estimate of potential increase in productivity was obtained.
This estimate became a pivotal variable in the cost-benefit assessment.

C. Automation Technology Assessment. The technologies required

to impletent each automation opportunity were identified. This effort
included an investigation of appropriate sensor technology, development of
mathematical models for each affected system component to establish the
location and data feedback for each sensor, and conceptual design of guidance
and control systems. The technology assessment included preliminary cost
estimates and schedules for developing the various automation candidates.
Mitigation of potential health and safety hazards was also part of the
technology assessment. This was done by considering how each area autc.aated
would reduce worker exposure to hazards (e.g., through removing the worker
from a proven hazardous area or by providing more protection). A projection
of the expected reduction in serious injuries was then_ made and converted into
potential cost savings.

d. Cost-Benefit Projections. Costs and benefits were estab-

lished for each automation opportunity by comparing the projected productivity
improvements against capital and operating costs. Other variables derived
from health and safety impacts, market potential, and market penetration rates
were also factored into the final cost benefit calculation. The automation
opportunities were then prioritized to allow formulation of a development plan.

2. Results of Analysis

The automation candidates were derived by integrating the findings

of the NASA Marshall Space Flight Center longwall sensor development program,
the user needs as identified by the industry survey, considerations of
excessive costs or system complexity, and potential health and safety
improvements. The following resulting automation opportunities are summarized
as follows:

(1) Shearer automation (cutting, face aligment, coal-rock

interface detection, last cut memory, shearer arm
articulation, and tramming).

(2) Shield advance (remote operation or semi-automated).

(3) Conveyor and pan-line advance.

(4) Computer monitoring (information management for sources of

system delays).

(5) Preventive maintenance system (fault isolation and failure

diagnostics).

(6) Face sensing ahead of the shearer (to detect hard rock
partings).

1-5
 - IL

(7) Remote seam mapping (as related to sine planning and

development).

(8) Semi -automated movement of equipment components to a new

panel.

As stated earlier in Section I.B, one of the study guidelines was that
the longwall automation study consider only those opportunities which could be
incorporated in ti,e near term using state-of-the-art technology. The technol-
ogy assessment of both existing and evolving automation indicated that only
the first five of the above areas would fit the near term category. These five
areas (shearer automation, shield advance, conveyor/pan-line advance, computor
monitoring, and preventive maintenance) became the focus of the automation
study.

Once the sensor, guidance, and control technologies were identified and
conceptualized for each automation area, approximate development costs for
implementation in commercial longwall systems were estimated. The costs and
benefits were then calculated using the net dollar worth of improved produc-
tivity, minus the capital and uperating costs. Miner health and safety, the
other major areas impacted by automation, were also quantified where possible.
Table 1-1 provides a summary of the quantitative and qualitative benefits in
the areas of productivity, health, and safety.

Table 1-1 shows that the smart (remotely operated) shearer and shield/
conveyor opportunities are the most promising of the five automation options.
This is because the largest system delays are associated with the shearer,
shields, and conveyor, and these components have the largest impact on safety
and productivity. The above results do not imply that the remaining automa-
tion options should not be pursued. Once the appropriate sensors and data
retrieval systems are built into the shearer, shields, and conveyor components,
the benefits associated with the management information and fault isolation
options will also be realized. In conclusion, the recommended development
program for longwall automation has three thrusts:

(1) Pursue the development of sensors, data retrieval, guidance and

control technology for the smart (remotely operated) shearer,
since the shearer is a pivotal component which controls the
position of the shields and conveyor.

(2) Develop t'ne additional sensors (e.g., shield-to-shield alignment),

guidance, and control technology required to automate the shields
and conveyor pan-line advance.

(3) Uti1 4_ze the sensor feedback information from the face components
and add the desired additional sensors and feedback linkups for
the computer monitoring and fault isolation options.

E. STRUCTURE OF DOCUMENT

Section II provides an introduction to contemporary longwall system

components and operation through a detailed discussion of system architecture,
interaction of components during operation, and other support activities.

1-6
 Table 1-1. A Su ary of Total Longwell Automation Benefits
(1989-2000)

Safety
Productivity Injury
Opportunity ($1,000,000) (51,000900 Reductions) Health's

Smart (remotely-operated
Shearer 158 7.7 385 +

Automated Shearer 48 2.3 114 +

Shields and Conveyor 124 8.3 416 +

Computer Monitoring 28 0 0 0

Fault Isolation 0 0 0 0

Combined Automation
(all of the above) 199 13.7 683 +

Smart Shearer/Shields/
Conveyor 193 15.1 753 +

Smart Shearer/Shields/
Conveyor/Computer
Monitoring 226 15.1 753 +

aA plus (+) indicates an improvement in health conditions by reducing worker

exposure to dust.

This information forms the basis for understanding the key areas where
automation could enhance system operation. Section III develops the automation
opportunities based on NASA Marshall Space Flight Center longwall experience,
industry survey, cost and complexity considerations, and health and safety
impacts. Once the basic system components, operations, and automation areas
are identified, the productivity, technology assessment, health and safety, and
cost benefit aspects of the study are developed. In preparation for presenta-
tion of the final study results, Section IV briefly introduces the evaluation
tools used in each aspect of the study to allow the ultimate ranking of automa-
tion opportunities. Section V then provides the detailed results of: (1) the
network and productivity analysis, (2) the automation technology assessment,
(3) health a-id safety evaluations, and (4) the cost benefit analysis. Section
VI presents a recommended plan for implementing the automation opportunities,
based on the results presented in Section V, and also provides concluding
comments on the overall study with a detailed summary of the results.
Supporting data for the anal y sis are provided in the Appendix.

1-7
 SECTION II

DESCRIPTION OF THE CONTEMPORARY LONGWALL SYSTEM

A. OVERVIEW

The longwall coal mining method is a combined system of excavation, roof

support, and bulk material handling across a wide coal seam face (12).
Although the method can be adapted to seams sloped as much as 45 deg or more,
the usual application in the United States is in flat or nearly flat coal
seams. Given a good match between mining conditions and selection of mining
equipment, the longwall can produce as much as 1,000 tons per operating hour.
The long-wall section crew is ordinarily composed of 10 or 11 miners.

Longwall mining requires the development of a rectangular panel in the

coal seam, with the narrow side as the excavation face. Face dimension varies
from 400 to 800 ft. The longer side of the panel, running between sub-main or
main entries, ranges from 2,500 to 4,000 ft in length (12). Preparation of
the panel requires end entries for starting and ending the extraction, and
multiple side entries to accomodate worker access, ventilation, coal removal,
and supplies. Panel entry patterns and supporting pillars vary according to
the mining conditions encountered (13). In America, the panel is usuall-•
mined in a retreat mode. Retreat mining allows the roof to cave behind the
protected face line of the longwall. The roof is supported mechanically with
hydraulic self-advancing support structures (shields) immediately behind the
coal excavation line. The successful and orderly collapse of the immediate
roof strata behind the roof supports is critical to the safe and efficient
operation of the longwall. An array of mining system hardware, sub-systems
and equipment compcnents is available from both European and American firms.
The two well-developed longwall excavating systems presently in use are the
plow and the shearer. The plow is basically a line of blades which are dragged
bi-directionally across the face. The shearer is a cutting drum which is
moved across the face, cutting a web 28-30 in. deep. The cutting drums can be
mounted at each end of a chassis, and can either be rigidly set in a vertical
position or havo ranging capabilities that elevate and depress the drum.
Control of the horizontal speed of the shearer and the cutting height of the
drums is the responsibility of the shearer operator.

The extracted coal is dropped onto a continuous floor-mounted chain

conveyor haulage system. The conveyor structure also serves to support and
guide the excavator, and provide an anchor point for the roof support units.

In this discussion the shearer (rather than the plow) will b= used as
the excavating component because it presently represents the most widely used
longwall extraction device in America (12). The shearer described below
consists of a double drum type with ranging arms and will travel across the
face unidirectionally (since this configuration offers the most favorable way
of controlling dust at the face). The shearer, its companion conveyor, the
line of roof supports that protect the face area, and the crew constitute a
system which, when complemented with support systems, allows the longwall
system to operate. Inefficiency, poor productivity, injuries, dust problems,
and equipment breakdowns sometimes result from the cramped quarters and poor

2-1

u
visibility. The following sections examine each of the system components in
detail and provide the foundation for the automation considerations developed
later in the document.

B. SHEARER OPERATION

The dual-drum shearer is designed to do two things as it traverses the

longwall face: to cut coal from the seam in a 30 -in. de, vertical slice from
a selected upper limit to a lower limit near the bottom of the coal seam, and
to load the cut coal onto the chain conveyor (12). The shearer is operated by
converting electrical power from a transformer to a gear or hydraulic motor
drive. The dual-drum shearer is capable of varying travel and drum cutting
speeds.

The dual-drum shearer can cut coal in seams as thick as 18 ft. Ranging
arms, which elevate and depress the drums, enable the shearer to change drum
cutting postures, thereby allowing bi-directional movement. The drum leading
in the direction of shearer travel usually cuts to the upper limit while the
trailing drum cuts to the lower limit. These limits are determined and
adjusted by two operators, who accompany the shearer as it moves along the
face. Each operator also controls the position of the drum cowling, which
helps direct broken coal toward the face conveyor. In bi-directional practice,
the cowls are repositioned when the cutting direction is reversed. The follow-
ing discussion demonstrates shearer operation as it is commonly practiced in
the United States.

The operating pattern of the double-drum, shearer, and loader calls for
a cut across the entire face. Although the depth of cut is limited by drum
dimension, a shallower cut depth can be achieved by changing the position of
the face conveyor structure that supports the shearer. Detailed attention is
given to face alignment normal to the entries on either side of the panel as
well as linearity, because Severe stress is placed on the shearer and conveyor
components if the face is not straight (12, 13). In addition, production is
adversely affected if clean-up (alignment) cuts are necessarv. The operators
position the drums, cowls, and water sprays, prior to starting a face pass.
The major operator concerns are:

(1) The shearer travel rate as judged by the power draw, volume of
coal leaving the face, and coal lump size.

(2) Roof and floor cut limits and adjustments.

(3) Condition of the freshly exposed roof.

(4) The dust condition.

The shearer generally cuts and loads at 10-12 ft/min. The full face
pass ends when the leading drum breaks into the far entry (tailgate). At that
point, the operator backs off the shearer, lowers the leading drum, and cuts
the short bench segment of coal that the trailing drum could not cut. The
shearer then retreats (flits) toward the starting gate until it arrives at a
point some 50-100 ft from the starting entry (headgate). The trailing drum

2-2
is raised to the upper position and the shearer cuts the top coal into the
headgate entry to prepare for another pass. Meanwhile, the entire conveyor
pan-line is moved in position against the freshly established face in pre-
paration for another pass. These periodic shearer stops at the headgate or
tailgate offer the best opportunity for maintenance or inspection.

C. ROOF SUPPORT OPERATION

Roof support (shield) operation is a parallel activity of the longwall

system (14). While the shearer and the chain conveyor components provide for
the removal of the coal from the face, the roof supports allow the system to
retreat or advance. The supports shield the operating personnel and equipment
from roof falls.

The roof supports are hydraulically lowered, advanced, and raised

against the exposed roof by a support operator to allow the freshly exposed
roof to be supported as closely as possible to the working face. The shield
move is initiated after the shield canopy is lowered hydraulically. A
hydraulic ram fixed to the support base pulls the shield forward across the
floor. When the new face alignment is achieved, the operator raises the
canopy by activating the cylinders carefully to avoid disturbing the roof.
This series of steps can be initiated as soon as the shearer has passed the
shield (14).

The conveyor positioning maneuver is done by thrusting the pan

components forward in conjunction with shield advance after the shearer has
passed.

The rate of shield and conveyor repositiorings is determined by the rate

of shearer cutting and loading. This rate varies with seam height, efficiency
of the operation, and operator judgment. Shield movement time ordinarily
ranges from 8 to 10 s, and permits a support operator to maintain pace with
the shearer. In low coal (less than 48 in.), the face line may be worked by
two shield operators to reduce some of the physical stress of following the
shearer. Controls for each shield, or for a group of shields, are mounted on
the adjacent: unit on the supply air side, so that when the shield move is
started, the operator is placed in a safe area where there is less dust.

The su pport operator's console includes a hydraulic supply system

pressure gauge, a valve block having independent adjustable pressure relief
valves for each cylinder, main support activity valves, a face conveyor ram
valve and a valve for any canopy extension control. If mining conditions
require, the mine operator may incorporate optional capabilities and controls
to establish enhanced side shielding, face shielding, and horizontal steering
assistance.

Supports may be also placed at both ends of the pan-line in the gate
entries to protect power components such as the conveyor drive, the gear
reduction units, the head and tail pulley assemblies, and the chain tensioning
device. Often these additional supports are fitted with hydraulically
activated canopy extensions to provide greater protection at the critical
face-gate junctions. In some conditions, roof supports may be deployed fully
across the entries to protect against roof or rib collapses caused by stress
buildup in the overburden.

2-3
D. PACE CONVEYOR OPERATION

The chain conveyor is the unifying element in the longwall systes. It

guides the shearer and removes the sheared coal (12). It is positioned in its
working place by the roof support shields, and in turn provides anchorage for
moving the shields. The shearer operator often relies on the conveyor's
position to obtain a reading on face alignment. Also, the productivity of the
operation is gauged by how much coal the conveyor is carrying.

The chain conveyor unit is composed of a core component which is a wide

H-section pan, usually open on the bottom, and a septum which supports the
coal-carrying upper flight and guides the flight bars of the haulage chain.
The conveyor run is formed by joining all the individual pan units, each
approximately 5 ft long, so that each corresponds with the width of a roof
support shield. The pans are closely joined with couplings and pins that
maintain alignment while allowing a few degrees of horizontal and vertical
movement between pans.

In operation, the pans remain in their "cutting pass" position until the
shearer returns to the headgate. As soon as the shearer flits back toward the
headgate, the pan-line is snaked to its new position near the face, ready for
the next shearer cut cycle.

E. OTHER COMPONENTS

Several lesser subsystems complement the principal components of the

longwall mining system. The following sections describe these necessary
subsystems (12). The group excludes basic services such as electrical supply
and distribution, ventilation, transportation, water supply, and communica-
tions which are common to all underground mining systems.

1. Hydraulic Supply Sub-system

A constant supply of oil-water emulsion is provided to the

mechanized roof-support group by a skid-mounted power pack, usually placed in
the headgate. The pack serves as reservoir, emulsion controller, filter
station, and pump station for the closed loop that serves and interconnects
the roof shields. The hydraulic supply sub-system is a fail-safe design in
that the supports will remain locked in place if a supply pump or hydraulic
line fails.

2. Stage Loader

The stage loader is a short, heavy-duty chain conveyor positioned

perpendicular to the face at the headgate end of the face conveyor. The stage
loader accepts coal off the face conveyor and redirects the coal down the
adjacent entry toward an outby dumping point. The stage loader width and
travel rate are often greater than the face conveyor so that it can contain
coal surges from the face. The unit moves in step as the longwall shearer and
face conveyor are positioned.

2-4
 Because the stage loader is designed to contain coal surges, it offers a
convenient location for a coal breaker. The breaker provision is widely
accepted as the necessary means to size coal ahead of conveyor loading to
prevent belt damage. The lump breaker is positioned on the stage loader
astride the coal flow. Large lumps of coal are mechanically broken as they
approach on the loader pan.

3. Temporary Roof Support

The head and tail gate areas are generally zones of maximum roof
stress, and require placement of supplementary roof support to maintain roof
3
integrity and geometry. Temporary roof support is typically accomplished via
the use of hydraulic props, sometimes supplemented by "'tuber cribs. Cribs are
normally constructed by stacking timbers in a vertical box configuration
between the floor and ceiling. The hydraulic prop is a metal column which is
used as supplementary and temporary roof support. It is manually placed and
activated by a hand-pump. A range of sizes and load-bearing capacities
(commonly 10-40 tons) is available.

The preceding description was used in constructing a baseline operating

scenario for longwall systems in the United States. This baseline scenario
was employed in the identification of potential automation areas, development
of an operational network to assess productivity impacts of automation,
development of automation technical concepts, and cost-benefit projections.
Each of these areas is discussed in greater detail in the following sections.
 SECTION III

IDENTIFICATION OF AUTOMATION OPPORTUNITIES

A. OVERVIEW

The preceding section provided descriptions of the complete longwall

system, the operating cycle of each of the major components, and potential
system problems, and established a basis for structuring the approach to
identify and develop automation options. The first step was to examine
existing efforts to automate longwall systems, particularly the NASA Marshall
Space Flight Center (MSFC) longwall study, and utilize this research
experience to identify non-problem and problem areas. The second step was to
survey the mining industry to confirm initial estimates of problem areas and
obtain additional information on operation and maintenance problems that could
possibly be solved by automation. Potential longwall cost impacts and system
complexity were the next considerations for establishing automation opportuni-
ties. In the third and last step, potential automation opportunities for
reducing worker exposure to hazards associated with longwall systems were
identified. The results of these efforts are presented below.

B. RESULTS OF THE MARSHALL SPACE FLIGHT CENTER LONGWALL DESIGN EFFORT

1. Identification of Non-Problem Areas

Review of the MSFC longwall study suggested three main areas where
automation was not particularly advantageous. First, operators apparently do
not have a problem with roll. This stems primarily from the relatively slow
rate of advance, thus allowing sufficient time to compensate, and the fact
that some existing equipment already has hydraulic actuators built into long-
wall systems to level the machine. In addition, a slight amount of roll does
not cause major machine stress or any reduction in shearer cutting efficiency.
Second, the MSFC study pointed out that slight machine pitch does not parti-
cularly cause any operating problems. It is recognized, however, that major
vertical undulations in the pan line do put stress on the conveyor. The third
non-problem area involved identification of the coal-rock interface. If the
operator is cutting up to the rock interface, the generation of sparks provides
immediate identification of the interface. This area does become a problem if
the operator wishes to leave coal on both the top and bottom.

2. Identification of Problem Areas

The MSFC study identified some major guidance and control problems
with existing longwall systems. As stated above, under certain conditions,
machine pitch and coal-rock interface location can be serious problems. In
addition, the MSFC study suggested that machine yaw and cut following can lead
to serious control problems, and can result in excessive stress on the pan-line
and face conveyor. Not maintaining a straight line between the headgate and
tailgate was recognized as contributing to roof control problems at the
interface between the face and entries, and to interference between shields
during shield advance. Cut following problems usually occur while cutting in
thick seams where it is necessary to leave a certain amount of coal on both

3-1
 a
F '

the roof and floor. These problems also occur when the operator attempts to
cut around a parting in the coal. Improperly following each previous cut
results in inefficient cutting, shield advance problems, and excessive stress
on both the shearer and conveyor systems as the system gets out of alignment
with each successive cut.

C. RESULTS OF THE INDUSTRY SURVEY

The survey sample was structured to obtain inputs from equipment and
system designers, mine managers, equipment operators, and maintenance
personnel. The equipment and system design organizations contacted were Joy
Manufacturing, U.S. Steel Research, and Lee Engineering. The longwall mines
contacted were Carbon Fuel #34, W. Va., Utah Power 4 Light (Deer Creek), and
lT tah Power b Light (Wilberg). These equipment and design companies were
selected because they were either presently involved in longwall automation
research, or had considerable knowledge of longwall mining and recent develop-
ments in automation. The longwall mines were selected based on personal
contacts with individuals at the mine management level.

