Mantenimiento de Bomba Horizontal de Lodos Warman

University of Southern Queensland
Faculty of Engineering and Surveying
Pump Gland Improvement Project
A dissertation submitted by
Damon Thomas Bishop
In fulfilment of the requirements of
Courses ENG4111 and 4112 Research Project

Towards the degree of
Bachelor of Engineering (Mechanical)
Submitted: 28th of October, 2010

ENG4111 & ENG4112 Research Project
Limitations of Use
The council of the University of Southern Queensland, its Faculty of Engineering and Surveying
and the staff of the University of Southern Queensland do not accept any responsibility for the
truth, accuracy or completeness of material contained within or associated with this dissertation.
Persons using all or any part of the material do so at their own risk, and not at the risk of the
Council of the University of Southern Queensland, its Faculty of Engineering and Surveying or
the staff of the University of Southern Queensland.
This dissertation reports an education exercise and has no purpose or validity beyond this
exercise. The sole purpose of the course pair entitled Research Project is to contribute to the
overall education within the students chosen degree program. This document, the associated
hardware, software, drawings and other material set out in the associated appendices should not
be used for any other purpose: if they are so used, it is entirely at the risk of the user.
Prof Frank Bullen

Dean
CERTIFICATION
I certify that the ideas, designs and experimental work, results, analysis and
conclusions set out in this dissertation are entirely my own effort, except where
otherwise indicated and acknowledged.
I further certify that the work is original and has not been previously submitted for
assessment in any other course or institution, except where specifically stated.
Damon Thomas Bishop

Student Number: 019222337
______________________________
Signature
______________________________
Date
ABSTRACT
The need for this research project has arisen principally due to the economic cost incurred from
poor reliability of a pump group within the Rio Tinto Alcan Gove Alumina refinery. This poor
reliability issue concerns the glands on a pump group whose reliable operation is critical to the
success of the business. This research project has identified the root cause of the problem and in
turn provided a solution.
The Alumina industry relies heavily on slurry pumps to transport liquids and solids throughout the
refining process. The failure of these pumps can significantly impact on a refineries ability to
produce and consequently the reliability of the equipment is a major focus. Slurry pumps often
utilise tradition stuffing box style glands to seal the wet end of the pump from the atmospheric and
are usually the first part of a pump to fail. At the Rio Tinto Alcan Gove alumina refinery the mill
injection pump group have been experiencing repetitive gland failures which have been costing the
operation approx $12,000 per day. It is these pumps that the project has concentrated on.
The critical objectives of this research project include
Identification of the reliability project
Identification of the root cause
Proposal of a solution to mitigate the effect of the root cause.
Implementation of the solution.
To achieve these objectives several methodologies have been employed.
Comprehensive review of production and maintenance cost associated with these pumps.
Literary review to examine past works involving pump gland on site and globally.
Detailed failure investigation.
Conducting of an Apollo root cause analysis to identify the root cause and provide the best
solution to mitigate it.
Implementation and validation of the solution.
With all objectives achieved the completed project has obtained it desired results. Since the
solution has been implemented there has been a significant reduction in gland failures with no
recorded production losses. The time and capital expedited on this project has returned many
times its value.
ACKNOWLEDGEMENTS
Appreciation is given to the following groups for information and materials used in this
project.
Rio Tinto Alcan Gove (RTAG)

Weir Warman Pumps
KSB Pumps Australia
Chesterton Mechanical Seals
TABLE OF CONTENTS
Page
Certification ..... 3
Abstract .. 4
Acknowledgements ...
List of Figures ...
10
List of Tables and Charts ................
11
List of Appendices ................. 13

Nomenclature... 14
Chapter 1 INTRODUCTION ..
15
1.1 Outline of the Study.......... 15

1.2 Introduction..
15
1.3 The Problem..
15
1.4 Research Objectives..
16
1.5 Conclusions: Chapter 1..
16
Chapter 2 SAFETY REVIEW..
18
2.1 Safety Introduction...
18
2.2 Identify Hazards..
19
2.3 Identifying Risks...
19
2.4 Controlling the Risks.
19
Chapter 3 MILL INJECTION PUMP OVERVEIW..
21
3.1 Aims of the Chapter...
21
3.2 Introduction...
21
3.2.1 Location.
21
3.2.3 Purpose of Refinery...
21
3.2.3 Bayer Process..
22
3.3 Identification of a Reliability Project..... 24

3.4 Mill Injection Group... 27
3.4.1 Expansion of Refinery....
28
3.4.2 Pump Makes...
28
Chapter 4 LITERATURE REVIEW...
31
4.1 Aims of the Chapter.... 31

4.2 Introduction to Centrifugal Pumps. 31
4.3 Introduction to Slurry Pumps. 33
4.4 Gland Configuration and Performance in Slurry Pumps...
35
4.5 Gland Failure Characteristics. 36

4.5.1
Worn Shaft Sleeves..
36
4.5.2
Poor Packing Material Selection.....
37
4.5.3
Incorrect Gland Water Supply....
38
4.5.4
Incorrect Packing Installation.
38
4.5.5
Wrong Pump Application...
39
4.5.6
Mechanical Fault
39
4.6 Industry Methods in Improving Slurry Pump Gland Reliability...
40
4.6.1
Grooved Flow Restrictor Bushes
40
4.6.2
Live Loading on the Glands ..
40
4.6.3
Grease Purging of the Glands..
41
4.6.4
Stuffing Box Bearings
41
4.6.5
Slurry Packing
41
4.6.6
Mechanical Seals....
41
4.6.7
Stiffened Shaft Assemblies...
42
4.7 Previous Projects Concerning the Mill Injection Pumps......
42
4.7.1
Gland Water Differential Project.......
42
4.7.2
Installation of a Grooved Flow Restrictor..
43
4.7.3
Repack MST...
43
4.7.4
Mechanical Seal Installation......
43
4.7.5
Slurry Packing Trial
44
4.8 Conclusion
44
Chapter 5 DATA COLLECTION and ANALYSIS...
45
5.1 Aims of the Chapter.......
45
5.2 Data Required....
45
5.3 Data....
47
5.3.1
Suitability of Shaft Sleeve.......
48
5.3.2
Suitability of Packing Selection ..
48
5.3.3
Suitability of Gland Water Supply... 49
5.3.4
Suitability of Pump To Application....
50
5.3.5
Mechanical Inspection on Gland Failure ..
52
5.4 Conclusions..
52
Chapter 6 APOLLO ROOT CAUSE ANALYSIS..
53
6.1 Aims of the Chapter...
53
6.2 Incident Report..
53
6.3 Reality Chart..
54
6.4 Solution Selection
54
6.5 Solution Description....
56
6.6 Conclusion...
58
Chapter 7 SOLUTION IMPLIMENTATION AND VALIDATION 59

7.1 Aims of the Chapter.... 59
7.2 Mill Injection Shaft Upgrade Cost. 59
7.3 Acceptance From Refinery Management 60
7.4 Plan for Installation..... 60
7.5 Validation of the Project Six Months After Implementation.. 62
Chapter 8 CONCLUSION
1
64
Review of Project......
64
8.1.1
.. 64
8.1.2
.. 65
10
8.2 Further Work to Complete Project.. 65

8.3 Further Benefits and Opportunities. 66
8.4 Closing.... 67
REFERENCES..
68
APPENDICES..... 69
11
LIST OF FIGURES
Figures
Number
Title
Page
3.1
Northern Map of Northern Territory.. 21
3.2
Arial View of Rio Tinto Gove Alumina Refinery....
3.3
Process Diagram of the Bayer Process... 23
3.4
Mill Injection Group Schematic....
28
3.5
Mill Injection Group Pumps..
29
3.6
D Group Pumps.
30
4.1
Pump Dynamics.
32
4.2
Flow Curve..... 33
4.3
Slurry Pump Parts List..
34
4.4
Stuffing Box Configuration...
36
5.1
Shaft Sleeve Condition after 10 days operation.. 47
5.2
Shaft Sleeve Condition after 77 days operation.. 47
5.3
Packing after 5 days run time . 48
5.3
Packing After 21 days run time.. 48
5.5
Gland Water Pressure..... 49
6.1
Existing Shaft Assembly..... 56
6.2
Stiffened Shaft Assembly 56
7.1
New Assembly in Box... .. 61
7.2
Installed Assembly in P633-4B... 61
22
12
LIST OF TABLES and CHARTS

Tables
Number
Title
Page
1.1
Time Line for Completing Research Project Milestones. 17
2.1
Likelihood / Consequence Matrix...
18
2.2
Risk Control Table..
19
2.3
Hierarchy of Control...
20
3.1
Exert from Production Loss Report.
24
5.1
Data Collection Plan.. 46
6.1
Incident Report...... 53
6.2
RCA Team. 54
6.3
Summary of Root Causes.. 54
6.4
Solution Generated From RCA. 55
6.5
Solution Assessment.
6.6
KSB Pump Deflection Calculation.... 57
7.1
Cost For Project. 59
7.2
NPV... 60
7.3
Plan For Installation.
55
62
13
Charts
Number
Title
Page
3.1
Total Pump Losses . 25
7.2
Mill Injection Pump losses by Failure Mode.. 25
3.3
MTBF for selected Refinery Pump Groups.... 26
3.4
Repack Events on the Mill Injection Group ... 26
4.1
Mill Injection MTBF... 42
5.1
Flow Curve for 6-8 LSA-25 ... 50
5.2
Stage 1 and 2 Mill Injection Flows for a 1 Month Period... 51
5.3
Calculated pump Shaft Deflection for the 1 Month Period. 51
6.1
Calculated Shaft Deflection With Stiffened Assembly ...... 57
7.1
Comparison of MTBF for P633-4B Before and After Shaft Upgrade.... 63
8.1
Mill Injection Utilisation 64
14
LIST OF APPENDICES
Number Title
Page
Project Specification.
69
Industry Products to improve Gland Reliability.... 70
Previous Improvement Projects..... 73
Gland Water Requirements for Slurry Pumps... 76
Shaft Deflection Calculations.... 77
Mechanical Inspection Reports..... 80
Reality Chart Summaries... 84
Letter of Offer concerning Stiffened Shaft Assemblies.... 91
Power Point Presentation to Management 94
Capital Project Forms for Mill Injection Shaft Upgrade... 98
15
NOMENCLATURE
Reliability- the probability that a device will
Scat A worn segment of Mill Rod or Ball.
perform its intended function during a
NPSHr Net positive suction head
specified period of time under stated
requirements for a specific pump to run with
conditions.
adequate suction pressure and avoid
Utilisation - Utilisation is a measure of the
cavitations.
time that the plant is utilized, albeit
BEP The best efficiency point (BEP) is
productively or non-productively.
the flow rate where a pump has its highest
(Operating Time / Calendar Time)
efficiency.
Availability - Availability is a measure of
Head Liquid force measured by elevation.
the time that plant and equipment is
P.P.E Personnel Protection Equipment
available to perform its intended function
EHS Environment, Health and Safety
over a defined period. (Available Time /
MST Maintenance Scheduled Task
Calendar Time)
NPV - Net Present Value
MTBF - is used to quantify the reliability in