The survey questions were designed to identify major machinery design

problems. guidance and handling problems, tasks or activities that are too
demanding and therefore lead to operator or maintenance errors, and limita-
!ons of automation incorporation due to practicality (i.e., cost or system
complexity) or non-acceptance by the worker. One overall result of the survey
that lent weight to the acceptance of automation was a general consensus that
there is a greater desire (sometimes due to corporate policy) to have more
electronics and less hydraulics for system actuation. Apparently, industry
experience suggests that electrical systems have greater reliability and lend
themselves more readily to modular design and, subsequently, easier
maintenance.

The number of organizations surveyed was small because of project

funding constraints and the limited time available for project completion.
However, the answers provided by the various participants were highly
consistent. The following survey results are presented in a form similar to
the questionnaire. A summary is provided at the end of this section to allow
a comparison of the responses from the various individuals.

1. Results of the Equipment and System Designer Survey

n. Identification of Non-Problem Areas. Results of the

equipment and s y stem designer survey suggested that coal-rock interface
detection was not a problem where operators remove coal up to the rock,
because the cutter picks strike the rock and display easily visible sparks to
the operator. This is consistent with the MSFC study results. The manu-
facturers indicated that the y did not feel shield-to-shield alignment was a
problem because it is generally easy to see when the shield pads are not
parallel. Overall. the equipment manufacturers discouraged total system
automation. The reasons for this were twofold: (1) total automation would
considershiv increase s y stem complexit y , and consequently, system cost; and
(2) the increased s ystem complexit y most probably would cause additional
maintenance downtime.

'i-2
 b. Identification of Problem Areas. There were several areas
where the equipment manufacturers felt automation could assist existing
longwall operations. The first areas identified were dust and methane
control. The manufacturers were particularly aware of the problem present
longwalls have with meeting the 2 mg/m 3 dust regulation. Removal of the
operator and helper from the immediate face would allow compliance with the
intent of the dust regulation, and concurrently decrease exposure to possible
methane ignition. Other areas highlighted were equipment guidance and moving
equipment from one panel to the next. Improper guidance and lack of appro-
priate support equipment for moving components result in considerable loss of
production time. Associated with equipment guidance are horizontal and
vertical alignment. These functions are presently performed visually, and
reasonably well, by experienced operators. However, undulations in the seam
aggravate these guidance problems. The manufacturers also indicated that a
mine diagnostics and fault isolation system would greatly assist in reducing
high maintenance downtimes. Two other relatively long-range areas considered
as reasonable automation opportunities were remote roof quality sensing and
remote seam mapping. These areas were suggested as important time savers for
planning ground control schemes and mine development. A key equipment design
concern apparent in the interviews was the need to include worker inputs on
design problems so as to provide design changes that are useful and acceptable
to the worker.

2. Results of the Mine Management Survey

a. Identification of Non-Problem Areas. The mine management

interviews suggested two primary areas where operators do not experience major
problems. The first area is machine roll. Apparently, the low machine
advance rate, coupled with the availability of the hydraulically operated
leveling device, reduces this problem consideraoly. Again, this is consis-
tent with the MSFC findings. The second area is visibility. Although
visibility of both the face and machine orientation is a problem associated
with the dusty environment, it was felt that an experienced operator could
compensate through having a good feel for the machine.

b. Identification of Problem Areas. Mine management identified

several areas where automation, or remote control, could assist longwall
operations. The first apparent area is dust control. As with the equipment
manufacturers, mine management acknowledged the present problems associated
with meeting the 2 mg dust standard. Another major problem area is the time
involved in moving equipment from one panel to the next. The management
suggested the design of smaller, more modular machinery. In the area of
ground control, mine management recognized that the extremely variable nature
of this task makes it difficult to automate. However, this is a major safety
and time problem. They suggested some type of remotely operated self-
advancing roof support for the entries interfacing with the longwall face
(this suggestion also applied to room and pillar operations). Coal-rock
interface detection was recognized as a major problem for machine guidance in
thick seams. Management generally agreed that some type of fault isolation
and diagnostics system would greatly enhance operation by allowing better
scheduling of machine maintenance. Another recommendation addressed the
problem of conveyor overload and damage due to oversized lumps of coal falling

3-3

Li
off the face. Management suggested a need to control the lump size through
—me type of automatic crusher. The mine management universally 4greed that
there wa a great need for better remote seam mapping technology which woul:
aasiat ii mine planning and equipment selection. The last area addressed by
management was miner acceptance of new technology. All of the managers
cautioned against total system automation. The reasons for this were:

(1) Miners must perceive a need for a new system before they will
effectively utilize it. This may take time.

(2) Trying to integrate a new, very complex system into the mine may
exceed the present maintenance familiarity and knowledge of the
miner.

(3) Results of the Operator and Maintenance Survey

a. Identification of Non-Problem Areas. The machine operators

identified only one area as not being a major problem of the coal-rock
interface. Apparently, operator experience with visually seeing sparks and
feeling machine vibration are sufficient when removing coal up to the rock
interface. In thicker seams, where coal is left on both the top and bottom,
interface identification was recognized as a problem.

b. Identification of Problem Areas. In the area of machine

control and guidance, the overall driving factor was productivity.
Consequently, most of the problem areas addressed by the operators were
generated from impacts on tons of coal produced in one shift. "he operators
generally felt that differences in operator experience and manner of operating
the machine caused problems on a shift-to-shift basis. For example, one
operator may not maintain face alignment, causing the next operator, to have to
make a clean-up cut. The net effect of this is a loss in shift production.
Therefore, a means of standardizing machine operation seemed Important. All
operators indicated that dust generation was a major problem from the
standpoint of health, and because it obstructed visibility of the face
(causing guidance problems). The operators also stated that not maintaining
face alignment could cause considerable stress on the conveyor system,
resulting in machine downtime, and that system guidance and alignment was a
problem when poor (soft) bottom conditions were encountered. Overall, the
operators suggested that some assistance in shearer and shield alignment would
help improve productivity and reduce stress on machine components,
particularly the shearer and conveyor. Another problem area that surfaced was
the need to have better knowledge of hard partings ahead of the shearer. This
problem was cited as a large contributor to shearer damage and subsequent
downtime. The operators also indicated that some type of feedback on machine
failures (overheating, cutter pick overstress, etc.) would be useful to allow
better control of cutting rates or alignment problems before the machine
actually broke down, which would ;y elp reduce downtime. In the area of machine
design, the operators suggested greater modularity and size reduction as a
means of allowing easier machine movement and maintenance (again related to
increasing productive time).

3-4
 The inputs from the maintenance personnel revolved largely around
reducing machine failures and subsequent downtime. As with the operators,
suggestions were made to incorporate diagnostics and fault isolation, greater
modularity, and reduced size for easier component handling and maintenance.
All of the above suggestions were considered useful to the worker for improv-
ing productivity. Overall, the survey suggested that:

(1) Design improvements in the areas cited were acceptable to the

operator and maintenance people.

(2) Although workers were interested in providing their experience to

the designer, they did not feel a need to become involved in the
redesign process.

In summary, it appeared that there was reasonable agreement between all

or most of the parties concerning the following probl"-solving suggestions:

(1) Remove the workers from dust and methane hazards in the
environment.

(2) Maintain face alignment to increase productivity and reduce strews

on machine components.

(3) Install diagnostics and fault isolation to allow better scheduling

of maintenance.

(4) Provide a means to move eq+tipment easily from one panel to the
next (either through size +a.nd weight reduction, or self-contained
motorized units).

(5) Provide for coal-rock interface detection in thick coal seams.

(6) Provide for remote seam mapping to allow better mine planning and
equipment selection (such as locating rock intrusions in the coal
seam or discontinuities in the overburden which could cause poor
roof conditions).

D. SYSTEM COMPLEXITY, POTENTIAL COSTS, AND PRACTICALITY

Determining potential longwall automation opportunities was also

examined from the standpoint of potential costs and impacts on system
complexity. Both of these parameters operate hand in hand. For example,
modeling results (Section V) indicate that total automation of the shield
operation would require the development of a rather complicated 5-axis robot.
The projected cost increase per shield (using known costs for existing robotic
technology) is at least $30K. For roughly 100-120 shields (for a 600-ft face)
the longwall cost increase in 1981 dollars would be $3 to $3.6 million. This
does not conaider the additional increases associated with other automation
opportunities. This cost represents approximately a 30% increase in the cost
of present longwalls.

Another area to consider is the labor efficiency of automation from the

standpoint of classical man-machine tradeoffs (15). There are three basic
man-machine systems:

3-5
 (1) Uncompensated (man does all the information integration).

(2) Aided control (man and machino integrate information).

(3) Quickened control (total machine integration of information) (15).

Whethe ► or not a system should be totally automated is determined by a measure

of the input signal and output response of the system, resultant error, and
impact of the error on the system. For example, the preceding survey indi-
cates that total human integration of information on longwall operation is
inadequate in that the resultant error in system alignment results in produc-
tion inefficiency and possible machine system failure due to overstress.
Therefore, these data suggest that an operator aided, or totally automated,
system may be desirable. The deg ►•ee of additional automation should be
examined by first considering the advantages of total automation. Generally,
quickened control systems are employed where:

(1) The task demands are greeter than the operator can cope with.

(2) The task must he performed in an unacceptable environment for the

operator.

(1) The resultanr system error has catastrophic consequences.

(G) The allowable error is much less than the operator can insure.

(5) The operator is not required to be p-esent.

For example, automatic pilot systems on high performance jet aircraft are a
necessity to compensate for the relatively slow human response as compared to
the demand for control of high speed aircraft. The potential error in net
automating this system is catastro phic; both the pilot and the aircraft wouLi
be lost. This is not the case with guidance control on longwall systems.
First, the longwall system operate3 and advances at a relatively slow rate.
Second, an examination cf the longwall operation and activity aetwork indi-
catos that workers must he present simply because man y non-routine tasks are
required (i.r., adjusting the conveyor, additional ground control, cleanup
wider the shearer, etc.). Classical human engineering principles suggw that
if the human being must be present, then it is more efficient and cost
effective to use the human's integrative abilities rather than the machine's,
given that the potential impacts of a human error are not significant. In
longwall operation, the only impact of a human error is loss in production
time. Th^^refore, in view of the potential cost, the fact that none of the
longwall activities and potential err7rs are of a critical nature (i.e., life
rand system endangering), and wo- • kers must be present, it appears that the best
approach is to develop an aided control system. This approach is also con-
sistent with the survey results, which indicate that workers desire aids to
help them perform their tasks more efficientl y , not aids that replace the
worker completely.

F. POTENTIAL HEALTH ANT) SAFETY IMPACTS

The final aspect of determining automation opportunities concerned

health and safet y considerations. In the area of dust, longwalls presently do

1-6

i
not conform with the 2 mg standard. Automation, or remote control, of the
shearer and shields would remove both the operator and helper from the dust
plume at the face. In the area of safety, the four major accident classes in
order of severity, are: (1) handling material, (2) machinery, (3) roof/face/
rib falls, and (4) slips and falls. The bulk of the serious handling material
injuries are associated with non-routine activities such as cleanup, handling
supplies, and machine maintenance. Consequently, the more routine tasks which
could be automated (such as operating the shearer, moving the conveyor, and
advancing longwall shields), removing the worker from the hazards, would only
marginally effect the total number of s..rious injuries. The major improve-
ments in serious injury reduction through automation occur in the last three
accident classes. The major causes of serious injuries in the machinery
accident class are: (1) operators struck or caught while operating the
longwall, and (2) operators struck or caught while advancing longwall
shields. These two hazards cause 752 of the total disabling injuries. These
same two hazards, along with being struck or caught while moving the conveyor,
compose 242 of the roof/face fall disabling injuries, and 322 of the slip and
fall disabling injuries. Overall, automation of the shearer, shield, and
conveyor operations could reduce the average yearly longwall injuries by
approximately 242.

F. SUMMARY OF LONGWALL AUTOMATION OPPORTUNITIES

The final list of potential automation opportunities resulted from

integration of all of the above findings. Identification of the major problem
areas from the MSFC study, the industry survey, understanding that workers
must be present for numerous other tasks, and the health and safety evaluation,
led to the formulation of the fo'lowing list of automation candidates:

(1) Shearer automation (cutting, face alignment, coal-rock interface

detection, last cut memory, shearer arm articulation, and
tramming).

(2) Shield advance (either remote operation or semi-automated).

(3) Conveyor and pan-line advance.

(4) Computer monitoring (information management for sources of system

delays).

(5) Preventive maintenance system (fault isolation and failure

diagnostics).

(6) Face sensing ahead of the shearer (to detect hard rock partings).

(7) Remote seam mapping (as related to mine planning and development).

(8) Semi-automated movement of equipment components to a new panel.

(e.g., self-propelled towing units for shields).

3-7
 One of the project constraints was that the longwall automation study
consider only those opportunities amenable to near-term incorporation (the
1984-1985 time frame) using state-of-the-art technology. The technology
assessment of state-of-the-art automation suggested that only the first five
opportunities listed above would be available for the desired time frame of
incorporation. The last three areas shown will require much more development
effort.

3-8
 a `

SECTION IV

EVALUATIVE TOOLS FOR RANKING AUTOMATION OPPORTUNITIES

A. OVERVIEW

Sections II and III provided a longwall baseline design and operating

description, and the background for selection of the automation options. The
next step in the analysis was to establish a means of translating the desired
options for longwall automation into measurable parameters which could be used
to rank the various opportunities. Several tools were developed to accomplish
this.

One important measure of improvement is productivity. It is hoped auto-

mation could streamline the longwall operation by deleting operator delays and
hazardous tasks, thus, providing more production time at the coal face. The
tool employed to measure productivity impacts was a network which described
each activity and delay in the longwall system as time elements. Time savings
realized by automation were then added to available production time and
converted to coal tonnage. Another important aspect of production was the
cost of incorporating automation. If the costs of automating a longwall
system far outweigh production benefits, it would not be beneficial to change
the existing design.

However, before the costs of developing and incorporating automation

could be determined, the technology had to be identified. Appropriate
automation technology was identified by examining robotic technology in other
industries with similar applications. These applications were then redesigned
to meet the harsher mining environment before being translated into
development costs. The use of similar industrial applications as a tool for
identifiying and costing the required longwall automation technology was a
straightforward means of establishing cost tradeoffs.

The use of sophisticated automation technology in a harsh environment

such as a coal mine does not serve a useful purpose unless the worker realizes
a need for job assistance. Production and technology development should focus
on areas where workers perceive a need for some type of job assistance. As
discussed in Section III, the survey provided a means to isolate those key
areas where the worker would view automation as a way to improve task perfor-
mance and safety, and therefore accept it.

An important aspect of worker acceptance was whether the introduction of

automation technology into the underground mining environment would improve
health and safety. The assessment tools applied toward the health and safety
areas were methodologies previously developed and tested under the DOE
Advanced Coal Extraction Systems Project. These assessment schemes evaluate
the change in hazard exposure between the new design and a similar existing
design.

As indicated earlier, cost is an important measure of the po:'Intial

benefit of automating the longwall system. The major economic tool employed
in measuring cost benefits was the difference between productivity improvement
(converted to 1932 dollars) and the total projected capital, operating, and

4-1
 . a•

ORIGINAL PAGE 6;
OF POOR QUALITY

development costs. In the economic impact study the development costs were
deleted since they consumed onl y a small fraction of the total costs.

Each of the above analytical tools is explained in greater detail in the

following discussion. The elements of the analytical approach are outlined in
this section to provide a foundation for understanding the detailed analysis
presented in Section V.

B. NETWORK ANALYSIS

In order to identify areas where automation might aid productivity, a

network describing the key tasks in longwall operation was developed. The
first large-scale application of networks to production problems was the PERT
diagram (lb, 17). PERT diagrams are useful tools for separating a large,
complicated process into activities and delays in order to understand the
major variables necessary to complete a project or process. The activities
and delays are structured in a sequence similar to Figure 4-1.

Figure 4-1 illustrates how various events (i.e., activities and delays)
are sequenced in the network. Events that occur in a straight line (events 1
and 2 above) are in series with each other because event number 1 must occur
before event 2 can commence and end (17). However, event 3 can start and end
while events 1 and 2 are proceeding. Therefore, event 3 is placed in parallel
with the first two events (17). In a similar manner, the complete process is
diagrammed from beginning to end, with the legs in the network representing
the various times for task start, duration, and end.

START END
EVENT EVENT
3 , DURATION OF EVENT 3 3 d

START ND START END

EVENTENT EVENT
1 ^DURATION 2 . DURATION OF EVENT
EVENT 22

Figure 4-1. Example Network

4-2
 The complete application of the PERT method also requires that.best,
worst, and desired event completion times be used. Multiplying the proba-
bility of an event being completed (based on historical experience) by the
three completion time estimates results in the expected value for finishing a
given task (16, 17). When all the expected values (the beat estimates of task
completion) have been placed in the network, the minimum time to complete the
whole process can be determined. This is called the critical path and is an
important baseline to estab.ish in order to examine ways of streamlining the
process (17).

Probabilistic simulation calculations are extremely important for

networks that have several parallel activities. These calculations help
prevent evrors in determining the critical path. However, networks that are
basically a series structure greatl y simplify the process of identifying the
critical path. In these cases a deterministic approach is acceptable. Given
that the process is basically a series of events, the mean values of the
activities and delays become a good representation of the critical path. The
longwall process examined in this study, being basically a series type
operation, was therefore amenable to using a deterministic type network.

As stated earlier, it is very useful to determine the critical path

before a process can be examined. The critical path for the longwall process
represented the main tool for evaluating the productivity impacts of the
automation opportunities. This was accomplished by first determining which
delays and activities on the critical path were affected by automation. Each
automation option was converted into a time savings, where applicable. These
savings were then added to the available face production time. The estimates
of the production time improvements were later reviewed by a group of experts
from the mining industry to ensure that these projections were reasonable.
The automation options were ultimately ranked by examining which one, or
combination, gave the beet improvement in productivity. The results of the
longwall critical path ana;vsis, productivity projections, and expert review
are explained in greater detail in Section V.

C. Al ►TOMATION FEASIBILITY AND DEVELOPMENT COST INDICATORS

The evaluation tool for determining the feasibility of applying

automation to underground mining was based on finding the required tech-
nologies in other industries. The evaluation of the automation options
included an assessment of: (1) the required sequence and time for technology
development, and (2) the respective development costs for similar equipment of
comparable complexity.

Generic automation technologies already exist in varying degrees in

industr y . A brief description of these technologies was provided in Section
I.0 and is continued in the following discussion. The first area, digital
electronics, is one major component of automated systems. Digital electronic
computers appeared first in the scientific world, followed by the business
world. It took the invention of the large-scale integrated circuit before
digital electronic computers invaded the timekeeping world (i.e., digital
watches). Roughly 10 y r after the microprocessor was introduced to industrial
applications, a longwall shearer was developed with a microprocessor installed
(18). Since then, the British Coal Board has developed a mine monitoring and

G-3
control system called MINOS (Mine Operating System) with over 10 units
delivered to mines. The fact that U.S. mines have started to introduce
similar systems underground is a solid indication that digital electronics are
evolving toward a practical application in the mining industry.