operation of equipment. (Utilised Time /
Number of Failures)
SFL Strong Feed Liquor
16
CHAPTER 1
INTRODUCTION
1.1
Outline of the Study
The need for this research project has arisen principally due to the economic cost incurred from
poor reliability of a pump group within the Rio Tinto Alcan Gove Alumina refinery. This poor
reliability issue concerns the glands on a pump group whose reliable operation is critical to the
success of the business. Previous projects which have been carried out with the aim of improving
the gland reliability have had little success. Additionally there is often frequent discussion within
the Alumina industry concerning pump gland reliability and the lack of resources that
specifically target slurry pumps glands as opposed to clean fluid pumps. The intention and
scope of this project is laid out in 1.4 Research Objectives.
1.2
Introduction
Whilst the field of Reliability engineering is not new there has been a relatively recent escalation
on its importance which has been born out of the need for industry to maximise their return on
their investment in their equipment. Equipment failure can impact a business in many ways such
as production loss cost, maintenance costs and environment, health and safety and therefore the
solution to preventing these failures can yield valuable proceeds. To find the optimum solution
the root cause will need to be understood.
1.3
The Problem
The Alumina industry relies heavily on slurry pumps to transport liquids and solids throughout
the refining process. The failure of these pumps can significantly impact on a refineries ability to
produce and consequently the reliability of the equipment is a major focus. Slurry pumps often
utilise tradition stuffing box style glands to seal the wet end of the pump from the atmospheric.
Pump glands often are usually the first part of a pump to fail and often are merely symptom of a
bigger issue within the pump or the process of which the pump is subjected. This research
project will examine a gland failure issue and attempt to produce a solution which will in turn
positively impact the reliability of the pump group.
17
1.4
Research Objectives
The critical objective of this research project is to identify a reliability project, isolate the root
cause and propose a solution to mitigate the effect of the root cause.
The format of this project will be closely aligned with the project specification (Appendix A). To
ensure that the project meets all poignant completion dates a plan has been step up so as to track
the progress (Table 1.1).
1.5
Conclusions: Chapter 1
The results of this study is expected to demonstrate a problem solving methodology which
enables the reliability engineer to solve a problem with an unknown solution.
The background and literature review will consist of a review of Slurry pumps, principally
focusing on the gland composition, failure modes and performance and provide background in
the equipment reliability issue which the project is involved. This will assist in the identification
of the root cause and also a solution to the problem.
18
Active
Due
Inactive
Start of Week
Project Proposal Submission
Read Project Reference Book
Project Specification Submission
Time Line for completion of Research Project milestones (Mill injection gland reliability project)
Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10 Week 11 Week 12 Week 13 Week 14 Week 15 Week 16 Week 17
28-Feb
7-Mar
14-Mar
21-Mar
28-Mar
4-Apr
11-Apr
18-Apr
25-Apr
2-May
9-May 16-May 23-May 30-May
6-Jun
13-Jun
20-Jun
Due
Develop timeline to Complete Project

Build Document Shell
Research slurry pump sealing in industry (preferably
the Alumina industry) with respect to designs,
performance and reliability.
Investigate previous improvement projects which have
been previously performed on this pump group
Collect and analysis the data concerning the cost to
the business that have resulted in the mill injection
pumps poor gland performance.
Develop an evaluation plan for the collection and
monitoring of historical and current data concerning
the gland performance on the mill injection pumps.
Due
Project Appreciation Submission

Analysis the field data for the purpose of isolating the
root cause or causes.
Commence developing potential solutions
Start of Week
Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10 Week 11 Week 12 Week 13 Week 14 Week 15 Week 16 Week 17
19-Jul
26-Jul
2-Aug
9-Aug 16-Aug 23-Aug 30-Aug
6-Sep 13-Sep 20-Sep 27-Sep
4-Oct
11-Oct
18-Oct
25-Oct
1-Nov
8-Nov
Read Project Reference Book

Develop timeline to Complete Project
Perform Root cause Analysis (Write up report)
Submitt Extended Abstract and Work

Experience
Due
Complete Mathamatical analysis to validate solution

(Write up report)
Obtain costing reports to impliment solution (Write up
report)
Present solution to management (Write up report)
Impliment solution (Write up report)
Validate solution (Write up report)
Due
Draft Submission Submitted

Submitt Complete Project
Due
(Table 1.1)
CHAPTER 2
SAFETY REVIEW
2.1
Safety Introduction
An Alumina refinery is an inherently a dangerous place, the caustic which is used to dissolve the
alumina also reacts with human tissue essentially dissolving it. The product is pumped at
temperatures of over 200c and pressures of over 4000kPa. This makes any incident where the
exiting of the product from the system into the atmosphere potentially a dangerous one. The
process that is required to be followed in order to manage these risks in the refinery which is
found in the EHS procedural manual is essentially simplified as
Identifying the Hazard
Identifying the Risk
Placing controls to manage the risk
The Likelihood / Consequence matrix (Table 2.1) is a widely used tool which enables risk to be
categorised with the intent of determining the level of control required.
Likelihood
Consequence
1 - Minor
2 - Medium
3 - Serious
4 - Major
5Catastrophic
A - Almost
Certain
Moderate
High
Critical
Critical
Critical
B - Likely
Moderate
High
High
Critical
Critical
C - Possible
Low
Moderate
High
Critical
Critical
D - Unlikely
Low
Low
Moderate
High*
Critical
E - Rare
Low
Low
Moderate
High*
High*
Table 2.1. (Rio Tinto Alcan Gove procedural Manual GPM-EHS-001)
2.2
Identifying Hazards
Pump gland failures often result in the product exiting the stuffing box region of the pump under
high pressure. This presents a hazard particularly when the product is caustic slurry at 105 c. In
the event of a human to be in the vicinity of a sudden gland failure a realistic consequence would
range from a minor chemical / thermal burn to a serious chemical / thermal burn which would
result in a loss time injury. This identifies a hazard with a serious consequence if nothing is done
to control it.
2.3
Identifying risks
To quantify the hazard so appropriate controls can be in place the Likelihood / Consequence
matrix is used (Table 2.1) this compares the severity of the hazard with likelihood of it occurring.
Fortunately glands typically fail gradually therefore while the hazard is potentially serious the
likelihood of it occurring would be unlikely. Therefore the risk can be classified as moderate.
2.4
Controlling the Risk
To manage a moderate risk Table 2.2 is referred to. Table 2.2 suggests that the risk can be
managed through control which routine monitoring of the effectiveness of these controls and
periodic review.
Residual Risk
Rating
Critical
Tolerability
Action Message
Intolerable
Significant and/or urgent action is required to eliminate or reduce the risk level.
Area Superintendents must allocate resources and review the risks. Consider
Quantitative risk assessment.
High
Intolerable
Develop a Risk Reduction Plan to eliminate or reduce the risk to a tolerable level.
Consider Hierarchy of Controls and ALARP (as low as reasonably practicable)
principles.
Tolerable
Maintain controls, and ensure risks are periodically monitored and reviewed.
Moderate or Low
21
In Table 2.3 the preferred selection of controls process is illustrated. This enables the most
optimum and practical control to be put in place.
Hierarchy of controls
Control Flow
Hierarchy
Step 1
Eliminate
Step 2
Substitution
Replacing one substance or activity with a less hazardous one.
Step 3
Engineering
Physical risk control methods complying with legal requirements,

preventing any contact with power source or the situation and which can
be changed by human activity. They reduce or limit the exposure
Step 4
Administrative &
Training
Approved working methods, procedures, inspection, observations, and

regulations. Examples: JSEAs, clearance procedure, confined space
procedure, coaching, mentoring, competency based training. .
Step 5
Control Definitions
Physical risk control methods complying with legal requirements,
preventing any contact with power source or the situation and which
cannot be changed by human activity.
Signage, barricading, pedestrian crossing, definition of work area,
Protective Devices posters, PPE etc

In the case of the Rio Tinto Alcan Gove refinery the following protocol is to be followed when
working in and around online pumps.
Perform relevant inductions Visitor is aware of potential dangers and what to do if they
are encounter. I.E. Safety shower location, Radio channels, ECT.
Seek permission from operation before entering the area This will enable
communication of any hazard so the visitor is aware.
Perform a personnel hazard assessment upon entering the area to ensure all is safe.
Gland guards are to be in place This will divert any product which exits the stuffing box
towards the ground.
PPE to be worn In the case of an incident some protection will be given.
Area to be barricaded in the event of a gland failing.
With these controls in place the risk is now understood to be managed and allow safe access for
workers.
22
CHAPTER 3
MILL INJECTION PUMP OVERVEIW
3.1
Aims of the Chapter
The specific aims of this chapter are to:
Provide an introduction to the Rio Tinto Gove Alumina refinery, its location, function and
the Bayer process.
Provide the justification for the reliability issue concerning this research project.
Explain the purpose and composition of the Mill injection pump group
3.2
3.2.1.
Introduction
Location
The Rio Tinto Alcan Gove Bauxite Mine Alumina Refinery is situated on the Gove Peninsular in
the N.E region of Arnhem Land in the Northern Territory next to the purpose built town of
Nhulunbuy.
Figure 3.1. (http://www.sailing-story.com/images/darwin-nhulunbuy-map.jpg, 17/05/2010)

3.2.2.
Refinery Function
The refinery which was first commissioned in 1972 utilises locally mined bauxite to produce
Alumina. The refinery is currently producing 3 million tonnes of Alumina per annum which
makes it the 5th largest Alumina refinery in the world
23
(http://en.wikipedia.org/wiki/List_of_alumina_refineries). The refinery extracts the Alumina

from the Bauxite using the Bayer Process.
Figure 3.2. (http://www.riotintoalcan.com/ENG/resources/image_library_1456.asp,

17/05/2010)
3.2.3.
Bayer Process
At the Gove refinery the Bayer process (See Pictured) achieves it purpose through the following
process.
Bauxite is strip mined (6.8 million tonnes annually), crushed to <25mm and loaded on a
system of conveyors which deliver the ore to the refinery stockpiles 18km away.
The Bauxite is reclaimed and delivered into the mills where the dry bauxite is blended
with a Caustic solution then ground to a fine paste. It is here that the extraction of the Alumina
from the bauxite begins as the caustic dissolves the alumina.
The caustic slurry is then feed through a series of pressure vessels and heat exchangers to
allow the maximum absorbsion of alumina from the bauxite.
The slurry is then separated into two streams where the mud solids are sent out to the
residue disposal and the alumina rich liquor is feed through filtration and into the white side
for crystal growth.
24
In the white side the liquor is allowed to cool and then passed through a series of filters
so as to extract the hydrated (wet) alumina.
The hydrate is then conveyed to calcinations where the hydrate is passed through fluid
bed calciners. At +1000c the kilns remove all bonded water to produce a dry alumina
ready for export.
Figure 3.3. (http://www.redmud.org/images/0.C2.gif, 15/05/2010)

To perform the Bayer process the refinery relies on large volumes of caustic slurry and liquor to
be transported through an extensive and complex piping system. To achieve this, the refinery has
dedicated approximately 400 of its approximate 1000 pumps to pump product in various forms
throughout its piping network.
Due to the dependence of the process on slurry pump performance poor performance can have a
significant impact on the output of the refinery. When an event occurs though poor performance
it is measured as a production loss.
25
3.3
Identification of a Reliability Project
Production targets for the refinery are forecast on a daily basis. When the refinery fails to
achieve the targeted tonnage rates the events which led to the short comings are tabled and
published as the daily production loss reports. These reports identify the production loss event.
Table 3.1 displays an exert of an event from the Production Loss report.
Start
1-Apr10
End
1Apr10
Event
Area
Category
100401
Dig 2 Injection
Group
P633-4B
Gland
Failure
633Mills/MSH
MechBreakdown
Equipment
Root Cause
Event
Category
P633-4B
Construction
Quality
Control
Failure Gland
Actual
Variation
-416
Table 3.1 Exert from Production Loss report

These losses are reviewed weekly by the Reliability engineering team which uses the data to
highlight the areas where the equipment reliability is performing poorly. Projects are then
developed using this data. Pumps which make up the lions share of the rotating equipment on
site can and often do feature heavily in the production reports and hence numerous projects
concerning pumps are developed for the Pump reliability engineer to work on.
One pump group in particular has continued to dominate the production loss reports and on
reference to Chart 3.1 it is evident that the mill injection group has been the number one
contributor to production losses for pumps in the time period measured. These losses total 34,500
tonnes which equates to $10,350,000 @ $300 per tonne of alumina (2010-04-16 spot price of
Alumina from LME) or $12,621 per day.
26
In
je
c
eh tion
qu o P G
or
um rou
M Inje p g p
i ll
c
Sl tio rou
p
ur n
St
ry Gr
g
ou
D
S
1
& tg i sc p
3
h
2
IS IS arg
C
C
e
St
M
Sl Pu
g
ud u r
m
1
ps
In ry
&
P
je
2
IS c tio um
ps
C
n
W
G
D
e Mi ll ate rou
p
W
r
a Slu Pu
H teri rr y mp
TD ng
H
s
D P u ea t
il u
in
m
LT
g
te
p
G
D
S
ro
P
S lu
C ure FL rry u p
C oar Co Fe Gro
la
ed
n
u
ri f se
G p
S de
ie
r U ee nsa ro
u
d
t
nd
p
e
R
Ev erfl etu Gro
up
ap ow rn
P
P
C
on um ump
d
p
s
SF y P Gr
In
L um oup
st
B
p
ru
o
m ost Gro
en e
u
Tu t P r G p
rb u rg rou
id
p
e
St
Li
g Stg
qu Wa
1
t
an 3 C HR or er
d
on D U gro
2
up
n
P dy
ro
R de
LT
ce et rflo
u
s
D
w
r
s
Pr
C nG
oc
r
D o
es CW nd y oup
Pr
s
es
G
C
su
on Und rou
re
de er p
D
fl o
n
H
s
TD ec
a TC at ws
Pr nte A e G
P
ro
Pr oce r O u
es
s s v e mp up
Co rflo gr
C sur
ou
yc
e
D nd w
p
lo
ne ec en Gro
s
a
u
a
U
nd nter te p
G
er
flo Fee rou
w
p
d
G
P
um ro
p up
G
ro
up
Li
M
i ll
Total Pump losses 01/01/2008 - 01/04/2010
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
Chart 3.1 Pump Losses for all groups
High production losses are not the only indicator of high cost equipment in the refinery. Further
examination of the process loss reports reveals that the of the 34,500 tonnes of alumina losses
attributed to the Mill injection group 99.3% of them relate to Gland failures and 0.7% are
attributed to Drive belt failures.
Mill Injection Pumps Losses by Failure Mode

(Tonnes of Alumina)
-208
Belts
Gland
Failures
-31306
Chart 3.2
Using the MTBF calculation of (Run Time/Failure count) for a given time period the MTBF can
be calculated for similar pump groups across the refinery.
27
MTBF for se lecte d refine ry pum p groups
Run Time (Days)
120
100
80
60
40
20
0
Mill Injection
Group
Stage 3 Inter
Stage Cooling
Pumps
Liquor Booster Liquor Injection

Pumps
Pumps
Mill Slurry
Pumps
Chart 3.3
The results pictured in chart 3.3 displays the fact that the mill injection group have one of the
lowest MTBF for glands on site and coupled with their high availability requirements (4 groups
feeding three stages) lead to why the Mill injection group feature so heavily on the production
loss reports.
This low MTBF not only cost the business in lost production tonnes but also contributed to high
maintenance expenditure. An examination of the maintenance record associated with this pump
group shows that from the period of 01/01/2007 there were 207 gland repack events carried out
(see Chart 3.4). At an average labour rate of $100.00/hr and using an average of $600.00 worth
of materials the total maintenance cost to perform a 4 hour repack for each gland failure event is
$207,000 or $2000 on a weekly basis.
Repack events on the Mill injection group since 01/01/2007
30
25
20
15
10
0
Q1 2007
Q2 2007
Q3 2007
Q4 2007
Q1 2008
Q2 2008
Q3 2008
Q4 2008
Q1 2009
Q2 2009
Q3 2009
Q4 2009
Q1 2010
Chart 3.4
28
Another important fact to be gleaned from the chart is the fact that the occurrence of gland
failures has remained fairly stable over the last several years this indicates that there has been no
real improvement in MTBF.
With the data which has been presented the following summarisation can be made.
The Mill injection Gland failures are currently costing the refinery approximately $13,000
dollars per day in production losses. (2010-04-16 spot price of Alumina from LME) and
maintenance costs. The successful completion of this research project concerned with
identification of the root cause and implementation of a solution would give significant return on
the invested time and energy by the reliability engineer.
3.4
Mill Injection Group
Mill Injection Pumps transport caustic slurry from the Mill relay tanks to the Low temperature
digesters in stage 1, 2 and 3. Due to the elevation of the digesters and the pressures associated
within the vessels the discharge pressure of the pumps need to be in the vicinity of 2000 kPa.
This is achieved by having two pumps in series. One pump operates as the low pressure pump
and feeds slurry to the high pressure pump. For the three stages to maintain flows there are 4
individual pump groups A, B, C and D. This provides redundancy in the system in the case of a
pump group requiring maintenance so as not to affect flows. In the event two pump groups
experience a mechanical or electrical failure flow cuts to a stage will be experienced which will
then be recorded as a production loss.
The pumps are identified using the site standard equipment identification.
A GROUP Low pressure pump P633-4A, High pressure pump P633-5A
B GROUP - Low pressure pump P633-4B, High pressure pump P633-5B
C GROUP - Low pressure pump P633-104C, High pressure pump P633-105C
D GROUP - Low pressure pump P633-304D, High pressure pump P633-305D
P equipment type (Pump)
633- Area of refinery equipment is found
4 Low pressure, 5 High Pressure
A - Group identifier
29
Mill Injection Group
Figure 3.4. (Mill injection group schematic. Obtained from Process book)
3.4.1
Expansion of Refinery
In 2007 the refinery underwent an expansion where a third stage was built. To accommodate this
the mill injection pumps was increased by another group (D).
The consequence of the expansion meant that instead of three groups feeding two stages which
required 66% utilisation it then required that 4 groups feed three stages which requires 75%
utilisation. This increased the reliability requirements of the pump groups.
3.4.2
Pump Makes
A, B and C Group. Comprises of three K&L 6X8 LSA25 on the low pressure side and three
K&L 6X8 LSA25 HP on the high pressure side.
K&L refers to the pump manufacturers (Kelly and Lewis), 6 x 8 refers to the discharge and
suction pipe size respectively (inches), LSA refers to slurry application , 25 refers to the impeller
diameter size (25 inches) and HP refers to high pressure application.
30
Figure 3.5. Mill Injection Pumps (note the steel shroud over the pump which acts as a safety guard in the event of a scat
from the mill enters the pumps and causes catastrophic failure)
These pumps have been in service since 1972 (Plant commissioning) and are generally subjected
to the following conditions
Maximum Flow 366m3/hr
Normal Flow 340m3/hr
Minimum Flow 206m3/h
Slurry SG 1.824 / Solids SG 2.4 / Liquid Density 1.3 / Concentration 54.8%
On the low pressure side the pumps are driven by a fixed speed 160kw AC motor the power is
transferred through a reduction belt drive and delivered to the pump at 811rpm. The high
pressure pumps are similar except for the fact that a variable speed fluid drive is utilised between
the motor and the v-belt pulley this enables the High pressure pump to regulate its discharge
pressure.
31
D Group. Comprises of a Warman 6x8EE - AH on the low pressure side and a Warman 6x8FFAHP on the high pressure side. Warman refers to the manufacturer (Weir Minerals), 6 x 8 refers
to the discharge and suction pipe size respectively (inches), EE or FF refers to the frame size
(Bearing cartridge) respectively and AH indicates that is for an abrasive heavy duty slurry
application and the P indicates high pressure.
These pumps have been in service since 2007 (3rd stage commissioning) and are generally
subjected to the following conditions (same as A, B and C group)
On the low pressure side the pump is driven by a fixed speed 185 kW AC motor, power is
transferred through a reduction belt drive and delivered to the pump at 1028rpm.The high
pressure pump is identical except for the fact that the motor is controlled by a VSD. This way the
speed as with the A, B and C group the speed can be regulated to control pressure to the
digesters.
Figure 3.6. D Group Pumps

This pump group forms a critical part of the refinery. The unplanned failure of these pumps
immediately limits the ability of the refinery to make alumina which then leads to loss of
income.
32
CHAPTER 4
LITERATURE and BACKGROUND REVEIW
4.1
Aims of the Chapter
Research the basic features of a centrifugal pump.
Describe the unique features of Slurry Pumps
Provide a description of pump glands their composition and purpose.
Research failure modes concerning pump glands
Outline current industry methods in improving slurry pump gland reliability
Indentify previous work or reliability projects which have been carried out on the mill
injection pump glands.
Provide an evaluation of the information researched.
4.2
Introduction to Centrifugal Pumps
Pumps are used to transfer liquids from low-pressure zones to high-pressure zones (Bachus,
2003, p.1). To do this effectively the liquid needs to be at the suction of the pump at a required
pressure so the pump can work with the fluids attributes. Centrifugal pumps cannot draw fluid
into the housing.
P1/ + v1/2g+ Z1 = P2/ + v2/2g+ Z2= P3/ + v3/2g+ Z3
Bernoullis Equation
The principal components in a pump which gives it the ability to transfer liquids is the volute and
the impeller. The impeller rotates on a shaft and allow liquid to enter its eye (See Figure 4.1).
The liquid then travels along the impeller vanes and is accelerated towards the outside of the
impeller. Following Bernoullis equation this creates a low pressure zone at the eye of the
impeller due to the increase in velocity. When the fluid reaches the end of the impeller the
velocity rapidly decreases which increases the pressure. Due to the positioning of the impeller
and the shape of the volute the fluid is then forced along a path by the impeller and direct
towards the discharge. As the fluid travels around the volute the velocity decreases due to the
increasing area available for it to fill the effect of this is an increase in fluid pressure.
33
This pressure differential that has developed between the suction and the discharge now allows
the liquid to be transported through a piping system at a specific flow rate with a specific head.
Figure 4.1 Pump Dynamics (Bachus, 2003, p.3)