Servomechanisms and actuators are also key components of automated

systems. An input signal (such as a digital electronic signal) causes an
action (via an actuator). The observation (feedback sensor output) of the
action is compared to the input. If the desired action is not achieved,
additional action (done through a comparator) is taken until the feedback
matches the input command. The comparator subtracts feedback sensor output
from the input. When the result is zero, the system is in the desired state
and the process stops. An automobile speed control system is a simple example
of a modern electro-mechanical servo system. A typical servomechanism is
shown in Figure 4-2.

Over the years the summing network, or zero seeking comparator, has
evolved from hydraulic or mechanical components through analog electrical
components to digital electronic computer components. With evolution has come
a reduction in cost and size, and an increase in performance and speed of
operation. Taken one step further, several digital electronic servos,
incorporated in an assembly which processes several incoming signals for the
purpose of performing a distinct set of functions, form a fully automated
device. The servos which actuate the system may be electrical, pneumatic, or
hydraulic units.

The above discussion attempts to show that the basic automation

technology components have existed for some time. The key to automating
longwall systems lies in transferring the proper combinations of present
digital electronics, servos, and comparators to the appropriate shearer,
shield, and conveyor machine functions. This transfer is provided with
supporting designs, later in the document.

Once each automation area has been conceptualized using applicable

state-of-the-art technology for the selected machine functions (based on the
previously developed automation areas), the development costs can be reliably
estimated. However, since some applications may require extensions of
existing technology such as a more rugged circuit design suitable for the
mining environment, coal industry experience was incorporated to provide a
more realistic indication of development time and cost.

INPUT wi COMPARATOR 1 Wi ACTUATOR

FEEDBACK SENSOR

Figure 4-2 Servomechanism Data Flow Diagram

4-4
D. MINER ACCEPTANCE

As stated in Section III, the survey was designed to sample opinions

from equipment designers and manufacturers, mine managers, and miners. The
key opinions concerning worker acceptance of new technology were obtained from
the interviews with mine management, equipment operators, and maintenance per-
sonnel. The survey, if structured properly, is a useful tool for determining
user needs and preferences. The basic approach taken in developing the survey
was to make site visits and personally question the managers and workers.
Each interview was structured to identify major machinery design problems,
equipment guidance problems, and tasks or activities that were excessively
demanding. Major machinery design problems were important to isolate because
of their relation to potential operational errors and maintenance problems.
Understanding these problems lent credibility to the need for an operator-
assist system or failure diagnostics. Equipment and guidance problems and
excessively demanding tasks were important for identifying other operator or
maintenance-assist opportunities. The interview was not intended to alienate
workers from the introduction of automation into the mining environment. It
was understood at the onset of the survey that miners are very sensitive to
any technology that might infringe on their jobs (19). Therefore, the
approach taken was to develop automation as an additional tool rather than as
a job threat. One final aspect of the survey was to determine the degree to
which miners should be involved in th^ design process. This last aspect was
included so that the design would jointly benefit from the worker's experience
and provide the worker a means of participating in the design process.
Through the survey approach outlined above the important automation acceptance
criteria were identified.

E. HEALTH AND SAFETY INDICATORS

The evaluation tools for assessing the health and safety impacts of
automation were developed separately under the JPL Advanced Coal Extraction
Systems Project (20, 21). Potential health improvements are evaluated using a
qualitative methodology. This approach was selected because any stage of
development prior to the prototype test does not provide information on actual
dust concentrations or toxic material emissions. However, by understanding
the comparative levels of exposure to health hazards of a proposed design and
a similar conventional system (considering worker activities which interface
with hazards and design improvements such as additional protection), the
^iethodology provides a relative means of rating the effectiveness of various
health design features (20).

The safety methodology is both qualitative and quantitative. The

qualitative portion ,erves the same purpose as the above health hazard
evaluation scheme. However, using the historical injury data base for the
comparable conventional system, the design analysis can be carried one step
further into a quantitative stage. The quantitative injury projection takes
the changes in hazard exposure time during task performance, changes in the
number of workers exposed, and improvements in protection and adjusts the
historical injuries accordingly (21). Once the injury projection is completed,
the relative contribution of each new design feature to the overall reduction
in injuries provides the means of raking the automation opportunities. This
procedure is developed in greater detail in the next section.

4-5
 L ^r

F. COST-BENEFIT INDICATORS

A number of factors contribute to the costs and potential benefits of

the automation opportunities. These factors are summarized in terms of the f
present value of net national benefits of the new automation technologies.
The present value of the benefits was determined by discounting at a 72 (real)
rate and summing the difference between benefits and costs (to American mine
operators) over the time period 1985 to 2000 for each option. This approach
also provided the means of ranking the various automation options. The
elements used to evaluate cost-benefits of automation were as follows:

(1) Influence on productivity.

(2) Changes in capital and operating costs.

(3) Size of potential market (1985-2000).

(4) Time of availability and rate of penetration into the potential

market.

(5) Health and safety impacts and costs.

The time of availability and penetration of automation into the potential

longwall market was based on several factors. These factors included the
following:

(1) Time required for RED and prototype development.

(2) MSHA approval time.

(3) Acceptability to mine personnel and mine owners (e.g., automation

must produce a high rate of return on equity in excess of 25%
after taxes, and will not increase health and safety hazards in
the mine).

(4) Historical penetration rates of longwall technology in the U.S.

Other cost factors which were not addressed in this study were: (1)
national security, as it is affected by domestic energy availability, (2)
international trade, as it might be affected by increased export of coal and
coal mining equipment, and (3) possible spinoff benefits from the proposed RED.

This section introduced the various methods of design and assessment in

the areas of productivity, technology development, miner acceptance, health
and safety, and cost benefits. Section V will show how these analytical tools
were applied specifically to integrating and evaluating the automation options
as part of the baseline longwall design.

4-6
 5

SECTION V

RESULTS OF THE LONGWALL AUTOMATION STUDY

A. OVERVIEW

The previous section briefly characterized the various tools used to

evaluate the impacts of automation on the longwall system. This section
presents the results of applying these tools to various automation options.
In the productivity area the deterministic network was employed, along with
the automation options developed in Section III, to determine potential
improvements in productivity. These improvements were later converted to
present-value dollars for the cost-benefit analysis. The necessary hardware
and software for each automation option were identified through the technology
assessment analysis. As stated earlier, state-of-the-art digital electronics
and servomechanisms were used where possible. The development and production
costs followed once the off-the-shelf components had been identified. These
cost data were later employed to establish a baseline for understanding the
relative payback on the research and development investment. The technology
assessment also included evaluation of health and safety impacts. Reductions
in serious longwall injuries were translated into cost savings for the cost-
benefit analysis. Last, the cost-benefit evaluation took the above data
inputs (with the exception of the development costs; these costs were
considered separately) and provided the net national benefits. Where
possible, conservatism was exercised in the calculations to prevent overly
optimistic projections. For example, market penetration rates of automated
longwall equipment were assumed to be no greater than historical longwall
growth, despite an extremely favorable payback time.

B. RESULTS OF NETWORK ANALYSIS

The longwall network analysis was divided into four basic steps. The
first step was the development of a longwall and operational network. The
automation options and respective estimates of time savings were inserted into
the network where it was felt a given delay could be affected. As these delay
reductions only represented estimates, the next step involved meeting with
experts in the mining industry and revising the projections based on actual
experience. In some cases the estimates proved to be optimistic, while in
other cases the estimates were rather conservative. Based on the input from
the experts, the final projected time savings (and productivity impacts) were
determined. Each of the above steps is explained in the following discussion.

1. Longwall Operational Network

The longwall operational network was based on the general

operating scenario provided in Section II, a set of operating projections
prepared earlier under the DOE Advanced Coal Extraction Systems Project, and
on-site observations. As stated above, the series nature of the longwall
process allowed a deterministic approach to be used in designing the network.
Additionally, control of dust and face ignitions (small explosions of methane
gaa at the face) constrained the cutting cycle to a single pass from the

S-1
headgate to tailgate. It is understood that ideal environmental conditions
(very little dust and no methane) in conjunction with a seen height below
48 in. does permit some longwalls to cut in both directions. However, this I
not normal practice. Using 480 min as the baseline shift definition, Figure
illustrates the basic longwall operation. The key for the definition of each
numbered activity and delay element shown in the network is provided in Table
5-1. It should be noted that the first seven activities listed in Table 5-1 do
not have mean times listed because these activities are not part of the normal
longwall operating cycle and are not affected by the automation options.

As illustrated in the network, event combinations 5 through 16 basically

represent the shearer cutting cycle. The conveyor and shield movement, events
13-16 and 11-14 respectively, operate in parallel with the shearer. Similarly,
the haulage operation (events 5-30), power and water delivery (event 5-25), and
new panel development (event 35-36) also occur in parallel with shearer opera-
tion. The "critical path" for the system is denoted by the dark arrows in the
network diagram. Delays associated with some of the parallel events can cause
total system shutdown (e.g., power outage). When these delays occur they
become part of the critical path and subtract from the available production
time. These delays, where they could have major impact on production, were
considered when examining the various automation options .snd their impacts on
shortening the critical path.

After development of the longwall network, the next step was to integrate
the various automation options into the network at the appropriate delay
points. As indicated earlier, the primary automation candidates were:

(1) Shearer automation.

(2) Shield advance.

(3) Conveyor and pan-line advance.

(4) Computer monitoring (management information system).

(5) Fault isolation and failure diagnostics.

Qualitatively, it would appear that automation or remote control of the shearer

would prevent the shearer from stalling due to heavy cuts, and allow continuous
operation during periods of normal operator delay. In addition, operator
errors in face alignment would also be reduced. Remotely controlling or
automating the shield and conveyor advance, in conjunction with maintaining
face alignment, would remove some of the delays associated with equipment
breakdowns caused b y poor alignment (e.g., chain broken/out of guide, pan
hangups, flight jamming, and shield interference). Examination of the longwall
operational network suggests that delays associated with shearer maintenance,
face convevor problems, and outby haulage are sufficiently large that a fault
isolation and diagnostics system would allow better maintenance scheduling on
the shields, pumps, section power/water, and electrical systems during these
periods. A simple model was developed to merge the above qualitative
assessment into the activities and delays on the critical path. A preliminary
estimate of the increase in production time was calculated by subtracting the
delay times associated with each of the above event descriptions from the
critical path. The following equations and flow diagram (Figure 5-2) explain
how the production increases were calculated.

5-2
 .,

ORIGINAL PAGE
OP POOR Q 11' S

xw
0
L
3
cu
z
^o
a
O
M
U
C
a
w

d
u
y

tO

C
O
..a

.r
u't

47
7
OD
4.

5-3
 Table 5-1. Longwell System Functional Network Key

Event Description Mean Time

Activity Description Per Shift (Min)

1-2 Drive initial entries

2-? Establish initial crosscuts

r
3-4 Haul equipment

3-35 Drive other entries

35-36 Develop new panels

4-5 Install support systems

Entry conveyor, stage loader,
hydraulic pumps, water pumps,
power center, cables, etc.

4-5 Setup and checkout longwall

Shields, face conveyor, shearer,
position mechanical/hydraulic
roof supports and lines, etc.

5-6 Sump to 100 ft from headgate 8.4

6-7 Amortized headgate move 3.6

Conveyor and shields-advanced

7-8 Face cut - 500 ft. @ 11 fpm 89

8-9 Amortized trim cut 22.5a

8.5b

9-10 Cut termination and turn around 4.2

Stop shearer travel, range drums,
reverse cowls, reverse shearer
direction

10-15 Flit travel back toward headgate 35

(500 ft @ 28 fpm), range drums

15-16 Return cut to headgate (100 ft @ It f;Ym) 18

a New section with inexperienced operators

bExperienced section

5-4
 Table 5-1. (Cont'd.)

Event Description Mean Time

Activity Description Per Shift (Min)

6-11 Shield delay 9.5

Hydraulic lines, weld repair, stuck
shields

11-12 Conveyor delay 31

Chain broken, out of guide,
pan damaged,
flite damaged, jammed

12-13 Conveyor jammed, adjusted 2.0

13-16 Conveyor advance 1

11-14 Shield advance Same as shearer

14-10 Shield advance lag 4.5

(after shearer movement)

16-17 Stop shearer, check and/or replace picks 35

17-18 Prepare for next cut to tailgate 3.3

Start and range drums, reverse cowls

18-19 Provide added ground control 18/36a

19-20 Shearer maintenance 26/66b

Gearbox, drive shaft, etc.

20-21 tleadgate overload 6.0

21-22 Cleanup under sheerer 22

22-23 Operational problems (oversized coal) 6.0

23-24 Stage loader/crusher hangup 5.0

Spillage, oversized lumps, etc.

24-25 Head/tailgate drive problems 11

5-25 Section power, water, face Same as shearer

Ventilation operation

a 18 min for good roof; 36 ruin. for bad roof

b 26 min for high coal; 66 min. for low coal

5-5
 a

OF 'MR Q `.ALI 'Y

Table 5-1. (Cont'd.)

Event Description Mean Time

Activity Description Per Shift (Min)

25-26 Section power and water shutdown 7.0

26-27 Electrical maintenance (overloads) 3.0

27-28 Pump failures 3.0

28-29 Rearrange supports at headgate and 3.0

tailgate

5-30 Outby haulage delays 47

Conveyor, railcar spotting, etc.

30-31 Outby haulage operation Same as shearer

31-29 Electrical maintenance 8.0

Power failures, cable defects,
warning/communication devices

32-33 Amortized retreat of support

System (for 10n £t of cut; 40 cuts)

Where,

ti - The time delay element affected by the automation

opportunity.

DT - Total system delay time, where D T ranges between 279-319

min/shift.
PT - Total average production time, where PT for one-way is 199
min/shift.
The initial projections for time savings (considering only one way
cutting) are shown in Table 5-2.

This production time increase, when placed back into the cut cycle
times, resulted in the shearer being able to perform one more 490-ft cut.
This represents a 50% increase ; a production time.

2. Discussion of Projections with Industry

The above model repeesented a first attempt at a projection of the

aggregate impact of automation opportunities on production time. It was also
assumed that several large delays would not be removed from the critical path
as a meanv of being conservative for the initial projection. The next step
was to involve a group of equipment experts using both the present operating
delays and projected improvements as an initial starting range through which,
given the rationale for the projections, they could extrapolate a more

5-6


OMMM • ^GE 13
CW MM QUALITY
(hart)

It i for all system component delays

Input revision@ to
projectten tree experts

Calculate

st i Calculate revised production

per shift

Calculate lecalculate f tuts

PT +=t i as [ R= +2ti)

final projection of automation

Impacts an operatin& time and
productivity

(stop)

Figure 5-2. Flow Chart to Project Automation Impacts

on Operating Time and Productivity

realistic answer based on their experience. This procedure allowed the

initial projections to he refined. The basic process was as follows:

(1) Experts were familiarized with the automation effort, rationale

for the opportunities, and list of automation opportunities.

(2) Technical backup and feasibility of incorporating opportunities

were presented.

(3) Overall results of -network analysis were presented; essentially,

the production time increases, when placed back into the cut cycle
times, resulted in the shearer being able to perform one more
490-ft cut travel in the case of a one-way cut. This represents a
50% increase in production time.

(4) Areas where automation opportunities apply were provided along

with the rationale.

(5) Experts were provided data on the present operating time scenarios
and the projected improvements based on automation, for:

(a) Shearer operation.

(b) Shield/conveyor advance.
(c) Fault isolation/diagnostics.

1
5-7
 ORIGINAL PAGE 15
OF POOR QUALITY

Table 5-2. Initial Projection of Production Tire Savings

Automation Component/Activity Time Savings (min /shift)

Shearer -
Readgate overload 6
Operator delays 2-10 (avg 6)
Lunch 60
Subtotal 72

Shield/conveyor advance
Conveyor jammed/adjusted 2
Chain broken/out of guide 6
Pan hangups (nominal) 10
Conveyor flight jamming 2
Shields stuck/interference 5
Subtotal 25

Fault isolation/diagnostics
Electrical maintenance 11
Pump failures 3
Section power and water
shutdown 7
Subtotal 21

Total 118 min/shift

The experts were provided both the present and projected time impacts as
a baseline from which they could make their own projections. The review
process commenced with asking each expert if he believed the actual effects of
automation would be closer to the projected or present operating scenario
based on his own experience and the design analysis. The range between the
two operating times was then cut in half, and the ^xperts were asked to refine
their estimates. This process cont4nued until the experts could not refine
answers any further. The final answer is typically provided as a range and
the reasons for each answer were also recorded in order to establish a sound
foundation for the revised projection.

5-8
 The experts chosen were from Foster-Miller (FM), Joy Manufacturing (J),
and Skelly and Loy (SL). All of the participants had a good working knowledge
of existing automation practices and longwall machinery operations. It is
important to note that although the experts varied in their perceptions of the
magnitude of the automatio.. impacts on production time, the types of responses,
in terms of losses or improvements, were consistent. The final production
time improvement estimates were displayed as a range, encompassing all of the
answers. Each group of experts was interviewed separately and not informed of
the results of the other interviews. This approach ensured that each partici-
pant was not biased by the other responses. Table 5-3 summarises the results
of the interviews.

3. Impact of Redc-.ed Operating Delays on Available Production Time

The potential time savings resulting from incorporation of

automation opportuniti*:y can he determined by subtracting the final projected
delays from the averng• delays presently experienced by longwall systems. The
only deviation fro g this is consideration of the trim cut delay. In the
original network L ­ cut sequence did not include a trim cut because no
experience was ave lrble. As a result, the calculated 199 min of cut time par
shift (for one way cut) represented the total production time available
considering th 4 other };mown delays. incorporating the trim cut delays
essentiall- reduced the available production time down to a range of roughly
170-190 min/shift, which is well within the range of presei.t longwall
operations. By removing the amortized trim cut delay through automated face
alignment, the production time is returned to the expected operating scenario
shown in the network. Any additional time savings due to automation are then
simply added on to the available production time.

a. Shearer improvements. Table 5-3 shows that although the

original projection for the preliminary scenario conservatively indicated no
improvement in shearer maintenance time, the experts disagreed. The two main
reasons given for a reduced delay time were: (1) automating the shearer would
control machine stress better, thus reducing maintenance (e.g., cutter pick
wear); and (2) installing fault isolation sensors on major components would
save time spent isolating an unknown failure and gaining access to the
component. The experts all agreed that controlling the load on the shearer
would greatly reduce the headgate stall problem. It was interesting to note
that all the experts similarly felt that automating the face alignment task
would alleviate the problem of having to take trim cuts.

In the area of operator delays, the original network suggested an

operator delay of 2-10 min/shift and a lunch delay of 60 min/shift. This
would result in a total operator delay of 62-70 min/shift. All the experts
indicated that delineation of the total delay by "operator" and "lunch" was
not accurate in that workers' lunch breaks ere staggered such that machine
operation is continuous. The experts did comment that staggered lunch breaks
usuall y result in a reduced face crew. The net impact of this reduction is a
slower operating speed so that the available face crew can perform the tasks
of the absent workers. Considering the loss in production time due to the
slower operating speed and other operator delays, the experts adjusted the
original baseline from 62-70 min/shift down to 20-40 min/shift.