On the manufacture of a pump impeller, designers will subject the impeller to hydraulic testing,
this testing involves the flows produced by the impeller gradually being restricted until flow shut
off is achieved. The data which includes flow rates, power consumption, discharge and suction
heads is collected and presented on what is known as a flow curve (see Figure 4.2), Each
impeller design has its own flow curve and enables engineers to design a pumping system to
match the flow and pressure requirements. In the interest of efficiency it is always preferable to
run the pump as close to the best efficiency point as possible. The B.E.P is the point where the
power coming out of the pump (water horse power) is the closest to the power coming into the
pump (brake horse power) from the driver. This is also the point where there is no radial
deflection of the shaft cause by unequal hydraulic forces acting on the impeller.
(http://www.mcnallyinstitute.com/Charts/Glossary-html/Glossary_B.html 18/05/2010)
34
Figure 4.2 Flow Curve (McNally Institute 2-3)
4.3
Introduction to Slurry Pumps
Slurry pumps are used to transport solids in a liquid medium. The solids may consist of bauxite,
silica and other minerals which can be up to10mm in diameter and are highly abrasive. This
aggressive application challenges traditional clean fluid pump technologies and has forced
pump manufacturers to incorporate numerous design changes into their components to perform
the duty as per requirements. The changes include the following.
Usage of high chromium alloys white iron to resist abrasive wear in the wet end
components.
Greater wall thicknesss in the wear components.
Greater internal clearances concerning the impeller to throat bush (reduced Efficiency)
Greater use of packing as opposed to mechanical seals for pump sealing.
Different pump stuffing box configurations as opposed to clean homogenous solution

pumps.
Greater shaft rigidity.
Larger bearing sizes to accommodate bigger loads.
Blunt tipped impellers.
Lower number of vanes used in the impeller.
Pump run at a lower speeds
(Slurry Pump Handbook. 2009)

35
Figure 4.3 (Obtained from a Warman parts guide for an 8/6 FF-AHP High Pressure Slurry
Pump.)
36
4.4
Gland Configuration and Performance in Slurry Pumps
The stuffing box houses the packing assembly and is located where the shaft penetrates the
casing that is under pressure. In slurry pump applications the stuffing box is bolted to the casing.
The stuffing box bore is concentric to the shaft sleeve and of a specific size so as to
accommodate rings of packing (Volk. 2005, p334).
Compression packing is most commonly used on rotating equipment. The seal is formed by the
packing being squeezed between the inboard end of the stuffing box and the gland. A static seal
is formed at the ends of the packing rings and the inside diameter of the stuffing box. The
dynamic seal is formed between the packing and the shaft sleeve. (Karassik et al. 1986, p2.114).
Seal water should be available at 10 psi (0.7 bar) above maximum pump discharge pressure.
(LSAS Technical Booklet).
A negative attribute associated with this type of sealing is the fact that it allows considerable
water into the wet end of the pump which dilutes the product. In the Alumina industry water
ingress into the slurry reduces the yield and means that the water will have to later be extracted
through either the evaporation or calcination process. Therefore it is important to run the gland
water to an optimum level.
There are essentially two different stuffing box configurations that can be utilised by slurry
pumps (See Figure 4.4).
Low flow version is the most common in the alumina industry as it requires the least
gland flush to work effectively.
Forward flush configuration tends to be used more in high pressure pump applications
(+2000kPa) due to a tendency for the packing to be extruded through the clearance of the
shaft sleeve and stuffing box by the high pressure gland water which is turn on before the
pump is started and pressurised. In this case water forms a barrier between the slurry and
packing.
37
Packing rings
Lantern ring
rings
Figure 4.4 Stuffing Box Configurations (Obtained by LSA Technical Book from GIW)
These two configurations are the most common method of sealing slurry pumps, mechanical
sealing is used widely however unless conditions are stable and the pump is allowed to run close
to its design, problems often arise and expensive repairs can occur. Stuffing box type sealing
provides a more robust configuration and is able to operate effectively under a broader range of
conditions.
4.5
Gland Failure Characteristics
The following points discuss common modes of failures for pump glands. The research into this
will enable a better understanding and assist when identifying the root cause of failure
concerning the mill injection pumps.
4.5.1
Worn Shaft Sleeve

o Stuffing box shaft sleeves are surrounded in the stuffing box by packing: the sleeve
must be smooth so that it can turn without generating heat (Karassik et al. 1986,
p2.117).
o For packing to operate properly, the finish on the shaft sleeve must be at least 0.4
m). The sleeve must be harder than the packing and chemically resistant to the liquid
being sealed. (Karassik et al. 1986, p2.117).
38
o For pumps subject to slurry the surface must be hard so as to resist wear, chromium,
tungsten carbide and ceramic and some of the materials used for severe service. If the
sleeve has a coated material for a hard wearing surface, the sleeve must also have
good thermal shock. (Karassik et al. 1986, p2.116).
Research into shaft sleeves suggests that the material of construction is critical to its ability to
resist wear. In the case of the Gove alumina refinery the type of pump which is used in the mill
injection area is also used many other applications some being considerably less severe in
application than the mill injection pumps. The same shaft sleeve material is in installed into
everyone of the pumps. This material while being suitable for most of the applications may not
be suitable in the mill injection area.
4.5.2
Poor Packing Material Selection.
Packing requires a number of attributes to work effectively. The correct balance of these
attributes is required for the optimum reliability to be achieved, the attribute are as follows
o
Conformability The ability for the packing material to adapt to its volumetric
constraints and provide an effective seal.
Lubricity The packing is impregnated with lubricant (usually graphite) it needs to

be able to retain it so as or start up it wont over heat.
Low coefficient of expansion As the pump shaft speed increases so does the
temperature. The packing needs to expand as little as possible.
Braid construction Different weaving configuration can affect the packings ability
to retain its shape during operation.
Low abrasiveness Material needs to be soft enough so as not to wear the shaft
sleeve
Ease of installation The easier the packing is to install the less likely failure will be
result
(http://www.impomag.com/scripts/ShowPR.asp?RID=7811&CommonCount=0, Al Guizzetti,
Product Specialist, W. L. Gore & Associates, Inc., Newark, DE, 17/05/2010)
39
Having many attributes selecting the correct packing is about finding the right balance of
qualities to suit the application. At the Gove alumina refinery there are two pump packing types
used. This packing material has been deemed suitable to work in a wide range of applications
from acid service, clean fluid and slurry. The definition of suitable needs to be investigated
4.5.3
Incorrect Gland Water Supply

o Under pressure Slurry will force its way underneath the packing quickly abrading
the packing undermining its ability to form a seal.
o Over pressure Gland water will extrude the packing through the clearance between
the shaft sleeve and the stuffing box into the wet end of the pump.
o Low flows An important function of gland water is to remove heat from the stuffing
box. Excessive heat build up will burn the packing.
o Poor gland water quality Impurities in the gland water can block piping restricting
the flow or impregnate the packing undermining its ability to seal.
o The quality, quantity and pressure of this gland sealing water is of prime importance
and must be carefully matched to the duty required.
(Slurry Pump Handbook. 2009, P2-12).
In large refineries where there are extensive gland water systems uniform pressure and flow are
difficult to achieve. Further to this as the refinery ages the piping deteriorates and scaling from
corrosion will progressively get worse. Pressure, flow and water quality will need to be measured
at the individual pumps to ascertain whether it is a contributing factor to poor gland reliability.
4.5.4
Incorrect packing installation

o Wrong sized packing length Excessive stuffing box clearances allows slurry to exit
wet end of pump.
o Over compressed packing Prevents gland water from cooling shaft sleeve.
o Under compressed packing Excessive stuffing box clearances allows slurry to exit
wet end of pump and into atmosphere.
This type of failure can be attributed to the training and skills of the pump tradesmen. The Gove
refinery employees approximately 100 fitters with varying degrees of skills. For a chronic issue
40
such as the mill injection glands as opposed to a sporadic one it is unlikely that poor installation
is responsible.
4.5.5
Wrong Pump Application

o High pressure spikes- Can be caused due to control valves on the discharge side
suddenly opening and closing. This can cause the discharge pressure of the pump to
momentarily to exceed the gland water pressure. This will inject slurry into the
packing deteriorating packing life.
o Pump operating off its best efficiency point At the best efficiency point the
balance between flow, pressure and pump area is in equilibrium. When the pump is
operated away from the BEP the flow, pressure and area ratios become imbalanced
which create radial forces. These radial forces will increase as the pump operates
further away from the BEP. As the radial forces increase so to will the amount of
shaft deflection which is occurring. The deflection will cyclically deform the packing
decreasing its ability to perform its sealing duty.
(Know and Understand Centrifugal Pumps Larry Bachus and Angel Custodio 2003)
This suggests that a pump which operated at or close to its BEP will run reliably. To do this the
pump and piping system needs to ensure the following.
Discharge valves remain in a constant position.
There is no pipe scaling to build restriction in the lines.
Operations require constant flows.
In reality this is not possible so some shaft deflection will be experienced and it needs to be
distinguished as to what is the acceptable level of shaft deflection.
4.5.6
Mechanical Fault
o Bearing failure Bearing wear will increase the radial run-out of the shaft. This runout will deform the packing decreasing its ability to perform its sealing duty.
o Bent Shaft Like a bearing failure and shaft deflection the bent shaft will deform the
packing decreasing its ability to perform its sealing duty.
41
These mechanical faults would be easily detectable and the symptoms of gland failure would be
expected to disappear after the corrective action has been performed.
Knowing and understanding the different failure modes which can occur in a stuffing box will be
an important tool when attempting to isolate a root cause in a gland failure. They will become
points of reference for later in the project.
4.6
Industry Methods in Improving Slurry Pump Gland Reliability
Slurry pumps are widely used in the resource industry due to their ability to transport solids in a
liquid medium. Australia being a major producer of raw materials relies heavily on slurry pumps
and consequently there is a large industry centred on the supply of products which are constantly
being developed for the purpose of improving gland performance. The following products are
relatively new on the market and give an example of possible solution that may be used in the
mitigation of a root cause of a reliability issue concerning glands. Further information about the
products can be found in Appendix B.
4.6.1.
Grooved Flow Restrictor Bushes These components are installed in forward flush
stuffing box configurations and replace the standard bush installed by the pump manufacturer.
The Grooved Flow Restrictor Bushes have a tapered spiral groove manufactured on the inside
diameter which is counter rotational to the shaft direction. Additionally the internal diameter of
the grooved section of the bush is conical, this then performs the following actions
o Gland water solids which are in suspension are separated to the outside of the bush
and channelled out of the stuffing box and into the wet end.
o Pumping action of the tapered spiral increases the pressure towards the wet end
maintaining an increased pressure differential.
o Reduces water usage by utilising tangential ports which distribute the water more
effectively.
4.6.2.
Live Loading on the Glands For a gland to work effectively the compression on the
packing needs to be maintained. Traditionally this was performed by periodic nip ups by the
42
maintenance staff. However there are often factors which prohibit this action occurring and the
reliability of the gland is affected. These factors include
o Poor access in the case of vertically mounted pumps or remotely located.
o Pump which present hazards such as high temperature, high pressure pumps.
o
Pumps which have low maintenance intervals and experience a long time between
inspections.
This product maintains a constant force on the gland follower through either spring or hydraulic
loading which ensures that the compression is maintained.
4.6.3.
Grease Purging of the Gland In applications where gland water quality is poor or the
pressure and flow is unsuitable grease can be used to replace the water as a flushing medium.
Grease pressure is maintained through a grease pump and pumped into the gland. Due to the
superior lubrication qualities of grease compared to water only a fraction of the normal flush is
used.
4.6.4.
Stuffing Box Bearings Pumps that have a low shaft stiffness ratio and that are
operating significantly away from their B.E.P can experience shaft deflection due to the
unbalanced forces in the wet end of the pump. The Stuffing box bearing is installed in the
stuffing box and provides support to the shaft close to the impeller. This minimises deflection
and increases gland life by maintaining packing shape.
4.6.5.
Slurry Packing- Due to the tendency for slurry pumps to be exposed to far more
aggressive applications then clean fluid pumps slurry specific packing has been developed.
Slurry packing incorporates re-enforcing fibres to maintain integrity, enhanced lubricants to
reduce friction and complex braiding to maintain shape.
4.6.6.
Mechanical Seals- Traditionally mechanical seals have been avoided in slurry
applications. This is due to the limited range of applications a mechanical seal can be subjected
to. Axial loading, shaft run out, poor flush water and high solids will all contribute to mechanical
seal failure. In recent years slurry seals have been developed to better cope with these
applications and consequently can be a viable solution in the case of poor gland reliability.
43
4.6.7.
Stiffened Shaft Assemblies These assemblies are provided by Pump manufacturers to
avoid shaft deflection occurring. In the case where a pump is required to operate significantly
away from its B.E.P shaft deflection can lower gland life through the deformation of the packing.
The stiffened shafts or larger frames have better rigidity and are able to handle greater loading.
4.7
Previous Projects Concerning the Mill Injection Pumps
Over the last ten years there have been several projects that have investigated the poor gland
reliability issue concerning these pumps. These projects were performed by engineering and
maintenance personnel with the aim of mitigating the production losses. The project description
and details have mostly been extracted after examining the historical records of the pump crew
meeting minutes, past work orders raised and interviewing of the staff who were involved with
this pump group previously. An example of this evidence of this can be found in Appendix C.
4.7.1.
Gland Water Pressure Differential Project - Project was carried out in 2006 as part of
a continuous improvement project. Due to process conditions the discharge pressure on the high
pressure pump side fluctuates making the pressure differential between the gland water and
pump discharge unstable. A pressure control system was installed to maintain a constant 200kPa
difference. This ensured that the pressure differential was constant. This had a positive effect on
the MTBF for the high pressure pumps (Pumps identified as having a 5 in their equipment
number). Referral to Chart 4.1 displays the high pressure pumps as having a significantly higher
MTBF as compared to the Low pressure pumps the exception to this is the 304D pump which is
a different make of pump compared to the other three.
Days
MTBF (Based on Run Time)

180
160
140
120
100
80
60
40
20
0
MTBF (Days)
Average
Target
P633- P633- P633- P633- P633- P633- P633- P6334B

4A
104C 5A
5B
304D 105C 305D
Chart 4.1
44
The implementation of this improvement would have little effect on the low pressure pumps as
the speed and therefore discharge pressure is constant. The performance of 304D will be worth
investigating as to why it performs significantly better than the other low pressure pumps.
4.7.2.
Installation of a Grooved Flow Restrictor Project was carried out in 2005 by the
workshops engineer at the time. This involved the reconfiguration of the stuffing box from a
standard configuration to a forward flush configuration where a grooved flow restrictor was
fitted.
Project was unsuccessful and the stuffing box was re-converted back shortly after. There is no
data available to analysis as to the impact the project had on the gland MTBF, however it has
been suggested that the grooved flow restrictors were contacting the shaft sleeves and damaging
the components.
4.7.3.
Repack MST This was instigated in 2006 by the pump maintenance co-ordinator with
the aim of repacking the Pump glands on a scheduled basis before the MTBF point. A work
order was raised on a fortnightly basis and the repack planned in. Using this strategy it was
thought that the production loses could be avoided.
With eight pumps in the mill injection group it meant that one pump was required to be isolated
ready for maintenance approximately 8% of the time. Due to process requirements operations
could not afford to release the pumps for that frequency. Additionally when a pump was
available pump crew often lacked the manning to perform the task. The MST was cancelled after
one year.
4.7.4.
Mechanical Seal Installation Prior to 2000 there is some evidence that suggests a
slurry mechanical seal was trialled on these pumps however verification has not been achieved.
Given the poor reliability associated with the pump glands it is doubtful as to the suitability of
this trial.
45
4.7.5.
Slurry Packing Trials Throughout the operational life of the mill injection pumps
there have been numerous packing trials that have been conducted. Different packing suppliers
have offered to help solve the issues by trialling their products. Slurry packing with different
weaves, yarn material and lubricant base have all been trialled with little effect.
Despite there being numerous slurry packing manufacturers the variation in make, quality and
composition is limited. Therefore the difference in performance will also be limited. To enact a
significant change in gland reliability changing the slurry packing brand is unlikely to have the
desired effect. However once the root cause is addressed the fine tuning process may involve
trialling different packing compositions.
4.8
Conclusions
The research conducted in this chapter has had the following purpose.
Understand as to how pumps glands work.
Understand how pump glands can fail.
Understand what products or upgrades can be retrospectively installed so as to mitigate

the gland failures.
Understand the successful and unsuccessful improvement work that has already been
conducted on this pump group and what effect it has had on the pump gland reliability.
The completion of this research has provided a broad understanding of the reliability issue at
hand and will enable future work to be focused towards the higher probability scenarios. Using
this strategy we can refer to Chart 4.1 which displays the fact that the poor gland reliability
mainly exists in the Low pressure A, B and C group pumps. Therefore to maximise the value
from this project we can exclude the other pumps from further analysis.
The fruits of this chapter will now lay the foundations for the following chapters which include
data collection, root cause analysis and the development and implementation of potential
solutions.
46
CHAPTER 5
DATA COLLECTION and ANALYSIS
5.1
Aims of Chapter

Determine the relevant data which is to be collected based on the research conducted and
existing data previously displayed in this project.
Display data which has been collected or calculated
Report and discuss the of relevance of the collected data
This data analysis will then be combined with the next chapter to enable a complete Root Cause
analysis to be conducted.
5.2
Data Required
Referring to Chart 4.1 which displays the MTBF of the mill injection pumps. The chart indicates
that the D group pumps (Warman Manufacture) and the High pressure pumps (Identified by the
5 in the equipment number) are significantly better performers than the low pressure A, B and C
group pumps. This identification allows the data collection to focus on those three pumps.
The literary review in Chapter 4 indicated that the following failure modes associated with
Glands are the mostly likely modes which will yield the root cause.
Worn Shaft Sleeves
Poor Packing Selection
Incorrect Gland Water supply
Wrong Pump Application
Mechanical Fault
To confirm the relevance of each failure mode to the mill injection pumps Table 5.1 displays the
information that needs to be gathered.
47
Data Collection Plan

Failure Mode
Data Requirement
Data Source
By
When
Collection
Freq / Span
Worn Shaft Sleeves
Shaft Sleeve make

Shaft Sleeve wear
Drawing Register
D.Bishop
30th Jun
st
Once
Visual Investigation
D.Bishop
01 Aug
2 Examples
Shaft sleeve MTBF
Process History and Ellipse
D.Bishop
01st Aug
Once
compared with similar.
work orders
Current Packing used
Equipment Parts listings
D.Bishop
01st Aug
Once
Packing condition after
Visual Investigation
D.Bishop
01st Aug
2 Examples
D.Bishop
01st Aug
5 Samples
D.Bishop
01st Aug
5 Samples
attributes
Poor Packing
Selection
failure
Incorrect Gland
Required Gland Water
Manufacturers Specs (LSAS
Water supply
flow
Technical Booklet).
Required Gland Water
Manufacturers Specs (LSAS
pressure
Technical Booklet).
Actual Gland Water flow
At Pump
D.Bishop
01st Aug
Once
Actual Gland Water
At Pump
D.Bishop
01st Aug
Once
Flow curve of pump
Pump Manufacturer resources
D.Bishop
01st Aug
Once
Actual flows pump is
Pi Historical trends
D.Bishop
01st Aug
1 Months data
Shaft deflection
Calculations using
D.Bishop
01st Aug
1 Months Data
experienced by the pump
Manufacturers specifications and
for given flows
flow data from process history
Mechanical inspection
Failure Investigation
D.Bishop
01st Aug
2 Examples
pressure
Wrong Pump
Application
subjected to. Relation to

B.E.P
Mechanical Fault
after gland failure
Table 5.1 Data Collection Plan
48
5.3
Data
5.3.1
Suitability of Shaft Sleeve
Make Shaft Sleeve is manufactured from mild steel, with Metco 34F hard facing, hardness tested
to Rockwell 'C' 60+.
Metco 34F is a Tungsten Carbide powder blended with nickel. The powder is thermally fused on
to the shaft sleeve and requires minimal grinding. This gives the wear surface both the hardness
and corrosive resistance requirements for this application.
Visual Inspections after use
Wear mark from contact with wear plate
gives evidence of Shaft Deflection.
Figure 5.1 After 10 days in operation
Figure 5.2 After 77 Days- rendered unusable