5-9
 4

Table 5-3. Expert Adjustment of Projected Automation

Impacts on Production Delays

Automation Opportunity Initial Adjustment by

and Affected Activity Projected Experts Reason
Delay (min/shift) (min/shift)

1. Shearer Automation High coal-26 20-25 FM/SL - Controlling the

Low coal-66 50-63 shearer load through
Shearer maintenance automation should
reduce maintenance
slightly.
J - Fault isolating
major components
should save access time

Cleanup under the 22 Same FM/J/SL - this activity

shearer will not be affected
by automation

Headgate overload 0 0-1.5 FM/J/SL - generally

resulting in shearer agree that controlling
stall the load on the
shearer will prevent
overload; some delay
might still be
experienced due to
roof conditions

Need for trim cuts to

realign face (one-way
cut):

• Inexperienced No historical a0ne trim FM/J/SL - This is an

operator, or data available cut every average trim cut re-
opening a new for projection 1.5 shifts, quirement based on
longwall section or a delay of experience; auto face
15-30 min alignment would solve
problem

• Experienced No historical One trim cut FM/J/SL - alignment is

operator data available every 4 limited by face visi-
for projection shifts, or a bility; auto face
delay of 6-11 alignment would solve
min problem

aExperience indicates that for every trim cut, approximately 0.5-1 cut per
shift is lost in the process of squaring up the face, adjusting ground
control, and positioning the head or tailgate sections.

5-10
i y

Table 5-3. (Cont'd.)

Automation Opportunity Initial Adjustment by

and Affected Activity Projected Experts Reason
Delay (min/shift) (min/shift)

Operator delays 0 0 FM/J/SL - Concur that

total shearer
automation would
remove operator delays
(remote control will
still require at least
one operator to be
present)

2. Shield/conve yor advance

Conveyor chain broken 0 0-1 FM/SL - There still may

or out of guide be a slight delay due
to chain stretch.
J - Agre^- with
projection; poor
alignment is the major
contributor to
conveyor chain problems

Conveyor pan failure/ 10 10 FM/SL/J - Agree with

hangups projection; poor
alignment puts stress
on the conveyor pan
connecting pins; it is
understood that other
kinds of pan failures,
such as gob falling on
the conveyor, will not
be affected by
automation

Conveyor flight 3 3 FM/J/SL - Agree with

damaged or jammed projection; this
failure is often
related to stress
induced by not
coordinating headgate,
shield, and panline
moves so that stress
on the conveyor flight
is minimized - other
failures such as gob
damage will not be
affected

5-11
 Table 5-3. (Cont'd.)

Automation Opportunity Initial Adjustment by

and Affected Activity Projected Experts Reason
Delay (min/shift) (min/shift)

Conveyor is jammed or 0 1 FM/J/SL - Although

requires adjustment this delay is
sometimes caused by
alignment, other
factors such as bad
floor conditions may
prevent this delay
from being eliminated

Shields stuck or require 5 5 FM /J/SL - Agree with a

maintenance medium improvement in
delay time since
shield interference
and guidance would be
aided by automation;
bad floor conditions
and maintenance would
not be helped by
automation

Head/tail drive fail- 11 8.5-11 FM /J/- Improving

ures requiring main- alignment will reduce
tenance (such as gear some of the stress
drives, fluid experienced on compo-
couplings, seals) nents (such as fluid
couplings); fault
isolation will help
identify major
component failures and
save some access time
SL - Automation will
have no effect

Rearrange support at 3 Same FM /J /SL

- Agree that
the head/tail automation will not
affect this delay

Cleanup and breaking 11 Same /SL - Agree that

FM /J
oversized lumps automation will have
no effect on this delay

5-12
 Table 5-3 (Cont'd.)

_ Automation Opportunity Initial Adjustment by

and Affected Activity Projected Experts Reason
Delay (sin/shift) (min/shift)

3. Fault isolation/
diagnostics

Electrical maintenance
(cable failures, power
overload) 0 15-19 J/SL - Electrical fail-
pump failures urea are difficult to
section power/water diagnose in sufficient
time to schedule
maintenance (accept
for cable failures);
also, manpower would
most likely be at face
during face breakdowns,
so impact would be
marginal. FM-Only a
marginal impact would
be felt since the
maintenance manpower
would most likely be
concentrated at the
face

Outby haulage delays Less than 47 24-42 FM/J/SL - The conveyor

(conveyor breakdown, system is too large to
rail car delays) monitor, but some
improvements would be
realized through
better rail car
scheduling, loading,
and moving materials
in and out of the mine
more efficiently

5-13
 All three groups of experts agreed that by completely automating the
shearer and shield activities, operator delays would largely be removed.
However, time savings could also be realized if remote control were employed,
since one operator could still operate the shearer and have time to monitor
and control other events (e.g., one operator could activate the shearer
sequence and remotely control the shields).

Assuming that the full cut time of 199 min is available (by removing the
amortized trim cut delays), the total range of time savings related to other
shearer delays is as follows:

(1) Low coal: 27.5-62 min/shift, 44.8 min/shift average.

(2) High coal: 25.5-52 min/shift, 38.8 min/shift average.

b. Shield - Conveyor Improvements. The experts generally

agreed that most of the conveyor stress and shield interference problems were
associated with face alignment. Although the experts felt that normal
conveyor chain slack due to chain link stretch would still require adjustment
and therefore cause some delay, other major problems such as the chain coming
out of the guide would be largely mitigated. Similarly, although pan/conveyor
jamming and damage due to poor roof or floor conditions or falling gob would
still exist, the experts indicated that automated face alignment would largely
reduce stress on the panline connecting pins, reduce panline jamming, and
reduce stress on the head and taildrive components. In the area of shield
interference, the experts agreed that automated movement and alignment would
assist greatly in maintaining shield spacing and a level advance. The total
shield delay could not be removed because of uncontrollable variables such as
bad roof and floor conditions.

The projected time savings for all of the above areas is 22-25.5
min/shift.

C. Improvements in Component Failure Identification and

Logistics Support. The experts provided useful comments on the application of
sensor technology to system monitoring and failure diagnostics. All the
experts indicated that sensor applications are limited primarily by cost. For
example, it is possible to monitor a total conveyor system in terms of motor
performance, bearing wear, belt tension, belt stress, roller wear, etc.;
however, the overall size of the conveyor system requires a large number of
sensors and a complicated data retrieval and processing system. This system
would be costly and would introduce a greater number of failure possibilities.
At the onset, the experts made it clear that only major system components
(such as the main drive shaft bearing, circuit breakers, and pressure line
junctions) and key associated failure modes can realistically be identified.
Examples of key failure modes might be bearing wear, midpoint conveyor belt
tension, motor temperature, change in resistance in a cable segment, critical
pressure loss in a group of shields, or main water line blockage.

The two major delays in the network that could potentially be reduced by
fault isolation or a management information system were: (1) electrical
maintenance, pump/water, and power delays, and (2) outby haulage delays such

5-14
 T INN

as conveyor breakdown or rail car problems. In line with the comments

outlined above, the experts generally felt that a diagnostics system, although
useful, would only marginally reduce outby maintenance delays. The reasons
for this were threefold. First, except for cable failures, electrical
failures are difficult to fault isolate or monitor before they actually
occur. Even though one could save some troubleshooting time by knowing where
the failure occurred, the delay would not be removed from the critical path
due to the extremely rapid progression of the failure. Second, a system such
a ^uthy conveyor is extremely difficult to monitor. Third, even if
­sere informed of impending failures that could be repaired during
App ays, the available maintenance crew would probably be at the face
> r shearer. One definite advantage of a monitoring system, as
P. the experts, would be in moving rail cars to and from the face
ar nagement would be able to schedule rail cars, control rail car
loaaitia , potting, and schedule delivery of support material much more
efficiently. Other areas pointed out as being favorably affected by a failure
diagnostics system are: ( 1) reduction in troubleshooting and access time for
major shearer components, and (2) reduction in troubleshooting time for major
headgate and tailgate components.

In total, the final projected time savings associated with utilizing

fault isolation and management information systems was 7-29 min / shift.

4. Summary of Projected Automation Impacts on Operating Time and

Productivity

In summary, the total projected reduction in delays is simply the

sum of all the above time savings. Assuming one way cutting as the baseline
system, this total is respectively 56.5-116.5 min/shift for low coal, and
54.5-106.5 min/shift for high coal. The largest segment of the time savings
associated with both low and high coal operations stems from removing operator
delays. The next largest impact on production time is derived from
alleviating the need for trim cuts.

Thus, for a one way cut,

ti s 56.5-116 . 5 (86.5 avg) min/shift for low coal

ti - 54.5-106 . 5 (80.5 avg) min / shift for high coal

Calculating P t + ti results in

255.5-315.5 min/shift for low coal, and

253.5-305.5 min/shift for high coal.

These available production times (when placed back in the cut cycle
times) result in the shearer being able to perform an extra three-fourths of a
490-ft cut. This represents an approximate 40% increase in production time.

5-15
C. AUTOMA;.JN TECHNOLOG`a ASSESSMENT

As illustrated in Sections III and IV, the operator and support

personnel are necessary components of the longwall system. For example, the
human operator of the longwall shearer functions as a comparator in a servo
system. His vision and other senses provide the control signal, and his
actuation of the control allows the ranging arm to operate. In analyzing the
results of machine operation, the operator receives inputs on shearer pitch,
roll, and heave along with geological inputs on coal seas dip. These inputs,
plus other variables, such as roof quality, require constant attention and
place a heavy load on the operator. Easing of this load should allow
operators to handle the remaining tasks more safely and efficiently.

Development of any new application includes a breadboard phase, a

prototype phase, a preproduction phase, and a production phase. The
breadboard phase is essential for quickly designing, assembling, and testing
digital electronic systems required for automation. The breadboard phase also
facilitates the development of a reliable prototype design. The use of
digital computers in conjunction with multiple servo systems (manually
assisted robotics) is sufficiently mature to develop sensor mountings with
data feedback on a production shearer during the breadboard stage. The
sensors and controls that are incorporated into a production machine must n ,3t
degrade machine performance, and must be able to be actuated or cut out by a
single switch. The choice of the digital electronic computer approach allows
an easy way of servo performance modification by software changes during the
breadboard stage. The software development is not as mature as hardware
development. Thus, the system software will continue to be refined during
in-mine testing of the breadboard designs and prototype. As each subsystem is
tested, it can be combined with other subsystems and retested. It is for this
reason that a breadboard phase is extremely useful. In this manner, a totally
integrated system can be developed. An approach to accomplish this goal is
presented in the following discussion.

1. Identification and Detailed Description of Opportunities

One type of automation technology for the longwall will be

electronic digital computers. The opportunity to apply this technology can be
subdivided into several hardware/software programs, and an experimental
configuration can be designed. This experimental set-up consists of sensors
located on the shearer, the headgate, the shields, and elsewhere as needed.
Each sensor sends its data to a general purpose computer away from the coal
face. The computer makes the calculations, prints summary charts with a
plotter, and sends commands back to actuators on the mining equipmerc. The
mining equipment will have a manual/auto switch so that the operator can
select modes. During the course of the test period, different forms of
automation can be tried and compared. The general purpose computer will make
the calculations necessary for the selected type of feedback and transmit
these signals to the mining equipment. In addition, the computer can format
the data for management information plots. The general purpose computer will
initially be used as a development tool for each of the concepts presented,
and represents a fixed t.ost in the development of any of the following
concepts. It is envisioned that computer monitoring/control and auto fault
isolation data handling and display will always be done at an above ground
location. Thus, a general purpose computer installed for software development

5-16
E

will also find use in general mining operations. Tn addition to the above
computer (which will process incoming sensor signals) and associated
communications, sensor and actuator development packages are required. The
development packages are described below; block diagrams are employed to
illustrate the complete conceptual design for each automated component.

a. Smart Shearer Ranginj Arm Articulation Package. The

fcllowing sensors are required for ranging arm articulation on the shearer:

(1) "Distance-along-the-face" sensor.

(2) Ranging arm to shearer body angle sensor.

(3) Shearer tilt angle sensor.

(4) aaearer roll angle sensor.

(5) Left armored face conveyor section tilt angle sensor.

(6) Right armored face conveyor section tilt angle sensor.

Figure 5-3 illustrates how the various sensors would be used to operate the
shearer rangin- arm.

b. Smart Shearer Cowl Articulation Package. in addition to

"distance-along-the-face" sensors shown above, the following sensors are also
required for cowl articulation.

(l) Left cowl to shearer body angle sensor.

(2) Right cowl to shearer body angle sensor.

Figure 5-4 below illustrates how the additional sensor feedbacks would be used
to operate t-± shearer cowls.

C. Smart Shearer Face Alignment Package. "Along-the-face"

sensors similar to the ones shown in Figure 5-3 may also be employed to
maintain face alignment. In addition to making face tilt and roll measure-
ments, the desired movement must be computed and then communicated to the
individual shields. Figure 5-5 demonstrates this iterative process.

d. Face Conveyor Alignment Computation Package. Using the

"distancealong-the-face" sensors and the face alignment sensors described
above, a software package can provide the communication with the smart shearer
via the computer, and calculate the desired movement of the ram on each shield
in order to adjust for face conveyor misalignment. This information would be
trans- witted to separate microprocessors located on each shield. The shield
micro- processors would then signal the respective actuators (i.e., the ram
servo mechanisms) to control each shield. Figure 5-6 illustrates this control
loop.

5-17
 . a

ORKii M PAN 0
OF POOR QUALITY

LEFT CONVEYOR

RIGHT CONVEYOR

OF COAL SEAM
DI

*The left and right conveyor sections are those i.mme r ii.ately left and right of
the shearer cowl.

Figure 5-3. Shearer Ranging Arm Articulation Flow Diagram

5-18
 OR GUM PAGE 9
OF POOR QUALM

MANUAL INPUT

COMPUTER MEMORY MUNCR

FOR EACH RANOINO ARM

FEEDBACK SENSOR FOR EACH COMPARATOR If EACH

RANGING ARM
AIM COWL I

ACTUATOR FOR EACH

RANGING ARM

Figure 5-4. Shearer Cowl Articulation Package Flow Diagram

B3h'TON TYPE
FACE ALIGNMENT SENSOR
COMMUNICATIONS
DISTANCE ALONG FACE

COMPUTATION OF DESIRED COMMUNICATIONS WITH

FEEDBACK SENSOR ON EACH

Figure 5-5. Shearer Face Alignment Flow Diagram

5-19
 ,a •

pRltiiN/►L ^A
OF FAR

HEAD END COMPUTER

MICROPROCESSOR ON EACH 3HIELD COMMUNICATES BOTH WAY3 WITH

COMPARATOR, PART OF

Figure 5-6. Face Conveyor Alignment Flow Diagram

It is expected that the task of calculating the shield Ldvance sequence

will also be done as part of the control loop described above. The shield
advance command is communicated to the microprocessors controlling pan-line
advance. Each shield will be modified to include a movement warning light, an
obstacle/emergency stop control, an unacceptable tilt sensor, and a roof
pressure sensor. The hydraulic controls must be modified to accept electric
control. The microprocessor controlled shield advance program will turn on
the warning light, and, while moving, continuously check that there are no
obstacles and that the tilt is within acceptable limits. The microprocessor-
controlled shield will go through the following movements: lower, move,
forward, raise, set pressure. If the movement is not within acceptable
limits, all movement stops and the warning light will change to flashing to
indicate that human help is needed. The block diagram is shown in Figure 5-7.

HEAD END COMPUTER

MICROPROCESSOR ON EACH SHIELD COMMUNICATES BOTH WAYS WITH

HEAD END COMPUTER, SENDS TIMED COMMANDS TO TOP RAM AND TILT

WARNING

OBSTACLE
DETECTOR
EMERGENCY

Flow Diagram
 a

e. Computer Monitoring/Control SYatem Software Package. As

indicated in the network analysis outby problems could cause reduced shearer
capability. One example might be the failure of the outby conveyor to remove
the coal received from the face conveyor. Computer monitoring/control would
allow a means of slowing down or otherwise modifying the operation of one
piece of equipment so that another piece of equipment can operate within
design limits and not delay the whole operation by failing. The simple block
diagram is shown in Figure 5-8.

f. Auto Fault Isolation Software Package. The longwall system

can be subdivided into subsystems, such as the electrical distribution system,
the water system, the hydraulic support system, and the methane monitoring
system. Each major system can be broken into measureable variables such as
power at the headgate and power at the shearer. By monitoring the key
variables via sensors, faults in the system can be isolated. Knowing the
general failure characteristics of each component via the fault sensors allows
timely dispatching of repair parts and maintenance personnel. The block
diagram is shown in Figure 5-9.

2. Discussion Of Technology Integration Into Existing Longwall Systems

In assessing any proposed new technology, technical feasibility at

an acceptable cost is paramount. In making this cost/benefit tradeoff, mine
safety, ruggedness, and reliability are the main goals of the equipment
designer. The technology outlined in the previous section is servo mechanism
and digital electronics technology. Servo mechanism technology is already
well established for a variety of products, and rugged methods of implementa-
tion are available. Electronics of comparable complexity have been built for
oil well drilling instrumentation, manufacturing robotics, and military
applications. The technology thrust is to adapt existing computer knowledge
to the specialized longwall application.

In ranging arm articulation control, the designer will be concerned with

routing wires safely in a protected enclosure. One present longwall shearer
design already contains a microprocessor with the ranging arm actuators
electrically controlled. All that needs to be installed on the shearer
convert the ranging arms into servo mechanisms are feedback sensors.
system represents an example of the reasonable feasibility of auton,*=ng
longwall components. The rhysical location of the feedback sensors, }-awever,
present a bigger problem than the problem of selecting the appropriate
sensors. The sensor physical location problem may require extensive
modification of existing shearer structures. However, once modified, the
provisions for these sensors would be insignificant compared to the overall
cost of the shearer. The same problem would be experienced by the face
alignment measurement sensors in relation to modifying the shields and face
conveyor. In summary, the sensor and servo mechanisms technology is
available, along with an identification of the necessary system modifications.
However, the system modifications could be costly.

In the following sections costs are grouped as smart shearer costs,

shield cost and information/fault isolation costs. This grouping also
reflects the " sequence of the development effort. The benefits of work on the
shield are impossible to achieve unless the smart shearer is completed first.

5-21
 'a

ORIGINAL PAGE IS
OF POOR QUALM

Figure 5-8. Monitoring/Control Flow Diagram

Figure 5-9. Auto Fault Isolation Flow Diagram

Similarly, meaningful information/control display relies upon information

coming from the shearer. Therefore, the shearer automation option was
considered the most important candidate, followed by the shield and conveyor
options. The management information and fault isolation candidates could be
accomplished very ^asily once the above options have been installed.

3. Estimation of Cost and Time to Develop Technology

a. Development Time. The method used in estimating cost and

time was to draw on industry experience for developing similar hardware.
Where applicable, direct experience from mining equipment manufacturers was
also used (18). For example, automation instrumentation for oil well drilling
involving a microprocessor and sensors costs approximately $300,000 and took
12 mo of development effort before a prototype unit was field tested. In the
robotic field, specialized single-function hardware can be designed and
constructed in approximately the same time frame.

The development schedule appears to be tied to resource availability.

A one-year development schedule seems to be nominal for a complex project
involving state-of-the-art technology. Hardware work on several automation
opportunities would most likely proceed in parallel, thus allowing the most
complex piece of hardware to drive the development schedule. There is some
software interaction between the opportunities (i.e., the shearer and conveyor
sensor information). Based on industry experience, it is estimated that
sufficient software can be developed in the one year following completion of
the hardware to provide for a meaningful test. Additional software can be
developed as the test program progresses (such as integrating the management
information sensor imputs into the computer).

5-22
 . a•

After initial feasibility testing using breadboard equipment, the

envisioned concept of development is a joint effort by a U. S. government
agency, an American longwall equipment manufacturing company, and a U. S.
longwall shearer customer. Since an industrial equipment manufacturer would
be utilized, an additional 6-mo effort is needed in contract definition and
award of a contract. Since it may be necessary to modify customer-owned
equipment, additional time may be involved while waiting for a cooperative
manufacturing agreement. It is essential that the proposed automation system
components be integrated with a production longwall shearer in order to permit
the concept to be tested using the most reliable and readily available
components.

In summary, it appears that the minimum time for development will be 6

mo for contract definition, 12 mo hardware development after award of
contract, and 12 mo software finalization after completion of hardware. This
gives a total of roughly 2.5-3 yr minimum development time. In-mine testing
would follow this development time.

b. Development and Production Cost Estimates. As stated in

Section IV, the cost estimates were based on similar technology applications
in other industries using components of equal complexity. The one refinement
made in the case of the longwall system, as compared to the previous oil
drilling example, was that the automated longwall system would be 3-4 times
more complex owing to the extra degrees of freedom (i.e., roll, pitch, and
yaw). Thus, it is conservatively expected that the longwall development costs
would be proportionally higher. The development and production costs shown in
Table 5-4 below reflect this conservatism. The detailed cost elements are
shown in Appendix A.

Based on the above cost figures it appears that the total development
phase costs are on the order of S4 million.

D. MINER ACCEPTANCE AND HEALTH AND SAFETY IMPACTS

Miner acceptance information was obtained from the survey described in

Sections III and IV. This segment of the document concentrates on the results
of the health and safety analysis of the various options, and relationships to
the final cost-benefit projections.

1. Description of the Health and Safety Evaluations

The health and safety evaluations used in this report were

developed in accordance with methods developed for the Advanced Coal
Extraction Project (20, 21). The evaluations are performed at both a
qualitative and quantitative level. The analysis revolves around: (1) an
identification of hazardous system failures, (2) human interfaces with these
system failures (which comprises the hazard analysis, and (3) both a
qualitative and quantitative evaluation of design changes including a
projected impact on injury reduction. The qualitative analysis helps the
evaluator determine whether a new design offers sufficient merit over existing
systems to warrant further development, and also directs the evaluator toward
both strong and weak design areas that should be examined more closely. The

5-23

.. a

ORIGINAL PAGE 19
OF POOR QUALrrf

Table 5-4. Estimated Longwall Automation Development and Production Costs

Development Production
Item Cost Cost

Common automation elements (Table A-1) $ 860,000 $ 48,000

SHEARER ADD-ON COSTS

Ranging arm articulation (Table A-2) 340,000 3,000

Cowl articulation (Table A-3) 160,000 3,000
Face alignment measurements (Table A-5) 320,000 21,000

Smart shearer cost (Including common elements) $1,680,000 $ 75,000

SHIELD ADD-ON COSTS

Face alignment control (Table A-6) $1,100,000 $215,000

Shield advance automation (Table A-7) 500,000 35,000

Microprocessor-controlled shield add-on cost $1,660,000 $250,000

Note: During the breadboard phase, the face alignment computation package is
run on the above ground computer. During the preproduction phase, it is
anticipated that this software package will be run on a dedicated micropro-
cessor located near the headgate. The exact equipment configuration will be
decided after completion of the shield development phase.

FAULT ISOLATION / MONITORING / CONTROL ADD-ON COSTS

Development Production
Cost
Item Cost

Fault isolation (Table A-9) $ 360,000 $62,000
Monitoring/control (Table A-10) 128,000 3,000

Information/control add-on costs $ 488,000 $65,000

important qualitative measures for safety improvements are: (1) reduced

axposure time by improved equipment design and/or streamlined operations, and
(2) more worker protection. The health evaluation of new designs ends at this
stage because actual levels of exposure to elements like dust and toxic
materials can only be determined when a prototype is developed and operated
(20).

The quantitative injury projection converts the qualitative measures of

exposure time and protection into a means of numerically measuring potential
changes in historical injuries (21). Historical data suggest that as both
exposure time and number of workers exposed increase, the actual number of
injuries also increases. The model developed to project injuries assumes that
as each variable increases or decreases, there is a proportional increase or

5-24
decrease in injuries. Although the actual relationships are more complicated,
tests of the model demonstrate reasonable accuracy (22). The field of data
elements includes task times and descriptions, crew sizes, and protective
devices for both the proposed design and a piece of contemporary equipment
chosen for comparison. Also tabulated are historical injury experiences
related to major hazards associated with the various conventional tasks. For
each task and hazard, manhours and populations at risk are multiplied by the
historical injury rates observed for similar equipment. Total system safety
performance is then estimated by aggregating rates for various tasks and
hazards (21). Expert judgment can also be incorporated as an additional
weighting factor if desired. For the purpose of this initial examination of
longwall automation opportunities, this factor will assume a value of unity.
The injury projection equation is as follows:

ni = Ni t i /T i figi

where,

ni - the projected injuries for a given task i and hazard in the new
system (injuries/yr).

Ni = the total number of injuries associated with a given task i

and hazard of the contemporary system used as a comparison
(injuries/yr).

ti/Ti = the measure of the fractional change in task exposure time

between the new (ti) and contemporary comparison (Ti)
systems (dimensionless).

fi = the manpower ratio for a given task i and hazard of the new
and contemporary comparison systems (dimensionless).

gi = the injury adjustment factor, which reflects the consensus of

the group of experts pertaining to the safety integrity of the
new system (expressed as a fractional change in injuries).

The "expert adjustment" variable related to protective measures is

assigned qualitative values of large, medium, or small, relative to the
projected affect on reducing injuries. Later, during the consultation with
experts, this variable is given a numerical value. The expression for
consideration of protective aspects of a new design is as follows:

bj - Bjdj

where,

bj = the injury projection considering the incorporation of new

protective design measures for a given hazard and accident
class j (injuries/yr).

s
5-25
 bj - the aggregate number of historical injuries to the body
(or the initial exposure time injury projection) for a
givea hazard and accident class (injuries/yr)

dj - the general consensus of the group of experts pertaining

to the integrity of the new protective device (expressed
as a fractional change in injuries).

Since the automation opportunities apply to the existing longwall

designs, all the supporting task time, labor, and injury statistics used in
the projection were extracted from present longwall operational data and
Mining Safety and Health Administration (MSHA) injury history. Mormally, the
first stage of the safety evaluation requires that a detailed system failure
analysis be conducted. This facilitates identification of the hazards through
an understanding of how workers interact with potentially dangerous system
failures (e.g., bursting pressure lines). This level of detail is often
required because new designs may not resemble existing equipment sufficiently
to use historical injury experience as a means of revealing the hazards. This
is not the case with the proposed automation opportunities. Since the basic
design of the longwall remains unchanged, the hazards will not change, allow-
ing the historical hazard experience to be used. As a result, the system
failure and human interaction evaluation was not required for this analysis.
The historical hazard and injury data base is shown in Table 5-5. Four years
of injury data (1976-1980) * were averaged to obtain the yearly expected
injury levels. Task time and general labor involvement data required to
assess hazard exposure were also tabulated. These data were developed from
the network flow analysis.

2. Qualitative Health and Safety Evaluation

After identification of near-term automation opportunities and

evaluation of supporting data for the injury projection, the next step was to
examine each opportunity and relate it to hazard areas where a potential
improvement could be realized. Table 5-6 summarizes the relationships between
the automation opportunities and the hazards that could potentially be reduced.

The qualitative health and safety analysis can be performed using the
above table. The health hazards associated with dust inhalation would be
greatly reduced by removing the need for the operator and helper to be near
the shearer while it is cutting coal. Similarly, by removing the operator and
helper from the vicinity of the shearer, the noise hazard is somewhat reduced.
This would only be a marginal reduction since many other kinds of machinery
contribute to the noise hazard. Fatigue and psychological stress caused by
working in a cramped working space (such as operating a longwall in low coal)
would be favorably affected by automation or remote control in that the
operator and helper would not have to traverse the face with the shearer.
Table 5-7 summarizes the effects of automation on the major recognized health
hazards.

* Data obtained from MSHA, "Injuries by Worker Activity," Report CM341L2,

1979-1980.

5-26
 ORIGINAL PAGE
OF POOR QALRY

Table 5-5. Average Yearly Longwall Disabling Injuries (DI) by Task/Hazard

and Accident Class

M C
O M .^
u a .^
w o ~ i• Y t
&J
Ua ti "
OM^ Y`
r .+ ^i in :+ w .r3 w w iv a o
e • -e w v U v N •• a u Ma u .+
May or Task s .4 u pp Co r U r w -,
Mr
O a s U •M A YOa Y 6 Y a.9 M.^
Accident and u a F Fw o v .• o o .t o o u o a Y o a
class Hasard a it 7 MO it a it O it •Mal id U it M M' it Y r y ,+
O O it U V
e Jw
1. Uw • E w 6 COU 9 a
M7 ~ MwO M> +1
u O ^C
Y r1 y
.-1 VO UO UD Y• wwO wwQ UOM
NU W n Ua ri N U V! Y^ N U InU .4 NU YYY S 6 6 N w w

Roof/ fans/rib l
1 0 2 1 6 6 5 0 Fatality 6
falls
1 SDI

Haulage 2 0 2 1 5 5 0 3 0 1

Machinery 1 0 0 O 3 9 33 2 3 20

Handling 5 2 11 10 141 38 3 10 15 10
Material

Explosion/Fire 0 0 0 0 0 O 0 0 0 0

0 10 0 0 1 3 0 0 0 0
Electrical

Slip / fall 2 0 1 0 14 8 2 5 0 5

Handtools 0 0 0 0 0 0 0 0 0 1

Pressure release 0 0 0 0 0 2 0 O 0 1

5-27
 OFMNAL PANE 8
POOR 0UAMY

Table 5-6. Hazards Potentially Affected by Automation Incorporation

Activity/Hazard Affected Automation Opportunity

Health
Operator/helper exposed to dust, noise, Automation of the shearer, shield
and cramped workspace while operating advance, and conveyor advance, and
the shearer and advancing shields total environmental monitoring

Safety
Struck or caught while operating Fully automated or remotely
the longwall controlled shearer

Struck or caught while operating Fully automated or remotely

the conveyor controlled conveyor advance

Struck or caught while advancing Fully automated or remotely

longwall shields controlled shield advance

Explosion/ignition occurs Remotely controlled shearer, shield

while operating shearer or and conveyor operation
or advancing shields

Shocked while performing Fault isolation/diagnostics system

electrical maintenance for component failure recognition

Struck by high pressure hydraulic Fully automated or remotely

fluid release wh-".le advancing controlled shield advance
longwall shields

Table 5-7. Results of the Longwall Automation Health Analysis

Improvement
Health Hazards Positive Neutral Negative

Dust X
Toxic Compounds X
Temperature/Humidity X
Noise X
Vibration X
Poor Lighting X
Psychological Stress X
caused by cramped workspace

5-28
 a•

ORIGINAL PACE W
OF POOR QUALITY

The results of the qualitative safety evaluation suggested that

automation or remote control could provide both an exposure time reduction and
an increase in worker protection. The exposure time reductions largely affect
the hazards associated with the activities of operating the shearer (and being
struck or caught), operating or moving the longwall (and being stuck or
caught), and advancing longwall shields (and being struck or caught). The
fault isolation and diagnostics system would provide protection for the worker
while he is performing electrical maintenance or moving the shields. By
knowing where an electrical failure is located (such as a shorted cable), or a
pressure line failure on a shield, the worker would be notified of the hazard
in time to avoid injury.

3. Quantitative Injury Projection

The injury projection is the last stage of the safety evaluation

and utilizes the qualitative findings provided above, along with the task time
and historical injury data. The exposure time calculation for the use of
remote control or full automation is unimportant since the worker would be
totally removed for certain activities. Therefore, the exposure time drops to
zero. The following list of activities, when automated, will reduce hazard
exposure time to zero.

Accident Class Automated Activity

Roof/face/rib falls Operating the shearer. Advancing

longwall shields.

Haulage Operating or moving the conveyor.

Advancing longwall shields.

Machinery Operating the shearer. Operating or

moving the conveyor. Advancing
longwall shields.

Handling MLterial Operating the shearer. Operating or

moving the conveyor. Advancing
longwall shields.

Explosion/Fire Operating the shearer. Operating or

moving the conveyor. Advancing
longwall shields.

Slips and Fall Operating the shearer. Operating or

moving the conveyor. Advancing
longwall shields.

Hydraulic Pressure Release Advancing longwall shields.

5-29
 a^

GWINAL PAGE 19
OF POOR QUALITY

The two major areas where automation would provide protection are
electrical and machine maintenance. The accident classes affected are
respectively electrical, and hydraulic pressure release. The actual expected
decrease in injuries is determined by consulting experts in the field of
machine design and safety. Two experts were chosen from the Bureau of Mines
and M3HA (23, 24).

The final injury projection, considering both exposure time reductions

and improvements in protection, are shown in Table 5-8.

Although the exposure time reductions are straightforward, the

protection aspects are not. The reasons given by the two experts for the
expected reduction in electrical and pressure release injuries are presented
below.

Table 5-8. Longwall Disabling Injury Projection Considering

Automation Opportunities

Average Yearly Average Yearly

Accident Class Historical Injuries Projected Injuries

Roof/face/rib falls 46 35
Haulage 19 15
Machinery 71 16
Handling Material 245 222
Explosion/fire 0 0
Electrical 14 11 -13 (avg 12)
Slips/fall 37 25
Handtools 1 1
Pressure release 3 2

Total 436 328

a. Bureau of Mines

Electrical. Based on the information provided, it appears

that some minor improvement would be experienced by knowing where cable
failures have occurred. However, the total electrical system will be too
large to completely monitor. Consequently, c:nrkers are still unaware of the
potentially hidden failure that could cause injury. Also, many electrical
maintenance injuries are associated with poor practices, such as not shutting
off main power before performing maintenance. Even with fault isolation, this
error could still occur unless an auto shutoff system was included. Given
this rationale, perhaps a 10-15% reduction is feasible. A much larger
reduction (50x) could be realized with an automatic power disconnect included
with the diagnostics system.

5-30
 Hydraulic Pressure Release. A fault warning for the shield operator
would definitely provide more protection when advancing shields. However, no
improvement would be experienced in the maintenance area because the complete
hydraulic system would be too complex to monitor, and even if a hose or
fitting failure was isolated. the worker must still interface with the hazards
associated with residual pressure release or home whipping during component
replacement.

b. MSHA

Electrical. By knowing where (or in which component) an

electrical failure has occurred, workers will exercise more caution when
performing maintenance. However, poor maintenance practices (such as trying
to save time and troubleshoot systems while the power is still on) are the
major contributors to electrical injuries. Without an automatic power
disconnect included in the diagnostics system, only a 20-25% reduction might
be realized. With a fait-sefe feature, probably a 50-75% reduction would be
obtained.

Hydraulic Pressure Release. Fault isolation and warning could

definitely protect the shield operator. However, no improvement would be
experienced in the maintenance area since the worker must still interface with
the pressure failure during repair.

In summary, the above health and safety evaluation suggests that several
areas could be improved through automation. In the area of health,. it appears
that the greatest improvement would be realized through the reduction in
exposure to dust. In the safety area, the largest injury reduction (in terms
of the percent of the total injuries in a given accident class) would be
experienced in the machinery accident class. This reduction would then be
followed, in decreasing order, by pressure release, slip and fall, roof, face
and rib falls, and haulage. In all of the health and safety areas listed
above the related automation areas are shearer automation and automated or
remotely controlled shield and conveyor advance. Both of these automation
opportunities have approximately the came overall impact on injury reduction.
The use of the fault isolation opportunity effects a reduction in the
electrical and pressure release injuries. However, the net impact on overall
injuries is considerabl y less than the other two automation areas.

4. Health and Safety Cost-Bent-.fit Considerations

The last step in the health and safety analysis was to provide a
means of incorporating the results into the fic.al -.)Rt-benefit projection.
This was done by establishing reasonable estimates 'or the cost of an injury,
as well as the amount of lag time required 'or MSHA spproval of new
technology. The MSHA approval period was !!a important contributor to the
market penetration rate of the automation options.

5-31
 a
.

a. Cost of Serious Injuries. The cost-benefit indicators for

the safety analysis are based on a value of human life" concept. This
concept is defined as the "willingness to pny" approach developed by a study
conducted by Thaler and Rosen (25). The approach infers what people require
as compensation for risk by analyzing acceptable wage rates for hazardous
jobs. Although it is felt that these values still underestimate the true
value, or willingness to pay, of the population at large, they are considered
more realistic than the human capital approach. Since serious injuries are of
greatest concern for proper design of equipment and developing meaningful
cost-benefit indicators, the following value: ippear to be realistic:

Cost of fatality - $200,000 (in 1973 dollars)

Cost of disabling injury = $6,500 - $10 9 00C (in 1973 dollars)

b. Steps and Timeframes Required for MSHA Approval. The steps,

and approximate timeframes, shown in Table 5 -9 were established through
discussions with representatives of the MSHA Technical Support, Approval and
Certification Center in West Virginia (26). Due to the wide range of
hydraulic, mechanical, and electrical equipment approved by MSHA, the
timeframes for approval could vary considerably. The timeframes presented
here, therefore, are representative of the approval process.

Based on the data shown in the following table, it appears that the
total approval period ranges roughly from one to three years, depending
largely on the complexity of the new system and the similarity of the
experimental and production designs. Discussions with the MSHA
representatives suggested that a longrrall system, automated to the degree
indicated in this study, would require probably three years to approve.

E. THE COSTS AND BENEFITS OF AUTOMATION

The automation opportunit y descriptions and production costs were

established in Sections III and IV. The opportunity descriptions were used to
estimate the maximum possible benefits* (with emphasis on productivity) for
representative mines in the UniteO States.** Automation cost estimates
(capital and operating) were used Lo derive net benefit, which is the
difference between maximum possible `enef ; ts and estimated costs in one mine
section for an average year. Other Factors included in this stviy were:
(1) the size of the underground coal industry and the potential market share
of longwalls in that industry between the years 1985 and 2000, (2) the gate of
penetration of new longwall technology into the potential longwall market,
(3) the time delays associated with R&D, MSHA approval, and industry adoption,

* Maximum possible benefits means: the benefits for one year at one mine
section usinh zero cost automation. This establishes an upper bound (or
budget) for automation costs.

**Representative mines include low coal (48-in.-high seams) and high coal
(72-in.-high seams) cases. In some instances, calculations are made for new
mine sections and experienced mine sections.