MTBF Current MTBF with the existing shaft sleeves is 70 days run time. This is poor compared
against other shaft sleeves in similar applications where 180 days has been achieved.
49
5.3.2
Suitability of Packing Selection
Make Current packing used is 1400r Graphmax 3/4 sq section manufactured by Chesterton. This
packing is rated to 550 deg and 120 Bar rating.
Failure Inspections
The square section packing looks to have been
deformed due to shaft deflection
Fig 5.3 5 Days run time
Packing has abraded on the shaft sleeve size

due to ingress of slurry into stuffing box
Fig 5.4 21 Days run time
50
5.3.3
Suitability of Gland Water Supply
Required Gland Water Flow - Maximum of 18l/min Gland flush be used No minimum
stated. The document suggests that considerably lower flows than the maximum stated should be
achievable with no adverse effects.
Actual Gland Water Flow Current Maric flow restrictor set at 12 l/min
Required Gland Water Pressure - 70kPa above maximum discharge pressure.
(Obtained from LSA Technical Booklet). At 811rpm the pumps maximum discharge pressure is
49m head. With a slurry S.G of 1.824 this equates to a pressure of 868 kPa. The means that the
gland water pressure of 1000 is close to the optimum recommended by the manufacturers of 940
kPa
Actual Gland Water Pressure - HP water source has been recorded at between 950 and
1000kPa depending on supply requirements.
Figure 5.6 Gland Water pressure at 950kPa
51
5.3.4
Suitability of Pump To Application
Flow Curve
Operating
Point
B.E.P
Chart 5.1 Flow Curve for 6-8 LSA -25

B.E.P for the current pump installed is at 687m/hr and 42.9 m head (See Chart?)
Average Operating Point is 275m/hr at 49m head.
52
Actual Flows pump is subjected to (Data extracted from Process history book)
Mill Injection Flows Using A, B or C Groups
330
310
m^3 / hr
290
270
Stage 1
Stage 2
250
Average
230
210
190
30
/0
6/
10
2/
07
/1
0
4/
07
/1
0
6/
07
/1
0
8/
07
/1
0
10
/0
7/
10
12
/0
7/
10
14
/0
7/
10
16
/0
7/
10
18
/0
7/
10
20
/0
7/
10
22
/0
7/
10
24
/0
7/
10
26
/0
7/
10
28
/0
7/
10
170
Chart 5.2 Stage 1 and 2 Mill Injection Flows for a 1 month period
Pump Shaft deflection (See Appendix for calculations involved)
Shaft Deflection at Pump Shaft End
1.30
1.29
1.28
mm's
1.27
1.26
1.25
P633-4A
P633-4B
Average Deflection
1.24
1.23
30
/0
6
/2
01
2/
0
07
/2
01
4/
07 0
/2
01
6/
07 0
/2
01
8/
07 0
/2
01
10
0
/0
7
/2
01
12
0
/0
7
/2
01
14
0
/0
7
/
16 201
0
/0
7
/2
01
18
0
/0
7/
20
20
10
/0
7/
20
22
10
/0
7/
20
24
10
/0
7/
20
26
10
/0
7/
20
28
10
/0
7/
20
10
1.22
Chart 5.3 Calculated Pump Shaft deflection for the 1 month period
Recommended Shaft deflection as per manufacturers specifications for a 8/6 LSA 25 is <0.5mm
from the end of the shaft.
53
5.3.5
Mechanical Inspection on Gland Failure
Refer to Appendix F for full maintenance reports.

The maintenance reports which were conducted in response to gland failures involved the
following checks.
Stuffing box to shaft concentricity
Bearing wear
Shaft sleeve wear
Gland water pressure checks
Gland water strainer checks
Impeller clearance checks
General condition checks.
The results of these check found that mechanically the pumps are okay and that it is unlikely that
any mechanical issues are causing gland failures. It is worth noting that one of the investigations
noted that the shaft sleeve had been contacting the wear plate. This indicates that the shaft is
deflecting..
5.4
Conclusions
It is important not to draw to many inferences from the collected data. The data which has been
collected will be an important tool when conducting the root cause analysis. Potential causes
which are brainstormed will require validation it is at this point that the data can be referred to, to
assess the validity of the cause.
54
CHAPTER 6
APOLLO ROOT CAUSE ANALYSIS
6.1
Aims of Chapter
The specific aims of this chapter utilise an RCA process which will enable the best and most
practical solution to be found. The RCA process to be followed is called Apollo. The Apollo root
cause methodology explores the relationships between the cause and effects and continues until
all plausible root causes are exhausted. The Apollo root cause methodology is a standard Rio
Tinto procedure and is facilitated by a software program.
The Methodology simplified is as follows
6.2
Incident report this helps to define the problem
Reality chart This consists of brainstorming all possible causes.
Solution generation Finding solutions for all possible causes
Solution selection Grading the solutions until the optimum one is found
Incident Report
INCIDENT REPORT
Purpose: To investigate production losses relating to the Mill injection Pumps, not to place blame.
For Internal Use Only
Report Date: Jul. 30, 2010
Start Date: Jul. 25, 2010
Report Number: 001
I. Problem Definition
What: Repetitive Mill Injection Pump Gland Failures
When: 01/01/2008 till present
Where: Area 633 of the Rio Tinto Alcan Gove Alumina Refinery
Significance: High Significances (Revenue loss)
Safety: No Injuries
Environmental: No EHS events
Revenue: $10,000,000 over a 28 month period in production losses (Current Alumina prices)
Cost: Maintenance $207,000 over a 28 Month period
Frequency: Almost a daily occurrence
Table 6.1 Incident Report
55
6.3
Reality Chart
To conduct a Root cause analysis a team is required. To be effective the team needs to represent
the full cross section of the refinery. In this case it is Operations, maintenance and engineering.
Team Members
Name
Email
Member Information
Damon Bishop
damon.bishop@riotinto.com
Rotating Equipment Engineer
Scott Smith
scott.smith@riotinto.com
Pump Crew Supervisor
Terri Dupe
terri.dupe@riotinto.com
Graduate Engineer
Dave Bennett
dave.bennett@riotinto.com
Mills Operations supervisor
Table 6.2 RCA Team

Refer to Appendix G for the completed reality chart
Reality chart Summary
The results of the reality chart indicate that the Root cause of the repetitive gland failures is principally due to
shaft deflection which is resulting from the hydraulic imbalance in the pump. This cyclic shaft deflection is
deforming the packing which is resulting in slurry exiting the pump due to the inability of the packing to seal. The
Hydraulic imbalance is due to the pump being operated away from the B.E.P. Secondary causes which have also
contributed include periodic gland maintenance not being performed. it has also been recognized that it may be
worthwhile in examining alternative shaft sleeve material with the goal of extending gland life.
Table 6.3 Summary
6.4
Solution Selection
The Solution assessment chart Table 6.5 displays the potential solutions and ranks them as per
the criteria listed. As is represented in Table 6.4 and 6.5 the installation of a stiffened shaft has
been seen as the most comprehensive solution by the team members. This selection of this
solution has been based on the following.
Subject matter expertise in the group (50 years pump experience in the group).
Failure investigations
Data collected
Calculations performed
Literary and Background review
56
Primary Solution
Causes
Poor Shaft slenderness ratio
Solutions
Install stiffened shafts
Solution Owner
Damon Bishop
Due Date
Aug 28, 2010
Table 6.4 Solutions Generated From RCA
SOLUTION ASSESSMENT REPORT

Criteria
Total Cost
Summary
Weight
2
Ranking
1 (Expensive)
to 4 (LowCost)
Score
Criteria
Ease of
Implementation
Weight
2
Ranking
1 (Difficult) to 4
(Easy)
Criteria
Probability of
Recurrence
Weight
Score
Score
Criteria
Effectiveness
Weight
2
Ranking
8 (98-100%) to
1 (0-2%)
3
Ranking
1 (Not Eff.)
to 4 (Very
Eff.)
Score
Criteria
Return on
Investment
Weight
2
Ranking
1 (<100%)
to 4
(>1000%)
Score
Total Score
Cause
Solution
Comment
Pump operating away from B.E.P
Install VSDs
25
Poor Packing Installation Techniques
Re- Training of Fitters
To Expensive. Major
Substation modifications to
accommodate VSD's
Fitters are trained
29
Work order cancelled

Shaft Sleeve Material to soft
Re-open Work order

Ceramic Shaft sleeve
4
4
4
4
3
2
2
1
4
4
36
31
See Above
27
Prohibitive costs involved
32
Excessive shaft overhang
Re-Open existing work orders

Examine different Shaft
Sleeve Materials
Install VSD on Low pressure
pump so as to be in tune
Change make of pumps to
Warman Pumps
Fit Stuffing box bearing
Unlikely to be a long term

solution as bearing will
wear in a slurry application
37
Poor Shaft slenderness ration
Install stiffened shaft
Install one then validate

solution
As above
42
42
This has been done

numerous times in the past
with little effect
Fitters well trained already
31
29
Daily Mill injection pump

checks to be performed
Ceramic Shaft sleeve
installed in Nov 2009.
Validate performance
36
31
Hi Pressure Pump restricting the

Low pressure pump
TKL pumps
Shaft Deflection
Install Stiffened Shaft
Packing has no elasticity.
Trial different Packing
Incorrect Installation
Provide training to maint

Personnel
Perform Nip ups
No Gland nip ups

Slurry Abrading Shaft Sleeve surface
Trial different shaft sleeve

material
Table 6.5 Solution Assessment
6.5
Solution Description
The current pump shaft installed in the mill injection pump can be seen in Fig 6.1 the shaft has
the same nominal dimension for the entire diameter excluding the impeller plug. The Bearings
are mounted on tapered adaptor sleeves and the sealing is a lip seal / labyrinth arrangement.
Figure 6.1 Existing Shaft Assembly

The stiffened shaft design as seen in Figure 6.2 displays a stepped shaft where the diameter of
the shaft has been increased where ever possible. To achieve this the following features are
enabled.
Parallel bore bearings as opposed to tapered.
Increased diameter between bearing to reduce shaft flex.
Larger sealing faces
Bearing housing base is thinner so as to enable larger bearings.
Tapered bearings on the drive end to reduce the end float due to axial thrust.
Figure 6.2 Stiffened Shaft Assembly
The calculated shaft deflection (See Appendix E for calculation details) for a Stiffened shaft has
been determined for the same period that was measured for the straight shaft.
Chart 6.1 displays an average shaft deflection of 0.48mm as oppose to 1.26mm which was
recorded for the straight shaft. The calculated 0.482mm is significantly lower than the straight
shaft value and is below the recommended 0.5mm shaft deflection suggested by the pump
manufacturers for optimum gland life. (LSAS Technical Booklet).
Shaft Deflection at Pump Shaft End

0.5000
0.4950
0.4900
mm's
0.4850
P633-4A
P633-4B
0.4800
Average Deflection
0.4750
0.4700
0.4650
30
/0
6/
2
2/ 01
07 0
/2
4/ 01
07 0
/2
6/ 01
07 0
/2
8/ 01
07 0
10 / 20
/0 10
7
12 /20
/0 10
7
14 /20
/0 10
7
16 /20
/0 10
7
18 /20
/0 10
7
20 /20
/0 10
7
22 /20
/0 10
7
24 /20
/0 10
7
26 /20
/0 10
7
28 /20
/0 10
7/
20
10
0.4600
Chart 6.1 Calculated Shaft Deflection with stiffened assembly