5-32
 Table 5-9. Steps and Average Timeframes for MSHA Approval

Steps Required Approximate

for Design Approval Timeframe for Completion
(Months after receipt of design)

1. Manufacturer submits application On receipt

for component certification to MSHA

2. Application received and logged On receipt

3. Drawings or design specs are

requested by MSHA 3-12

4. Specs forwarded to appropriate (depending on complexity

MSHA test center for review of system)

5. Design reviewed by MSHA

6. MSHA notifies manufacturer 0.5-1

of test dates for prototype

7. Design deficiencies are assembled 0.5-1.5

and reported to manufacturer
(resubmittal process would begin
after completion of this step)

8. MSHA forwards certification 0.5

letter to manufacturer

9. MSHA/Bureau of Mines select a mine

for experimental testing

1n. Experimental permit is issued to 6-12

user. (mine operator) (depending on complexity of ay7teai)

11.. System/component is tested and

approved/not approved

12. Manufacturer submits application 3-12

of production design b steps 2-11 (depending on sim 4- 1 2rity of test
are repeated design to production design)

Total Time Required 13.5-39

5-33
and (4) health and safety benefits. The steps in the methodology are shown in
Figure 5-10. RED costs are not addressed, but they are small relative to the
potential net national benefits.

The longwall market is composed of the expected future expansion of

longwall equipment usage plus the eventual replacement of existing units.
Retrofit automation is not considered here. The time interval addressed is
1985 to 2000 because any period beyond this would have to consider advanced
technology which would cause large uncertainty in the results. The approsch
taken we, to use the most recent projections of underground coal mining made
by DOE's EIA (27) and to develop a classic market penetration curve (28) that
is consistent with historical data (29, 30), a current census of longwall.s
(31), and known constraints on longwall technology.

Penetration into the potential market, another major consideration, will

not be instantaneous. Three steps must occur in order to obtain any market
penetration: (1) successful R&D; (2) MSHA safety approval; and (3) initial
mine experience. Normal market growth can then occur if the new technology
provides a satisfactory return on equity investment and is relatively safe.
The technologies that r,eet these criteria are expected to penetrate the
longwall market at the same rate as similar historically successful coal mine
innovations.

National benefits, the last calculation in the methodology, are the

product of net mine benefits times the penetration into the potential market
over the time period 1985 to 2000. These results are discounted with a 7%
real discount rate to 1982 dollars and summed to yield a net national benefit
which can be used to compare to the RED budget available for this program.
The following analysis provides the detailed background and results for each
of the above evaluation elements.

1. Cost Methodology, Financial Requirements, and Potential Market

a. Cost Methodology. The cost estimation methodology selected

was Improved Price Estimation Guidelines (IPEC). This is a complete cost
:methodology w<<ich estimates the annual revenue required after startup to meet
all direct and indirect costs, including profit and taxes. The IPEG metho-
dology was chosen because it included all the important variables. This
includes costs of profits and income taxes, and all the other financial and
operational overheads that exist in industry. The methodology must assign
financial. overheads to the actual capital investment, and operating overheads
to actual operating costs. This is an essential feature whenever capital
investment (e.g., automation) is used to increase productivity or reduce
operating costs.

IPEG was originally developed for DOE's photovoltaics program in 1977

(32) and has been updated (31) and adapted for coal mine applications (see
Appendix B). This adaptation required an analysis of the financial conditions
of the coal industry.

b. Financial Requirements. A survey of financial literature

(34) indicated that r. 15% after-tax return on equity investment has been a
normal rate of return for the majority of successful coal firms. A JPL study

5-34
 ORIGINAL PAGE 19
OF POOR QUALITY

Ln cn I1
W W
O Q W
]C
~
Q

r.+ ~
cL F— Q O Z
cn p
W CD
W Fp— J N
O
O ~ ~ F-
^"^ W 111 W
pp Z t/7
O CY) w O N
Cl - r--I CL
U
W W Cn ^ ro
1-- F—
CD
Q ob
C^
G
d
U- u
O cn co W F -
U-V)
v
W W W S U G
Q)
Z m
Q Q C- I

p w 0
X
Q ti
W > W U
<Z:W O G
r
J
W i-
W m Q
cn
t N
a
W_ a- V) v
F— W
^ CLl cw/') v^
W W W
:J7 Cn Cr Z W
W C/) W 0
I
1 Ln
J J F¢—
Q ^ Q O O
a
W w
S W
^
• • •
ZZ FW
— ^' w

O
Q Ln
F— Ln W
Z ^ -
F^U-.
Y- IO Z
-:Z--
j W
W ~ .-r .... Z
W I •-- CL
W
m
W
W O
C).-
' CL U
CL C/7
O w
C^

5-35
(35) derived a method for obtaining the higher rate of return required for
innovations, given a normal rate of return and the lifetime of the invest-
ment. That study was based on research conducted by Mansfield (36).
Mansfield determined that a 252 rate of return on equity, after taxes, would
be required to make an innovation such as automation attractive to the coal
industry.

Therefore, two cost equations were developed for the projection; one
with a normal rate of return (152), and the second, with a higher rate of
return (25x).

IPEG emplovs a fixed charge rate (FCR) for equipment which depends on
equipment lifetime, a second fixed charge rate for land, long-life investments
and facilities, an overhead rate for labor, and another overhead rate for
materials and supplies. A constant charge per ton of coal is also included in
order to account for union welfare costs. The expression for a 152 return on
a nominal coal mine is:

FCR(15) x EQPT + 1500 x ACRE + 2.16 x DLAB + 1.18 x MATS +

1.385
QUAN

and the expression for a 252 return on a nominal coal mine is:

FCR(25) x EQPT + 3200 x ACRE + 2.33 x DLAB + 1.23 x MATS +

1.385
QUAN

where the FCR number is taken from Table 5-10, EQPT is the delivered and
installed cost of equipment in 1982 dollars, ACRE is the area of the mine in
acres, DLAB is the annual cost of wages for direct labor, maintenance
personnel, and foremen, in 1982 dollars, MATS is the annual cost of material,
supplies, and utilities in 1982 dollars, and QUAN is the tonnage of coal
produced in a year.

The result of this calculation is the revenue required per ton of coal
to meet expenses and make a return on investment. The financial and
operational overheads include profit, income taxes and credits, amortization
of investment, insurance, startup costs and revenues, working capital,
indirect labor, fringe benefits, rovalty payments, and other miscellaneous
expenses.

Table 5-10. FCR For Equipment

Equipment Lifetime

3 4 5 6 8 10 12 20

FCR(?5) 0.58 .).49 0.43 0.40 0.36 0.33 0.32 0.29

FCR(25) 0.81 0.71 0.66 0.62 0.58 0.56 0.55 0.55

5-36
 t
1

c. Potential Market. The potential market for longwall

innovations consists of future expansion into new sections and replacement of
worn out units. Historically, longwalls have been increasing their share of
the underground coal mining industry since the 1960's at a rate which closely
matches a classic "S"-shaped curve, sometimes referred to as a logistics
curve. A general formulation of the logistics curve is the followings

K2
Market Share - K 1 + +BxT+CXTxT
1 + e A

where T is time expressed in years, and Kl + K2 is the upper bound for

market shar p . This equation is nonlinear with parameters K1, K2, A t B,
and C which must fit historical data or reliable theory.

The approach taken for calibrating this curve for longwall market share
was as follows:

(1) Establish an upper bour.l on market share based on the fact that
372 of today's underground production comes from mines which
produce less than 200,000 tons annually and are too small to use
longwalls (37). This statistic has not changed significantly
since 1962 (30). Furthermore, eves. in longwall mines, some 202 of
the coal is obtained by other means during mine development and
initi it tunnel construL , ion. This sets .,n uprAr bk..ind on
longwalls of SOX of total underground coal.

(2) Using 1995 as the point at which T - 0, a form of nonlinear

programming called a line search was used on each parameter to
determine the values which best fit the historical data (38).

(3) The result of the previous step was used as a trial solution, and
these results were compared to a projection made by Kuti (39), who
used a substantiall y different method. The projections used in
this report are somewhat more conservative than Kuti's. He
predicted 12% longwall penetration by 198S, the logistics curve
produces 11% for that y ear. Furthermore, the rate of expansion is
consistent with data that received from industry (31, 40).

Both the longwall industr y size, in terms of tons per year, and the
annual potential markets are given in Table 5-11. The annual market is
subdivided into expansion into new sections and replacement of old sections,
and is further subdivided into low and high coal. This breakdown is shown in
Table 5-12. The subdivision into low and high coal is based on the historical
data on longwall seam heights provided in Reference 41. This distribution is
shown in Figure S-11. The number of sections comes from the estimates of tons
per section, for low and high coal, directly resulting from the development in
Section V.B of this report. These figures are summarized :n Table 5-13. They
are somewhat higher than historical averages in the U.S. due to expected
improvements in future equipment and procedures (30).

S-37
 ' a

ORIGIWI L F 4'r Z'¢

OF POOR

Table 5-11. Underground, Longwall, and Potential Langwall Market

Tonnage Projections (millions of short tons).

Year Undergrounds Longwallb % Annual Marketc

1980 335 22 6.5 --

1981 350 25 7.71
366 20 8.r --
1983 382 34 8.9 --
1984 400 40 9.9 --
1985 418.0 46 11.0 7.3
1986 461 55 12.1 11.9
1987 509 68 13.3 14.9
1988 561 82 14.6 17.3
1989 619 99 15.9 20.0
1990 683.5 118 17.3 23.5
1991 728 137 18.8 22.5
1992 776 158 20.3 25.5
'193 826 180 21.8 28.4
1994 880 206 23.4 31.7
1995 937.6 234 25.0 38.3
1996 999 266 26.6 43.1
1997 1064 300 28.2 48.1
1998 1133 337 29.7 53.8
1999 1207 377 31.2 60.1
2000 1286 620 32.7 61.8

aBased on historical data reported by the United States Bureau of

Labor Statistics and on projection by DOB's EIA (the low oil price
scenario in Ref. 12). DOE projections were available for 1985, 1990,
and 1995 (30).

bnerived from the total underground tonnage times the percentage of

penetration. Penetration percentage was derived by fitting a general
logistics curve on historical data, plus an estimate of maximum
penetration by longwalls. See Appendix C.

c lncluac3 expansion plus replacement of 10-yr old units. See

Appendix C for the age distribution of most of today's units.
Estimates for years prior to 1985 are not necessary for this study.

5-38
 .y

ORIGINAL, PAGE t$
OF POOR QUAtt'IY

Table 5-12. Underground Longwall Annual Markets

(Number of Sections)

Replacement Expansion
Year Low Coal High Coal Low Coal High Coal

1985 1 2 7 9
1986 2 3 12 14
1987 3 4 14 17
1988 4 4 16 21
1989 4 4 19 25
1990 4 5 22 30
1991 4 6 21 27
1992 6 6 23 31
1993 7 7 26 34
1994 7 8 29 38
1995 10 14 31 43
1996 14 16 34 47
1997 15 20 38 51
1998 19 23 42 55
1999 21 28 45 60
2000 19 26 49 65

Table 5-13. Baseline Annual Section Production Rates

(1,000 Tons/yr)

Type of Longwall Section High Coal Low Coal

New Sections (inexperienced operator) 472 262

Replacement (experienced operator) 500 281

5-39
 1

O
O_
ORiu1NAL PAGE 19
OF POOR QUALITY co
m
co
o)
v
N
rn
O
m
co
w

00
Ict
m
CD a
N G
co
ao
a
O
co
co ro
n 3
(D C G
n •_ C

a ^ c
n =
N
n W G
d
O =
n ^ m v
0o Q
(D
cn ^S
CD
CD J

a
to FE
N :r
(D
O
(D
co
U) 1
.n
(D
U')
L
a
U)
N
U)
i
O
U)

00

(D

N
a
O
It
co
Cl)

(D

O a ao n 0 L V M N

S13NVd 11dMDNO1 3C 838vynN

5-40
 2. Cost-Benefit Estimates for R&D Options

This section provides the estimated net national benefits of each

R&D option using the methodology described above. In review, the following
automation options were examined:

1A. The smart shearer

IB. The totally automated (no operator) shearer

2A. Conveyor/face alignment

2B. Dumb shield advance

3. Monitoring and control of environment and haulage systems

4. Fault isolation and diagnostics

These options have common elements so that costs and benefits of each of
the six by itself will not add up to the correct cost and benefit for the
group as a whole. Furthermore, options lA and 1B are mutually exclusive. All
other combinations are mutually compatible. Options 2A and 2B should go
together. The maximum benefit comes from the combination of options 1A, 29
and 3. The combination of 1A, 2, 3, and 4 was also examined and is called the
"combined automation" option. Each option is also explored separately.

a. Productivity Calculations. Table 5-13 provided baseline

annual production levels for several kinds of longwall coal mines. As
explained earlier, each automation option increases productivity per section
by reducing some of the time spent on delays (see Section V.B). Table 5-14
summarizes the resulting annual productivity increase in terms of thousands of
tons of coal per section per year. Table 5-15 gives the same information in
terms of thousands of dollars of increased net revenue per year per section in
new mine sections. Net revenue is the difference between total revenue, or
$35/ton, and tonnage-related costs. Tonnage-related costs only include union
welfare ($1.385), materials and supplies ($9.16), and power costs ($0.44), for
a total of $11/ton (42). Net revenue is then $24/ton, in 1982 dollars. Table
5-15 strictly represents the maximum possible economic benefits from these
opportunities. Tahle 5-16 contains the estimated automation production
(capital) cost to a mine owner for each opportunity.

Table 5-17 gives the net benefits, including the automation costs, for
each opportunity. Opportunities 1 through 3 all have positive net benefits.
Opportunity 4, fault isolation and diagnostics, does not seem practical unless
it can share some of its microprocessor and communications costs with the
other automation opportunities. Table 5-17 is based on stand-alone automation
except for the ombined automation option. Table 5-18 shows the effect of
deleting various opportunities from the combined automation option.

Opportunity 1B, complete automation of the shearer, is not included in

these combinations because of difficulties in estimating the cost and time
required to successfully pursue this option. Complete automation presents

5-41
 Table 5-14. Summary of Productivity Increases
(1000 Tons Per Section Per Year)

New Section Old Section

Opportunity Area Low High Low High

IA. Smart Shearer 61 100 42 72

1B. Automated Shearer 80 127 61 99

2. Shields and Conveyor 59 89 40 61

3. Monitoring and Control 19 28 19 28

4. Fault Isolation 5 8 5 8

5. Combined Automation 114 177 95 149

Table 5-15. Maximum Possible Benefits ($1000 of Revenues

Per Section Per Year)

New Section Old Section

Opportunity Area Low High Low High

IA. Smart Shearer b 1464 2400 1008 1728

t
1B. Automated Shearer b 1920 3048 1464 2376

2. Shields and Conveyor b 1416 2136 960 1464

3. Monitoring and Control 456 672 456 672

4. Fault Isolation 120 192 120 192

5. Combined Automation 2136 4248 2280 3576

s Revenue is the average price of coal minus tonnage-related costs in

1982 dollars. Automation costs are not included in these figures.

bThese options could lead to a reduction in section workers of

1-2 persons per shift for an additional savings of $134,000 to $268,000
in direct and indirect labor costs. This is not included in this table.

5-42
 a
i

Table 5-16. Costs of Automation (Per Section)

Initial
Capital Labor Annual
Investment Per Costsa
Opportunity Area (1982 $) Shift (1982 $)

IA. Smart Shearer 300,000 b 1749000

1B. Automated Shearer 600,000 b 348,000

2. Shields and Conveyor 308,000 b 179,000

3. Monitoring and Control 158,000 0 92,000

4. Fault Isolation 207,000 1 254,000

5. Combined Automation 385,000 Oc 223,000

aBased on a 3-yr service life for equipment, the IPEG coefficient used
is 0.58.

bone to two persons could have been eliminated but this is not included
here due to concerns about union requirements.

c In this case, the extra mechanic required for opportunity 4 can be

exchanged for the available worker due to opportunity lA and 2.

Table 5-17. Net Benefits Per Section

($1000 Per Year)

New Section Old Section

Opportunity Area Low High Low High

IA. Smart Shearer 1290 2226 834 1554

1B. Automated Shearer 1572 2700 1116 2028

2. Shields and Conveyor 1237 1957 781 1285

3. Monitoring and Control 364 580 364 580

4. Fault Isolation -134 -62 -134 -62

5. Combined Automation 2513 4025 2057 3353

5-43

a`

Table 5-18. Combinations of Automation Opportunities

(Net Benefits Per Section Per Year in $1000)

New Section Old Section

Combinations Low High Low High

lA, 2, 3 9 & 4 2513 4025 2057 3353

1A, 2, & 3 2428 3868 1972 3196

1A, 2, & 4 2057 3359 1607 2687

lA & 2 1978 3202 1522 2530

substantial difficulties, in that it is possibly beyond today's state-of-the-

art sensor technology and is not particularly desirable to the worker. The
time required to conduct the necessary R&D is several years beyond what would
be required for the smart shearer.

Time delays before market penetration are an important factor in

calculating the present value of national benefits. Three periods of delay
were identified: (1) an initial R&D period leading to a research prototype;
(2) a MSHA approval period, and (3) time required by industry to get field
experience with a production prototype. A schedule of delays is given in
Table 5-19. This schedule could change after funding levels for R&D are
decided upon, but the relative values should not change.

After these delays, the new technology can diffuse into the potential
longwall markets described in Table 5-12. This study assumes that this
diffusion will occur at the same rate as original penetration of longwalls
into underground mines. Appendix C, Market Size Projections, describes the
approach taken in greater detail. The diffusion of automated longwalls into
the future longwall population is given in Table 5-20 for each of the six
opportunities and combinations listed in Table 5-19. The monitoring and
control opportunity (14) is most effective in mines which use rail transport,
which will be used in some older mines and in the larger new mines (43).
Unfortunately this is a small subset of the total potential market available
to the other opportunities. Therefore, this would have the net effect of
reducing the diffusion rate. Annual benefits and diffusion rates are then
converted into annual net benefit terms in Table 5-21, and these are
discounted at a 72 real rate for a present value result. The computations
were simplified by the following procedure; for each opportunity,it was:
(1) determined whether the mine operator would receive the return on

5-44
It

Table 5-19. Time Delays for Market Penetration (yr)

Opportunity Area RED MSHA Industry Total

1A. Smart Shearer 3 2 2 7

1B. Automated Shearer 9 2 3 14

2. Shields and Conve yor 3 3 2 8

3. Monitoring and Control 4 1 2 7

4. Fault Isolation 4 1 2 7

5. Combined Automation 4 3 2 9

6. lA a 2 3 3 2 8

investment required to tr y the new technology (i.e., at least 25%) under

different conditions, then (2) for those conditions for which the previous
requirement was met an average net benefit per ton was calculated, and,
finally (3) that result was multiplied times the fraction of the market given
in Table 5-20.

b. Annual and Total Health and Safety Benefits. Impacts of

automation on miner health and safety were in Section V.D of this report.
Those impacts were expressed in terms of average yearly injuries. This
section extends those estimates to annual and total impacts based on the
penetration model developed for economic benefits.

Historical injuries in longwall operations over the time period 1976 to

1980 averaged 436 injuries/yr. Because longwall installations are increasing
in number, this injury figure should also increase proportionately. Table 5-22
gives a projection of injuries in longwall coal mines, assuming that average
injury rates per ton of production will not be reduced. Table 5-23 provides
estimates of the total proportional injury reductions that could occur as a
result of automation opportunities, based on the projections made in Section
V.D.

In summary, opportunities 1A, 1B, and 2 offer good improvements in safety

costs. Opportunity 3 does not lead to injury reduction, and opportunity 4, by
itself, might reduce injuries slightly but fails to diffuse into the industry
for economic reasons. Opportunities 6 and 7, similar to 5, have identical
impacts on injury reductions.

5-45
 Table 5-20. riffusion of Automation Opportunities
Into the Longwall Coal Mining Industry (x)

Opportunity
Year 1A 1B 2 3 4 5 6

1989 0 0 0 0 0 0 0

1990 2.0 0 0 1.0 0 0 0

1991 2.3 0 2.0 1.1 0 0 2.0

1992 2.6 0 2.3 1.3 0 2.0 2.3

1993 2.9 0 2.6 1.4 0 2.3 2.6

1994 3.2 0 2.9 1.6 0 2.6 2.9

1995 3.7 0 3.2 1.8 0 2.9 3.2

1996 4.1 0 3.7 2.0 0 3.2 3.7

1997 4.6 2.0 4.1 2.3 0 3.7 4.1

1998 5.2 2.3 4.6 2.6 0 4.1 4.6

1999 5.8 2.6 5.2 2.9 0 4.6 5.2

2000 6.5 2.9 5.8 3.2 0 5.2 5.8

The avoided injuries in Table 5-23 provide economic end noneconomic

benefits. An economic measure of those benefits is $10,000 per injury in 1973
dollars (see Section V.D), or $20,000 in 1982 dollars. Noneconomic benefits
include reductions in dust by removing the operators. As an overview, Table
5-24 summarizes both the benefits productivity and safety for each opportunity.

5-46
 Table 5-21. Present Value of National Benefits
(Millions of 1 y82 Dollars)

Opportunity

Year 10 1Ba 2a 3b 4b 5c 6c 7c

1989 0 0 0 0 0 0 0 0

1 0,90 8 0 0 1.5 0 0 0 0

1991 11 0 9 1.9 0 C 14 16

1992 15 0 12 2.6 0 19 18 22

1993 19 0 15 3.2 0 25 24 28

1994 23 0 19 4 0 33 30 35

1995 31 0 24 5 0 41 38 44

1996 39 0 32 7 0 52 50 58

1997 49 25 40 9 0 67 62 73

1998 62 32 51 11 0 84 78 9i

1999 78 41 64 14 0 105 99 116

2000 97 51 79 17 0 133 123 144

Net
Present 158 48 124 28 0 199 193 226
Value

aStand-alone R&D options. Opportunity IA is the smart shearer, 1B is a

totally automated shearer, and 2 is an automated shield and conveyor
system. All three include face alignment.

b Stand-alone R b D options. Opportunity 3 is mine monitoring and control,

and 4 is fault isolation and diagnostics.

c Combined options. Opportunity 5 includes 1A, 2, 3, and 4. Opportunit y 6

includes IA and 2 only, and 7 includes 1A, 2, and 3.

5-47
 a^

Table 5-22. Annual Injuries in Longvall Coal Mines

Year Injuries

1989 3100

:990 3700

1991 4300

1992 5000

1993 5700

1994 6500

1995 7400

1996 8400

1997 9500

1998 10700

1999 12000

2000 13300

5-48
 Table 5-23. Annual and Total Safety Benefits from
Longwall Automation (Injury Reductions)

Opportunity
Year lA 1B 2 5 6 6 7

1989 0 0 0 0 0

1990 7 0 0 0 0

1991 10 0 11 0 19

1992 13 0 14 23 26

1993 17 0 18 31 33

1994 21 0 23 40 42

1995 27 29 50 53

1996 34 0 39 63 70

1997 44 19 48 82 87

1998 56 25 61 103 110

1999 70 31 77 129 140

2000 86 39 96 162 173

Totals 385 114 416 683 753

5-49
 a,

Table 5-24. Summary of Total Benefitsa

Opportunity Productivity Safety Healthc

($1,000,000) ($1,000,000) (Injuries)b

1A 158 7.7 385 +

1B 48 2.3 114 +
2 124 8.3 416 +
3 28 0 0 0
4 0 0 0 0
5 199 13.7 683 +
6 193 15.1 753 +
7 226 15.1 753 +

a Based on inflation calculated from the GNP price deflator index,

provided by the Department of Commerce.

bThe total reauction in injuries for the years 1990 to 2000.

C A plus (+) indicates an improvement in health conditions by reducing

worker exposure to dust.

5-50
 SECTION VI

SUMMARY AND CONCLUSIONS

A. OVERVIEW

The previous section provided the quantitative cost-benefits associated

with: (1) the difference between increase in production and projected capital
and operating costs, and (2) the numher (and cost) of the reduced disabling
injuries. The qualitative health and improve-Ants (primarily in reduced
exposure to coal dust) were also it"^cated. In order to appreciate fully the
implications of the various automation options, in terms of their respective
effects on the longwall mining industry, it is important to examine the study
findings in total. This section first summarises the results of the
productivity, technology assessment, health and safety, and cost benefit
evalit&tions. Second, all the various options are ranked by the magnitude of
heir respective potential impacts in each of the evaluation areas. Finally,
a technology development plan is suggested, based on the ranking of automation
C 11.0t.b.

B. SUMMARY OF RESULTS

The results of the productivity analysis suggested that removal of

delays associated with shearer stalling, face alignment, and operation would
greatly improve available shearer production time. Similarly, elimination of
face alignment problems which overstress the face conveyor or cause shield
interference would also enhance productivity. Some maintenance delays would
also be favorably impacted by incorporating improved management information
and fault isolation systems. Overall, almost a 401 improvement ir, production
was projected for the case where all the automation options were incorporated.

The technology assessment area indicated that all of the various

automation options could be developed using state-of-the-art technology in
sensors, servomechanisms, comparators, digital electronics, and computers.
This finding was based on the results of the conceptual designs presented in
Section V.C. Consequently, development and production costs were not
projected to be beyond a first order of magnitude, measured in millions of
dollars invested.

The health and safety impact evaluations provided some interesting

insights into longwall mining hazards. These results suggested that
automation of the shearer, shield, and conveyor operations would allow
approximately a 23% reduction in average y early disabling injuries as well as
reduced exposure to respirable coal dust.

Finally, the cost-benefit analysis examined the net economic benefit

realized from each of the automation options as a function of benefits
(increased productivit y and safety) less costs (capital and operating).
Overall, it was projected that incorporation of all the options would result
in a net national benefit of roughly $200 million. The major controlling
factor in the cost assessment was the influence increased production had on
offsetting the costs.

6-1
C. IMPLICATIONS OF RESULTS

The above results have strong implications for the structure of an

overall technology development plan. As a first step in developing the logic
of such a plan, it is useful to rank the various options in order of relative
impact.

1. Production Implications

Each of the automation options can be ranked by its respective

impact on production by using the results of Table 5-14 (from the cost-benefit
analysis). These values were calculated by taking the incremental time
savings for each automation area calculated from the network analysis, and
converting time into production. Each of the options is ranked in Table 6-1
as a 'unction of operator experience and seam height.

It is obvious from the above - asking that shearer, shield, and conveyor
automation make up the bulk of tht verall productivity improvement;
particularly, where an inexprrien^ ­ o operator ( having to make more trim cuts)
is working under ideal condtticn., tt► ign coal). In other words, automation of
the three major areas would allow ' ie longwall system to operate as projected
under ideal conditions, indep.?ndent of any operator delays associated with
skill. It should be noted that Lotal shearer and shield automation were not
considered practical due to the complexity of such a system. Additionally, it
was indicated in Section III that workers would still be required to take care
of non - routine tasks such as ground control. Therefore, it would be
impractical to attempt to fully automate a system where workers are a
necessary element to make the system operate properly.

Table 6-1. Ranking of Automation Options by Impact on Coal Production

(1000 Tons Per Section Per Year)

Inexperienced operator Experienced Operator

Automation Option Low Seam High Seam Low Seam High Seam

Remotel-r Controlled
Shearer 61 100 4? 72

Remote Shield anA

Conveyor Advance 59 89 40 61

Management Infotma-
tion System
(computer monitoring) 14 28 19 28

Fault Isolation 5 8 5 a

Combined Automation
(all of the above) 114 177 9S 149

6-2
 i

Table 6-2 presents the net impacts of all the automation options on
productivity in terms of percentage improvements over the original baseline
figures.

Table 6-2. Percentage Improvement Over the Baseline Productivity

Operator Skill High Coal Low Coal

New sections (inexperienced) 38 44

Old sections (experienced) 30 34

Overall average 37

2. Technology Implications

The technology choices are more difficult to rank. Table 5-10 in

the previous section indicated that the largest portion of the development
costs would be consumed by the shearer, shield, and conveyor options.
Although these components are the largest cost contributors, their relative
ranking in relation to the other automation options must consider other
factors besides cost. These other factors include: (1) choice of options
that achieve the best results for the money invested, and tt) choice of
options that represent the best development sequence to achieve the most rapid
pavback or investment. When these factors are considered, it becomes obvious
that automating the shearer, shields and conveyor achieve the best results in
terms of seeing a rapid payback through increased production. Based on these
conclusions, Table 6-3 gives the ranking of the options based on the impact on
production.

It is important to recognize that the total cost cited in Table 6-3 makes
no provision for amortization of expenditures. In fact, the total investment
cost could be raised by two to four times higher (i.e., up to $20 million) to
be conservative, and the total payback, based on the previously calculated
$200 million net benefit, would still be an order of magnitude larger.
Therefore, it is clear that the pote-tial for improved production far
outweighs the investment in RED.

3. Health and Safety Implications

The ra •.iki n g of the automation options as a function of health and

safe_v impacts resulted in a slightly different order than the above
rankings. The health impacts revolved largely around reduction in exposure to
dust. The safety impacts resulted primarily from the fact that hazaris

6-3

'a

Table 6-3. Suggested Automation Technology Development Sequence Based

on Impact on Investment Payback

Impact on Payback Total Investment

Automation Option Large Medium Small (million $)

Common Elements (computer,

sensor communication,
software) X 1.0

Remotely Controlled Shearer X 2.0

Remote Shield and Conveyor

Advance X 2.0

Management Information
System (computer
monitoring) X 0.1

Fault Isolation X 0.4

Approximate Total Cost 5.5

associated with being caught, being struck, or slipping and falling while
advancing the shields and conveyor, comprised the largest fraction of serious
longwall injuries. Table 6-4 provides the ranking of the automation options
as a function of the potential reduction in average yearly disabling injuries
and overall health improvements.

4. Cost Benefit Implications

The cost-benefit implications associated with each of the

automation areas were clearly indicated in Section V.E. It, therefore,
suffices to simply provide the ranking of the options based on the net
benefits cal-ulated for the productivity and safety area. This is provided in
Table 6-5.

Table 6-5 indicates that, overall, the net benefit of incorporating all
the automation options results in roughly a $200 million profit to longwall
sections. Most of this profit comes from the improved production potential of
automated longwall systems.

D. RECOMMENDa:rIONS

The findings presented in the preceding section have clear implications

for the evolution of longwall automation. Clearly, t'r:e best plan of action
should be to automate those areas where: (1) the most immediate and sizable

6-4
 a

Table 6-4. Ranking of Automation Options Based on Health and Safety Impacts

Reduction in Average Reduction in Exposure

Automation Option Yearly Injuries (X) to Dust
Large Small

Remote Shield and

Conveyor Advance 12 X

Remotel y Controlled
Shearer 10 X

Fault Isolation 1 X

Management Informa-
tion System
(computer monitoring) 0 X

Total Reduction in
Disabling Injuries 23

Table 6-5. Ranking of Automation Options Based on the Net Total Cost
Benefits (Nationwide) for the Years 1989 to 2000

Productivity Safety Total

Automation Option (million $) (million $) (million $)

Remotely Controlled Shearer 158 7.7 165.7

Remote Shield and Conveyor

Advance 124 8.3 132.3

Management Informa-
tion System
(c,mputer monitoring) 28 0 28

Fault Isolation 0 0 0

Combined Automation
(all of the above) 199 13.7 212.7

6-5
payback on R&D will be realized, (2) the most convenient and cost effective
expansion of the system is feasible, and (3) the largest improvements in
worker health and safety are realized. The one consistent element in all of
the preceding rankings was the positive impact shearer, shield, and conveyor
automation had on productivity, the technology development sequence, health
and safety, and cost benefits. Therefore, the first step recommended is
development of the smart (remotely operated) shearer. It is essential to
incorporate the proposed shearer automation package (see Section V-C.) because
it contains many of the spatial sensors which become integrated with the
shield and conveyor automation package. As part of the shearer package, it i•
also suggested that a surface central computer system be developed to: (1)
establish the appropriate sensor data processing links with the shearer, (2)
allow the appropriate command, guidance, and control software to be designed,
and (3) allow the sensor, guidance, and control subsystems on the shearer to
be tested for accuracy, and calibrated. Once the surface computer and shearer
package have been developed, it is recommended that the remaining sensors,
guidance, and control elements be developed for remote shield and conveyor
advance. Following the successful integration of the shearer, shield, and
conveyor automation, the management information and fault isolation options
should be developed. Although additional sensor and information retrieval
subsystems may have to be installed, the data linkup, processing, and display
system will already be in place via the surface computer system. This will
jointly simplify the process of installing and testing other desired fault
isolation sense-s, and permit expansion of the total information system. This
approach toward developing a management information and fault isolation system
seems reasonable considering that a similar system has already been
successfully implemented at the main Canelton Coal Company facility in West
Virginia.

It is appropriate at this time to compare the major findings of this

study with the results of the Skelly and Loy and COMINEC analyses. At the
onset it should be stated that the automation options developed from the
survey and network analysis were basically the same as those derived from the
previous studies. One major difference between the respective analyses was
the projected 40% versus 60% improvement in productivity. The original 60%
figure was based on: (1) a projected decrease in face personnel, (2) more
efficient shearer operation, and (3) an increase in shearer traverse speed (by
not being limited by operator mobility). In keeping with the overall
philosophy of conservatism throughout the document, it was decided that union
restrictions concerning Lhe number of workers required for various jobs could
adversely affect the chances of realistically reducing the face crew size.
Additionally, the study indicated that where one or two face personnel could
be removed by automating the shearer and shield operations, one or two extra
moving mechanics would be required to perform maintenance in support of the
fault isolation system (sea Table 5-16). Therefore, is % yas decided that the
potential For reducing face personnel was not a viable assumption in the near
term.

By comparison, all the studies were in agreement concerning the expected

boost in production related to improved efficiency in shearer operation (which
included maintaining face alignment). However, the results of this study
diverged slightly from the two previous studies on the last element concerning
increased shearer speed. It is char that under certain circumstances (such
as low coal), operator mobility can pace the cutting rate of the machine.

6-6
 .a

Similarly, many ocher variables such as soft floor, poor roof, or methane can
prevent the machine from operating at maximum speed. Therefore, rather than
assume that removal of the operator generally allowed a higher cutting rate,
it was decided to keep the traverse rate of the shearer (regardless of
automation) at the present average speed. Ultimately, this nominal value
provided a rate which was not only representative of the industry in general,
but it also allowed the study to retain its conservative (or worst case)
approach. One last minor point of difference between the results of the
respective studies stemmed from the management information and fault isolation
options. Although not addressed in the COMINEC study, the Skelly and Loy
analysis suggested that options would be required to monitor both the
environment (for methane) and machinery failures since workers would not be
present. The results of the network analysis in this stu..y show that these
• -ptions actually have an impact on reducing: (1) failure identification and
access time, and (2) bottlenecks in outby haulage and supply delays.
Consequently, it was discovered that these options could have an impact on
increasing production time.

Overall, it seems that the analysis and results provided in this

document not only confirm results of the previous studies, but provide a more
substantial basis for the conclusions. For example, the safety analysis draws
on a detailed historical injury data base (by worker activity) for the safety
assessment. Similarly, the cost benefit analysis considered variables such as
market penetration, future coal demands, and new longwall panels to accurately
project the potential costs and profits of automation. In both of the above
examples, well-constructed and tested methodologies were employed in the
system assessment. Of key importance throughout the study was the necessity
to not be too optimistic about the potential impacts of automation. This was
achieved by using conservative values for the productivity, cost, and
cost-benefit calculations where possible.

In conclusion, it seems that if the above automation development plan is

adopted, several key areas would be addressed: (1) large productivity
improvements could be realized, (2) cost benefits appear to far outweigh R&D
investment costs, (3) worker health and safety could be greatly improved, and
(4) the redesigned system would be responding to guidance, control and
maintenance problems presently recognized as production inhibiting factors by
operators, workers, and mine management.

6-7
 SECTION VII

REFERENCES

1. Bendix, "Automated Longwall Guidance and Control Systems," Phase I and

Phase IT Fin&. Report, NASA Contract NAS 8-32921, June 15 9 1979.

2. Little, A.D., "Last Cut Survey," DOE Contract No. ET-76-C-01-8880,

undated.

3. Mining Journal, "Methods and Machines-Natural Radiation Utilized in

Guiding Coal Cutting Machines," April 12, 1982.

4. SET, "Type 801 Natural Coal Thickness Indication System," Salford

Electrical Instruments Ltd., England, undated.

5. GIT" "Applications of Acoustics in the Measurement of Coal Slab

Thickness," Final Report, Georgia Institue of Technology, February 16,
1980.

6. Wyle Laboratories, "Baseline Studies on the Feasibility of Detecting a

Coal/Shale Interface with a Self-Powered Sensitized Pick," February 1,
1981.

7. Adjunct Systems Inc., "Design of a Laser System for Instantaneous

Location of a Longwall Shearer," January 1981.

8. Science Applications Inc., "Design for a Laser Noncontacting Last Cut

Follower for Longwall Mining Operations," December 10, 1979.

9 Eickhoff Corp., "Eickhoff Exhibits Trend Setting," Mining Exhibition in

Dusseldorf, Germany, December 1981.

10. Skelly and Loy, "Evaluation-NASA Coal Shearer Guidance System," Final
Report, October 24, 1980.

11. COMINEC, "Conceptual Design of an Automated Longwall Mining System,"

Voiumes I and II, Final Report, April 1976.

12. Kuti, J., "Longwall Mining in America," Mining Engineering, November,

1973.
13. Hislop, James, chief engineer, Mining and Exploration, Utah Power and
Light Company, private communication, 1982.

14. Hemscheidt American Corp., "Shield Supports; A Technology riscussion,"

Undated.

15. Van Cott, H., Kinkade, R., Human Engineering Guide to Equipment Design,
Joint Army, Navy, Air Force Steering Committee, Mcgraw Hill Co., 1972.

1;, Heskett, J., et.al ., Business Logistics-Management of Supply and Demand,

Ronald Press Company, New York, 1964.

7-1
17. Ackoff, R., Sasieni, M., Fundamentals of Operations Researchp John Ailey
and Sons, Inc., New York, 1968.

18. Joy Mfg., redesign of shearer model ILS4-43D59XE with microprocessor,

private communication, 1982.

19. Akin, G., "Psycho-Social Aspects of Productivity in Underground Coal

Mining," JPL Publication 81-113, Jet Propulsion Laboratory, Pasadena,
California, 1981.