A dynamic shaft analysis was carried out by the pump manufacturer using their software tool
SLYSEL. The analysis summarised in Table 6.6 Revealed a similar result using slightly different
data (They used theoretical flows where as we used actual) .
Shaft Deflection Table
366m3/hr Max
340m3/hr Normal
206m3/hr Min
LSA 6x8/25 315/16 Standard

Shaft
LSA 6x8/25 35/16 Stiffened
Shaft Design
0.777mm
0.898mm
0.946mm
0.277mm
0.290mm
0.292mm
Table 6.6 KSB Pump Deflection Calculations
59
6.6
Conclusion
The solution which has been selected addresses the root cause but does not remove it. To remove
the root cause it would be necessary to run the pump with less restriction however the
requirements of the refinery do not allow this so it is not feasible. The advantage of the stiffened
shaft proposal is that the root cause will be mitigated while not disturbing the shaft centreline,
impeller hub dimensions, stuffing box dimensions or bearing housing dimensions. This reduces
the cost of the upgrade considerably.
To implement the optimum solution it will now be necessary to perform a detailed financial
analysis of the costs involved it installation of the stiffened shaft assemblies and the expected
benefits that will come.
60
CHAPTER 7
SOLUTION IMPLIMENTATION AND VALIDATION
7.1
Aims of the Chapter
Report on the accurate costing data for the installation of the mill injection shaft upgrade.
Report on the process followed to gain acceptance from refinery management as to the
benefits of the project and to consequently gain capital expenditure approval.
Provide a plan for the installation of the Shaft Assemblies.
Validate the Assembly once installed.
7.2
Mill Injection Pump Shaft Upgrade Cost
As earlier mentioned the pumps which have the worst performing gland MTBF are the older
Kelly and Lewis A, B and C group pumps. The newer Warman D group pumps have an
acceptable gland MTBF and do not require upgrading. Therefore a total of seven stiffened shaft
assemblies will be required for purchase. The purchasing of an extra one will enable a rotating
spare which can be available for installation in the event it is needed.
On investigating KSB pumps had provided a quote of $27,890 to supply the stiffened shafts (See
Appendix H) with a lead time of 13 weeks.
The work required to install a stiffened shaft assembly includes the dismantling of the pump wet
end and the removal of the belt drive pulleys, the shaft assembly can then be unbolted from the
frame and removed. The re-installation is the opposite of the removal process. This task
generally takes two fitters a full 12 hour shift to achieve with a crane group to assist.
Singular
Multiple
Total
Purchase price for Stiffened Shaft assemblies
$27,890
x7
$195,230
Installation costs for 2 Fitters
$200/hr
x 12 x 6
$14,400
Crane Requirements or 1 driver and 1 rigger
$300/hr
x3x6
$5,400
Total
$215,030
Table 7.1 Cost for Project
61
7.3
Acceptance From Refinery Management
To gain an in principle agreement to proceed with this project a presentation was conducted to
the refinery management team (See Appendix H for power point presentation). The presentation
communicated that this pump group was the worst performing on site with the greatest losses
associated with it. Additionally the root cause was explained and supported with engineering
data. The aligning of the pump manufacturers shaft deflection calculations with the project
author gave further credibility to the potential success of the upgrading of the Mill injection
pump shafts. A conservative net present value calculation was performed (See Table 7.2) which
provided a return of $3,200,000 based on a 7% rate of return over 5 years.
Net Present Value

H529 Mill Pump shaft assemblies
CEA
Total project cost (Est)
170,340.0
3,207,562.1
Benefits
NPV (at 7 %)
Table 7.2 NPV

The reception was positive with the accord to immediately purchase one Stiffened Shaft
Assembly from the maintenance budget with the other 6 to be purchased via the capital process.
A $200,000 capital project was initiated (See Appendix H for completed paper work) and
approved within 3 months.
7.4
Plan for Installation
The stiffened shaft assembly which was purchased using the maintenance budget was delivered
and installed in P633-4B (worst performer) in early April (See Figure 7.1 and 7.2). This was
advantageous as it enabled the solution to be validated and any modifications could be conducted
on the remaining 6 assemblies.
62
Figure 7.1 New assembly in box
Figure 7.2 Installed Assembly in P633-4B
63
The 6 remaining shaft assemblies have had to be purchased via the capital budget which due to
the process requirement takes 3 6 months. The shaft assemblies are due to arrive to site
01/09/2010.
In discussion with the Pump Crew co-ordinator it has been deemed to suitable to perform the
shaft installations when the scheduled 3 monthly preventative maintenance tasks are performed.
This way the compliance to the planned maintenance which occurs throughout the refinery will
not be compromised. Of the A, B and C group pumps the low pressure pumps have the lowest
MTBF and therefore are the logical place to start with the improvement work. Table 7.2 displays
an approximate plan for the installation dates.
Plan to Achieve 90 day MTBF for Gland life (Mill Injection Pumps)
Mar-10
Apr-10
May-10
Jun-10
Jul-10
Aug-10
Sep-10
Oct-10 Nov-10
Dec-10
Jan-11
90 Day Gland Life Acieved
Pump Upgrade
Measurments and quotes
Materials Purchased and Work Orders
Raised (To increase laoding capacity)
104C &
4A & 5A 105C
5B
Installation P633-
Complete
Pump Conversion
Mechanical Seal Installation (Possible
Table 7.3 Plan for Installation
7.5
Validation of the project 6 months on
The stiffened shaft assembly which has been installed into P633-4B has significantly increased
the gland reliability of the pump. As can be seen in Chart 7.1 previous MTBF for the gland on
this pump was 19 days. As of 10/08/2010 the gland had not failed which gave it 120 day MTBF
based on run time. This has given the project confidence that the implemented solution has
directly addressed the root cause which is that the pump is operating away from the B.E.P and
shaft is deflecting as a result.
64
MTBF (Based on Run Time)

180
160
140
Days
120
100
2010 MTBF 4B
Upgrade
2009 MTBF
80
60
40
20
0
P633- P633P633- P6334B
4A
104C
5B
P6335A
P633105C
P633304D
P633305D
Chart 7.1 Comparison of MTBF for P633-4B before and After Shaft Upgrade
65
CHAPTER 8
CONCLUSION
Review of Project
8.1.1
Wins and Successes
By completing the objectives laid out in the project specifications the project has so
far achieved its original goal which was to reduce the production losses due to gland
failures to $0. Although only one stiffened shaft has been installed in this period the
MTBF increase has allowed that the 75% required utilisation to be met which can be
seen in Chart 8.1. Based on a continuation of these results the project cost of ~
$200,000 has been paid for with in the first month of the stiffened shafts installation.
Besides the benefits of increased production the implemented solution also reduces
the maintenance costs of the pump group and reduces the EHS risk associated with
gland failures.
Mill Injection Utalis ation
85
80
75
Utilisation Achieved since April
70
8.1
65
60
55
50
Chart 8.1 Mill Injection Utilisation

This project success can be attributed to good and clear communication between joint
stakeholders in the project. The stakeholders being:
Engineering who conceptualised developed and implemented the solution.
Production who provided the financial resources for the purchasing of the
solution.
66
Maintenance who provided the labour and crane resources to install the
solution.
8.1.2
Roadblocks and Delays
As earlier stated the production losses associated with the poor pump gland reliability
on the mill injection pumps were in the order of $12,000 per day. Therefore the
earlier that the stiffened shaft assembly was installed the quicker the return on the
investment could be realised. From the agreement from the area superintendent to
purchase of the first assembly to the installation took 6 months due to the following
delays:
The pump maintenance had exceeded their budget for that particular area and
there was a reluctance from the area to use their budget to enable the
purchase, this required clarification which took several weeks to resolve.
Once the purchase was enabled there was a 14 weeks delivery time (partly due
to the remote location).
Once arrived sat in the warehouse for 10 weeks due to several factors which
included lack of labour and crane availability for improvement work, on three
occasions the planned work was bumped to enable breakdown work to take
place.
8.2
Further Work to Complete Project
Following the Lean Six Sigma Methodology there are six stages of an improvement project, they
are as follows
Define
Measure
Analysis
Improve
Control
Validate.
67
Relative to this project we are still in the improve stage where the implementation of the solution
is taking place. Apart from the installation there are the following change management tasks
need to be completed.
Updating of the spares inventory to reflect the change.
Communication and training to the maintenance personnel concerning the different

impeller clearance requirements and bearing configurations.
Obsolescing of the previous shaft assemblies
Further to this it will be important to close the project with a validation of its success which
should be presented to the people responsible for the allocation of the required resources. This
way the positive return on the invested funds can be confirmed and will give confidence towards
the securing of funds for future reliability projects.
8.3
Further Benefits and Opportunities
Over the 30 period that this refinery has been in operation many of the original pump operating
points have changed due to optimisation or expansion projects. Consequently many of the pumps
are operating away from their B.E.P and experiencing hydraulic imbalance. One of the tools
developed in this project is a shaft deflection calculator (Appendix E). Provided the dimensional
features of the pump and flow characteristics are known the shaft deflection of any pump can be
measured and compared against the manufacturers specifications. This tool can now be applied
to other pump groups which are experiencing the same symptoms that have been experienced by
the mill injection pumps.
Further to this now that the shaft deflection on the mill injection pumps has been addressed
former projects which have previously been unsuccessful can now be implemented with the
objective of further increasing the gland life or reducing the parasitic water ingress into the
product from gland flush. These projects could include mechanical seal conversion or for the
purpose of reducing gland water the installation of a grooved flow restrictor.
68
8.4
Closing
The completion of this project has put to bed a long running reliability issue which has cost the
refinery significant losses. It is the authors belief that the main reason this issue had never been
solved previously is that although there is evidence to suggest that shaft deflection had been
identified as a cause the only solution which had been seen as possible to mitigate it was to
replace the pumps with an alternate design. This solution would entail the re-design of this area
of the refinery due to the piping, pumping and plinth modifications to adapt to the new pumps.
This work could have cost upwards of $5,000,000 to implement and then potentially still had the
same reliability issue at the end of it. The modification of the pump which the project
implemented was never previously considered and possibly not thought possible. Careful
adherence to the defect elimination process has enabled the previously hidden solution to be
exposed.
69
References
Michael W. Volk. 2005, Pump characteristics and applications Edition: 2, CCR Press,
Boca Raton.
Larry Bachus and Angel Custodio. 2003, Know and Understand Centrifugal Pumps,
ELSEVIER, Oxford.
Igor J Karassik, Joseph P. Messina, Paul Cooper and Charles C. Heald. 1986, Pump
Handbook second edition, McGRAW-HILL, New York.
Val S Lobanoff and Robert R Ross. 1985, Centrifugal Pumps design and application.
Second Edition, GULF PUBLISHING COMPANY, Houston.
William E Forsthoffer. 2005, Pumps, ELSEVIER Science & Technology Books, Oxford.
Ron Palgrave. 2003, Troubleshooting Centrifugal pumps & their systems, ELSEVIER
Science & Technology Books, Oxford.
Anthony Esposito. 1998, Fluid Mechanics with Applications, PRENTICE-HALL
INTERNATIONAL, INC, New Jersey.
Slurry Pump Handbook Fifth Edition. 2002, WARMAN INTERNATIONAL LTD.
LSAS Technical Booklet. 2001, GIW Industries Inc, Grovetown.
McNally Institute, Technical papers http://www.mcnallyinstitute.com, 18/05/2010.
Robert W. Fox, Alan T. McDonald and Phillip J. Pritchard. 2004, Introduction to FLUID
MECHANICS, JOHN WILEY & SONS, Hobroken.
Terry Wright, 1999, FLUID MACHINERY Performance, Analysis and Design, CRC
PRESS, Boca Raton
70
Appendix A
FACULTY OF ENGINEERING AND SURVEYING
Eng4111/4112 Research Project
PROJECT SPECIFICATION
FOR:
Damon Thomas BISHOP
TOPIC:
Improving Pump Gland Reliability
SUPERVISOR: Chris Snook