20. Zimmerman, W., "Health Requirements for Advanced Coal Extraction

Systems," JPL Publication 80-72, Jet Propulsion Laboratory, Pasadena,
California, September 15 9 1980.

21. Zimmerman, W., "Safety Evaluation Methodology for Advanced Coal

Extraction Systems," JPL Publication 81-54, Jet Propulsion Laboratory,
Pasadena, California, July 15, 1981.

22. Zimmerman, W., "Health and Safety Evaluation of a Modified Tunnel Borer
Design for Application to Single Entry Coal Mine Development," JPL
Publication 82-12, Jet Propulsion Laboratory, Pasadena, California,
February 15, 1982.

23. Serbousek, M., Mine Safety Research, Bureau of Mines, Spokane,

Washington, private communication, 1982.

24. Hutchinson, D., Industrial Safety, MSHA, Denver, Colorado, private

communication, 1982.

25. Thaler, R., and Rosen, S. "The Value of Saving a Life: Evidence from
the Labor Market," Department of Economics, University of Rochester,
1973.

26. Marshal, R., Director, MSHA Approval and Certification Center, West
Virginia, private communication, 1982.

27. Energy Information Administration, "1981 Annual Report to Congress,

Volume 3 Supplement 2," DOE report DOE lEIA-0173(81)/3 (S-2), February,
1982.

28. Kotler., P., Marketing Decision Making, Holt, Rinehart and Winston, Inc.,
1971.

29. 1981 Keystone Coal Industry Manuel, McGraw-Hill, p. 719, 1981.

30. U.S. Department of Labor, Bureau of Labor Statistics, "Technology,

Prudu-:tivity, and Labor in the Bituminous Coal Industry, 1950-79,"
Bulletin 2072, February, 1981.

31. Mining Progress Inc., private cocm,aunication, February, 1982.

32. Aster, R. W., "Interim Price E*.i*mation Guidelines, Version One," JPL
Report 5101-33, Jet Propulsion Laboratory, Pasadena, CalifornL-.,
September, 1977.

7-2
33. Firnett, P. J., "Improved price Estimation Guidelines (IPl'04) Coaputer
Program User's Guide, Release 2 9 " JPL Report 5101-156 Rev. A, Jet
Propulsion Laboratory, Pasadena, California, November, 1980.

34. Aster, R. W., "IPEG Coefficients for Underground Cozl Mining: First
Draft," January, 1982.

35. Goldsmith, M., and Lavin, M. L., "Overall Requirements for an Advanced
Underground Coal Extraction System," JPL Publication DOE/ET -12548-1,
pages 54-56, Jet Propulsion Laboratory, Pasadena, California, October,
1980.

36. Mansfield, E., Ind—istrial Research and Technological Innovation, W. W.

Norton Co., 1968.V

3'i. Energy Information Admistration, "Bituminous Coal and Lignite Production

and Mine Operations-1978," DOE Report DOE/EIA-01/8/78, June, 1980.

38. Ltwnberger, D. G., Incroduction to Linear and Nonlinear Programming,

Addison-Wesley, 1973.

39. Kuti, J. "Longwall Mining in America," Mining Engineering, pp.

1590-1602, November, 1979.

40. Rohaly, M., Consolidation Coal Company, private communication, 1982.

41. Goode, C. A., and Gross, M. A., "Longwall Face Support in the U.S.,"
Colliery Cuardian International, pp. 9-14; October, 1978.

42. Bickerton, C. R., and Wtiterfield, M. D., "A Moving Baseline for
Evaluation of Advanced Coal Extraction Systems," JPL Publication
DOE/ET-12548-5, Jet Propulsion Laboratory, Pasadena, California, April,
1981.

43. Lavin, M. L., Border, C. S., Duda. ). R., "A Life-Cycle Description of
Underground Coal Mining," JPL Report FE/9036-1, Jet Propulsion
Laboratory, Pasadena, California, April, 1978.

44. Moody's Value Line Investment Survey, November, 1981.

45. Standard and Poor's, 1981.

46. Bielicki, R. J., and Uhrin, U, C., "Coal Mine F.gdiptrent Forecast to
1985," Eastern Field Operation Center. 1.1rCE.3 of Mi.,e Information
Circular 8710, 1976.

7-3
 APPENDIX A

DETAILED AUTOMATION DEVELOPMENT

AND PRODUCTION COSTS

A-1
 k
I
t

3'Fiia appendix covers the coat - = alffere spproacbea ^o Itf - ------ ----- ---
automation. Each
The numbers provided are general order of magnitude coats.
common equipment/subprograms that
automation opportunity envisions the use of
are shared by all five automation opportunities and the use of apecial zed
add-an equipment/subprograms for each-automation opportunity. Table k--1 list
elements that are coeaon to each automation opportunity.

Table A-1. Total Common Element Costs

Development Production -
Item Cost Cost

General purpose computer $400,000 $209000

In-seine communicati on s 120,000 -13,000
Display equipment, Printers, Plotters 120,000 109000
Modifications to Shearer
Communications 809000 11000
Distance along face sensor 20,000 1,000
System software architecture
development 120,000 1,000

Total common element costs $860,000 $489000

Table A-2 is developed assuming that only the vertical profile of the
shearer cut is placed in machine memory and utilized on the subsequent shearer
cuts.

Table A-2. Add-on Ranging A::a Articulation Costs

Development Production
Item Cost Cost

Shearer ranging arm modification $100,000 $2,000

Ranging Arm Software 240,000 1,000

Add-on development costs $340,000 $3,000

PRECEDWG PME BUWK NOT A.tAED

C -^

A-3
 Table A-4. Combine: Costs of Ranging Are and Cowl articulation

Development Production

Item Cost Cost

Common costs (Table A-1) $ 860,000 $48,000

Ranging arm articulation costs
(Table A-2) 340,000 3,000
Cowl articulation costs (Table A-3) 160,000 3,000

Total combined costs $1,360,000 $549000

In Section III, the foilowing automation opportunities constitute a

single face alignment task:

(1) Smart shearer face alignment sensing package.

(2) Face conveyor alignment computation package.

(3) Face conveyor alignment hydraulic ram servo control package.

(4) Microprocessor controlled shield advance control package.

A-4


OMQ AL PAM IS
OF Pmt OUALfff

Each automationtrait y ^ suceessfel c letioa of the prior

r requires --
opportunities in order to function. Partial operation night be possible, s
as oicroprocessor-controlled ahis^ ahal # bobeftr, sash an.oparatiou is 46t
believed to be cost effective by itself. Table, A-s lists the 41Mroe 6"
needed for the smart rhearer face alitt ssi, packsge. this paekaga is
bated on the shower and night be iucluded in the total sbearer velapiaent;
costa of Table A-4*

E Table A:S. Add-on face_Ali& t Measurement-Casts

s -

F Development Production
,I Item Cost Cost

Shearer face alignment sensor $200,000 $209000

Face conveyor alignment sensing software 120,000 11000

Add-ca development costs $3209000 $219000

After the alignment of the face is meas •.red, these measurements most be
converted into commands to individual conveyor rams. Initially this
computation is done at the general purpose computer_ on the surface. In
productions versions, the computions would be done by a dedicated computer at
the headgate. The computations are used by the face conveyor alignment
hydraulic ran servo control package. Table A-6 shows the cost for face
alignment control. The work of Table A-5 must be completed before Table A-6
work can proceed.

Table A-6. Add -on Face AiignMoUt Control Costs

Development Production

Item Cost Cost

Modifications to Sheild
Communications
let unit $ 509000
100 units a 1,000 each 400,000 $1009000
Conveyor alignment ram
let unit - 50 1, 000 -----
100 units 0 1 9 000 each 400 9 000 1009000
Face conveyor alignment computation software 60 9 000 109000
Face conveyor ram servo control software 120,000 5,000

Add-on development costs $191009000 $2159000

i

A-5
 Face alignamot control includes commmication between mite a
each shield, and a microprocessor eat *aW shield. Sbield advanew ntili. es
face aligmwnt croprocessor an each smiaid to 10ont"I d47 shield tilt VIM
and the top rams. This will result in autt ted shield afivawai This.
capability is limited to good floor co6dittima t thus a tilt imiicater and M-
obstacle detector are included to detect malfunctions and a MarMIAMlit to
warn personnnel of ping t and/or to m w persounel if op ttwo
are faulty. Table A-7 covers the add-on cost of_microprocessot ontrolled
shield advance.

Control of Sheild top ram and tilt ran

Electric solenoid valves
100 units F 500 each
Warning lights
100 units E 100 each
Emergency stop obstacle detectors
100 units F 500 each
Shield advance software development

Add-on development costs

The total add-on cost for face alignment and shield advance is shown in
Table A-8.

Add-on Face Alignment/Shield Advance Costs

Development
Cost

Face alignment measurement (Table A-5)

Face alignment control (Table A-6)
Shield advance (Table A-7)

Add-on development costs

 Fault isolation
(pumps, transformers, etc.)
Modification 50 units @ 1,000 each 2009000 509000
Fault isolation software development 120,000 30,000

Add-vu development costs $3609000 $629000

A computer monitoring system would be a subset of sensors of the fault

isolation system. If control was also included, additional sensors and
controls would be needed. Assuming that one wished to vary the shearer speed
as a function of the pan-line load, the items of Table A-10 would be add-on
items. The addition of other types of control would cost about the same order
of magnitude.

Table A-10. Add-on Monitoring/Control System Costs

Develo-mlent Production
Item Cost Cost

Shearer tram motor control $ 4 9 000 $11000

Headgate drive current detector 49000 19000
Control system software 120,000 1,000

Add-on development costs $1289000 $3,000

A-7
 APPENDIX B

DETAILED DESCRIPTION OF THE IPEG

(IMPROVED PRICE ESTIMATION GUIDELINES)
MODEL -
bf
IPEG stands for itproved Pricestiaatia 6eti^as. IL_onsiats `of a
linear equation which provides averhead for equipment VA property
--
(A ) 9 direct labor (DRAB), and materials, supplies, AM utilities (NAM.
The coefficient for equipment overhead has different values for different
equipsitut lifetimes. Th=s equation car be used fee an entire sine or for a
portiou of a nine:

Two IPEG equations are p uvided here; one for a 1519I-rate.of return on - -
equity (a typical average for most sines at this tree), - - mr-a 25Z ra#=e
of return on equity. The 15% rate, and many of the . otherfinancial parameters
used to generate these equations, are consistent with the assumptions use in
which describe a 1978 coal mine (44). A test case yielded a result within 6%
of the result obtained in that report. -IPEG produted a result of $18.7/ton
and the more elaborate procedure gave a result of $17.6/ton when the same
values for EQPT, ACRE, 'DLAB, and BATS were used.

The two equations below represent the 152 and 25Z rates of return on --
y, res pe ctivel y, with a 10-year equipment ' life. The " ACRE coefficient
e quit
will be sensitive to land costs, site prep, and shaft development costs. The
other coefficients will be reasonably accurate for soot sines. A full list of
financial parameters used to generate these equations is attached.

(1) 0.330 x ESE + 1497 x ACRE + 2.16 x DLAB + 1.18 x MATS

QUAN

(2) 0.558 x EQPT + 3208 x ACRE + 2.33 x DLAB + 1.23 x MATS

QUAN

The financial and organizational overheads were derived from the values of
parameters given in Table B-3.

The improved Price Estimation Guidelines ( IPEG) were first derived in

1977 for photovoltaics ( 32). A computer program now exists to create new
models based on the financial and operational data for any particular plant or
industry ( 33). IPEG has row been extended to cover underground mining.

The intent of IPEG is to minimize the computational burden of

engineering economic studies without compromising the validity of the
analysis. This facilitates optimization and economic tradeoff studies. A
second intent is to permit valid comparisons between the true costs of
capital, labor, materials, and supplies. Each coefficient is a realistic
representation of the overhead burden that should be attributed to its
associated engineering parameter. This is a feature which is not always
available from engineering economic models.

The input parameters are: equipment ( EQPT), mine property size ( ACRE),
direct labor ( DLAB), materials, supplies, and utilities ( MATS), and the
quantity of coal produced in a year ( QUAN). These have specific definitions:

PRECEDING PAGE BLANK NOT FILAIM

B-3
 4. MATS is the annual cost of materials, supplies, and utilities
expressed in 199 2 ;ollars.

Rebuilding equipment (e.g, continuous miners are rebuilt every two to

three years at approximately 54Z of their original cost) can either be
capitalized (i.e., included as as equipment.ters) or expensed (i.e., included
in the annual materials and supplies term). Expensing will lead to lower coal
prices and, from an accounting viewpoint, is a more proper procedure. In
either case, comparisons of alternative technology should be done on a
consistent basis.

Financial and operational costs of doing business are included in the

overhead computed by IPEG. This includes: profit, income taxes, amortization
of investment, insurance, startup costs r ,id revenues, working capital,
indirect labor, fringe benefits, tax credits, royalty payments, and other
miscellaneous expenses. The average annual revenue required to meet all
cirect and indirect expenses is divided by the . averag,3 annual production
quantity (QUAIL). The annual revenue required is obtained by multiplying each
of the engineering parameters by an appropriate coefficient. The coefficient
for EQPT will change for different equipment lifetimes. In general, these
coefficients will change if a major financial variable such as the rate of
return on equity is modified. Table B-1 gives a set of coefficients based on
a 15% after-tax rate of return on equity, which is a typical coalmine
industry average. Table B-2 gives a set of coefficients based on a 25%
after-tax rate of return on equity which would be an appropriate rate of
return for relatively innovative, unproven investments. A complete list of
input data is given in Table B-3.

*The results of this.version of IPEG will be in 1982 dollars. If some other

year's dollars are desired, then EQPT, DLAB, and MATS should be expressed in
that year's dollars and the ACRE coefficient should be adjusted by the infla-
tion rate.

B-4
 Tabu 8=1. IM Coal Miss Coefficients# 151

Equipowt Fors. EQPT ACfB DLAE MATS

Lifetime

1 1535 1497 2.16 1.18

2 0.760

3 0.578

4 0.487

5 0.433

6 0.398

7 0.373

8 0.355

9 0.341

10 0.330

12 0.315

15 0.301

20 0.289
 Toble 3-2. MC-Coal Kim Coefficient•#

Equipment Far: EQPT ACRE MAD PA

Lifetime

1 1.869 3208 2.33 1.23'

2 1.014
3 0.610
4 0.711
5 0.655

6 0.619

7 0.595

8 0.578

9 0.567

10 0.558

12 0.549

15 0.548

20 0.547

B-6
FL Facility and mine lifetime, in rears 20

BETA Property tax rate, annual percent

of capital investment 0

EITCR Average equipment investment

tax credit rate 0.10

TC Construction lead tine (site prep,

surface buildings, shaft construction)
in years 2

TS Development and startup time, in years 1

L Average fraction of steady-state

production rate achieved during
development ','.A) 0.77

N Startup MATS usage fraction relative to

normal annual cost (44) 0.47

TM First year of full production

(no influence on coefficients) im

NU Insurance rate 0.01

RLAB Indirect/Direct labor cost, includes

402 benefits, 252 indirect labor
(plant visits) (44) 0.75

RUTIL Indirect utilities cost/acre for

ventilation, vater pumping in
$/acre/year (44) 28.6

B-7
c

--
ale 3.

Default
Variable Description dal

- i - Vorking capital tiwe lag, inyearn 0."

TAU Tneow tax rate 0.5
LAMM Leverage, expressed as total capital/
equity capital (default is 1000 equity
financing) 1

Interest rate on 4ebt, bands = -

(not used here) 0.14

R gate of return on equity . (44 9 45, 46). _-

A higher rate of return uy be needed to
introduce new technology into the
industry. 0.15

P1 Price of land $/acre 29500

P2 Price of buildings $/acre 187,500

D2 Building area/mine area 0.002

G Inflation rate (not influential to

coe+ficients) 0.09

XEC Contingency during construccion

(fraction of EQPT) 0.15

XFC Contingency during construction

(fraction of facilities cost) 0.50

X Miscellaneous expense fraction of

revenue (-0.10 federal tax credit, + 0 .05
royalties, + 0.02 local taxes) -0.03

XOPR Miscellaneous expense fraction of

operating costs (44) 0.15

CRT Capital recovery time, in years 20

S-8
C-1
s

-- - - - -- Historical data on loll market share in the coal

industry is summarized in Table C-1. In addition tothis information, a
histogram of longwall startups from 1965 to 1981.was d loped and is included
_ -here as Figure C-1. This information is consistent with traditional market
penetration curves ("S" shaped curves). This aendiz describes the detailed
procedure used to develop the particular longwall growth curve that was need
in this study.

The "S" shaped ;curve that is used in market growth studies can be
generated by a logistics function (28). A general formulation of the
logistics curve is the following:

Market Share =
E1 + R2
1+e A +BzT+CxTz?
where: T is time expressed in years,

R1 + K2 is the upper bawd on market share,

R19 R2 , A, B, and C are parameters which

must be estimated from data or reliable theory.

The first step was to use a simple fora of this equation by setting X1
and C to zero and then seeing if the resulting formula provided a good fit to
thn_ data. This simplification s<:smed appropriate because there- are only 13
historical data points available to estimate the model parameters.

The second step required deriving a reliable estimate of an upper bound

on longwall market s'tare. Even if all %oal mines eventually used longwalls,
there would be tunnel boring for shaft and panel development which coul4
produce roughly 20% of the coal that is extracted. However, many mines
produce less than 200,000 tons/yr, which is too small for supporting a modern
longwall section. In fact, 37Z of the 1978 production came from mines of this
size or smaller, and this percentage has not changed significantly since 1962
(30 9 37). This sets an upper bound on longwall's underground market share at
50%, and establishes a value of 0.5 for K2.

The next step required nonlinear programming techniques. Using 1995 as

the point where T - 0, selecting a value for the market share at T - 0
determines the value of A (because R 2 - 0.5 and the other terms drop out).

CEDING PAGE BLANK KOT FILMED

c-3
aEstimated from the longwall census (31).

C-4

71

17

16
z0
13
15
W Z
0W
Z
Ul
14 0 I'L

O
13

12 4 z -
0
11
ac

10 0
z Q
MA O
01 9 C w
zu
U.
0 i
w 8 0^
Ous
LU ZIE

O 7 R
z
z
6 R
Uj
5 CL
0 R R

4 R R R

3 R R cc

R
z
0-
R R fo
I I
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
YEAR

NOTE: There are 103 operating as of February 1982, 26 are on orderp and 145
have been operated.
Figure C-1. Longwall Startups in the USA

C-5
 0.375

1
0
I--

0.325

W
0
Z

0.300

0.275

0.250

20 ?2 24 26 28 30
PERCENTAGE OF UNDERGROUND COAL MINING
PERFORMED BY LONGWALLS IN 1995 (%)
Figure C-2. Market Curve Fit Variance as a Function
of Market Penetration by 1995

C-6
 Market Share - 0.5
1 + e -0.127
Where ? - 0 in 1995.

@SSIFICATION

The results of the market share curve mere compared to projections by

Ruti who used a different method in 1979. Euti's projection for 1985 is 122;
this study's projection for that year is 11x. A less formal verification cane
from examination of data received from industry sources (31 9 40). Current and
planned longuall expansions seem to be consistent with the above projection at
least to 1987.

C-7