SPONSORSHIP: Rio Tinto Aluminium
PROJECT AIM: To investigate the process parameters which
contribute to poor pump gland performance in the mill injection area and implement appropriate
corrective actions which will provide an acceptable MTBF.
PROGRAMME: (Issue A, 3rd March 2010)
1. Research slurry pump sealing in industry (preferably the Alumina industry) with respect to designs,
performance and reliability.
2. Investigate previous improvement projects which have been previously performed on this pump group.
3. Collect and analyse the data concerning the cost to the business that have resulted in the mill injection
pumps poor gland performance. An example of this data is lost production tonnes, maintenance costs and
EHS events.
4. Develop an evaluation plan for the collection and monitoring of historical and current data concerning the
gland performance on the mill injection pumps. An example of these data is MTBF, gland water use, flow
and pressure data.
5. Analyse the field data for the purpose of isolating the root cause or causes.
6. Develop potential solutions to mitigate the root cause and present recommendations to management for
approval. Provide cost estimates, benefits and time lines.
7. Submit an academic dissertation on the research.
As time and resources permit.
8. Implement solutions by way of co-ordinating improvements with the Maintenance Co-ordinators, Planners
and Area Production personnel.
9. Validate solutions through a continuation of monitoring the field data.
AGREED ___ Approved
Date :
(Student)
/ 2010
Approved
Date :
(Supervisor)
/
/ 2010
Examiner/Co-examiner:_______ Approved __________
71
Appendix B
Industry Products to improve Gland Reliability
72
73
74
Appendix C
Previous Improvement Projects
Pump Maintenance Meeting 36
(Week 18)
Subject: Pump Maintenance Strategy
Date:
05/05/04
Time:
1.00 pm
Location:
Maintenance Planning Conference Room
Attendees:
Aaron Edwards, D Mcdermid, Max Murdy, A Fleming
Minutes:
A. Edwards
Minutes of Meeting
EHS
First Priorities:
GOV04050015 Cut to right hand index finger.
PERSONNEL ISSUES
Pump Crew
Graeme Bean - A/L - 24/04/04 - 05/05/05
Skin Chessels L/S 29/05/04 10/08/04
- A/L 01/08/04 26/08/04
Machine Shop
Pop Reardon - LSL - returns 24/05/04
Rick Hutchinson A/L 28/04/04 06/05/04
KPI
KPI (weekly) Week 3, 2004
EHS - Injury, Enviro, Loss
5S
Work Backlog
% Planned work completed (schedule)
% Planned Hrs of available Hrs
Opportunity loss pump failure
Continuous improvement projects
Critical pump response time
Rotating spare schedule compliance
Investigations completed
Measurement
1
38
Last: 2830 hrs
This: 3043hrs
61
54
Mill Injection 1170T
(
1
12
0
1
Target
0
36
2000-2500hrs
Compliance
No
Yes
No
65%
65%
0 Reported
No
No
Yes
1 per quarter
12
50%
1 per week
Score
Yes
Yes
No
Yes
5/10
Unplanned Work - Critical Pump Reaction Time
75
Criticality
Critical A
Critical B
Critical C
Critical D
Unknown
PRIORITY CALLED ON
EQUIPMENT ID
P633-5A
P633-105C
P633-104C
P646-6
P643-116
P633-5B
P633-4B
P633-4A
P633-101
P653-145A
P633-5B
P641-176
Events
2
7
3
PERSON
RESPONSIBLE FOR
CALLING PRIORITY
??
??
??
??
??
??
??
??
??
S Savage
T Graham
M Easterbrook
Response
Immediate
Immediate
Immediate
Hrs
23
46
32
HOURS SPENT ON
UNPLANNED PRIORITY
8
4
20
4
4
4
4
2
8
15
4
24
PUMP CRITICALITY
B
B
B
C
C
B
B
B
A
A
B
C
Percentage of Rotating Tasks achieved against plan = 0%

Maintenance Strategy
Continuous Improvement
P633 Area continuous improvement project

- Crown Bushes on order for trial in area expected to be delivered to Garlock week start 17/05
- Replacement of scaled pipework commenced by area. Parts have arrived and are being fabricated
Investigation completed for premature gland failure on pump P652-107A Test points were installed last week. 7
pumps are Part of Green Belt project Dave Hill
Site investigation of gland water pressure Major continuous improvement project AE to complete.
Review of PMs to remove invasive inspection of pumps and reduce number of PMs To be undertaken as a
continuous improvement project AE to complete by second quarter 2004.
Major Priority Work
Vacuum pump overhaul
HX gearbox
General Business
Area 643 pump and gland failures Follow up pressure checks once all stators replaced,
P646-22A/B Mechanical seal failure 22a piping in place to convert to stuffing box, awaiting gland water supply to
convert 22b
K&L 8*6 pumps Still awaiting bowls, on order since mid Feb. Will chase quote and lead time from GIW - AE
The following issues were brought up at the meeting for further consideration:
Millmax conversion of P633-4a Trial pump installed. Guard not refitted to pump (FP raised by area FP04040062)
Next review meeting

Date:
19/04/04
Time:
10.00am
76
Location:
Maintenance Planning Conference Room
SIZE PACKAGED AW mm LBS.10% KG.10% ITEM NO.
Fortnightly Repacks for the Mill Injection Pumps

Model
EQUIP_N
O
COMP_COD
E
MAINT_SCH_TAS
K
JOB_DESC_COD
E
SCHED_DESC_
1
K&L 6X8
LSA25
K&L 6X8
LSA25
K&L 6X8
LSA25
K&L 6X8
LSA25 HP
K&L 6X8
LSA25 HP
K&L 6X8
LSA25 HP
P633-104C
P100
0002
PM
P633-4A
P100
0002
PM
P633-4B
P100
0002
PM
P633-105C
P100
0002
PM
P633-5A
P100
0002
PM
P633-5B
P100
0002
PM
GLAND
REPACK
GLAND
REPACK
GLAND
REPACK
GLAND
REPACK
GLAND
REPACK
GLAND
REPACK
WORK_GROU
P
MPC
MPC
MPC
MPC
MPC
MPC
Packing Trial Tag to be hung from the pump
Trial Co-ordinator
Damon Bishop (5718)
Equipment Number
P633-4B
Trial Title
Mill Pump Gland Optimisation
Description of Trial
To Increase the gland life of the Mill injection

pumps we are trialling 1830SSP packing.
Components being Trialled
1830 SSP Packing
Old Components
Graphlite Packing
Comments and
Instructions.
If Pump is re-packed please

Retain old packing and deliver to
Damon Bishop or Reactive
supervisor.
Take note of Shaft Sleeve
condition and give feedback.
Please re-pack with 1830SSP
packing. The packing is available
from the Reactive supervisors and
only to be used on this pump.
77
Appendix D
Gland Water Requirements for Slurry Pumps
78
Appendix E
Shaft Deflection Calculations
Specific Speed
Specific Speed Ns = (N x (Q^(1/2)) / (H^ (3/ 4)) / 1.9
N = The speed of the pump in revolutions per minute (rpm.)
Q = The flow rate M /hr ( for either single or double suction impellers) @ B.E.P.
H = The total dynamic head in m's @ B.E.P.
N
Q
H
Ns
811
630
42.9
2307.28775
79
Bending Force Formula

S.G = Specific Gravity of the Liquid
H = Total head @ B.E.P. (Meters)
B2 = Width of Impeller including Shrouds (mm)
D2 = O.D of the impeller (mm)
Kq = 1- (Q/Qn)
K = Radial thrust factor 0.3 to 0.35 (See Chart)
Q = M /hr actually Pumping
Qn= M /hr @ the B.E.P.
S.G
H
B2
D2
Kq
K
Q
Qn
1.824
42.9
150
635
0.989843179
0.38
200
630
P = (Kq x K x H x S.G. x D2 x B2) / 9.81
2857.775408
W = Weight of Impeller Kg's
W
F
250
F = Hydraulic Radial Imbalance, Kg's (In previous)

D = Shaft Diameter mm's
I = Moment of Inertia ( x D^4 / 64)
E = Modulus of elasticity of the shaft material (Kg's / cm)
L = Distance from the impeller centreline to the inboard bearing, mm's
F
D
I
E
L
3107.775408
100.01
491.0702311
2039432
500
Y = Shaft Deflection at the Impeller centreline mm's
P = Kg's of Force generated
3107.775408
Shaft Deflection For Straight Shafts

Y = (F x L) / (3 x E x I)
1.292961362
80
Shaft Deflection for Stepped Shafts

Stepped Shafts
F = Hydraulic Radial Imbalance, Kg's (In previous)

A = Distance from the impeller centreline to the steps of the shaft, mm's
S = Span between bearing centrelines, mm's
Il, Im, In = Moments of inertia of the various diameters, mm's
X= Distance to Stuffing Box Front mm's

D = Shaft Diameters mm's
L = Distance from the impeller centreline to the inboard bearing, cm's

E = Modulus of elasticity of the shaft material (Kg's / cm)
Y = Shaft Deflection at X mm's
F
A
S
Is
Il
Ia
X
Ds
Dl
Da
L
E
3107.775408
200
600
7853.981634
1885.74099
491.0702311
0
200
140
100.01
500
2039432
0.494914524
81
Appendix F
Mechanical Inspection Reports
Kelly & Lewis Pump PM / Condition Report
Work order
Equipment No.
Type
Checked by
Date
Failure / date
History
Root Cause
Pump Criticality
642510
P633-4b
6x8 top discharge
R. Whitham P.Fourie N. Finlay
21/04/10
19/04/10
Gland failure
b
Suction Liner
Casing/Bowl
Impeller
Suction Liner-Impeller Clearance

To be checked before pump is pulled down.
Bearing Lifts
Packing to be removed, drive uncoupled
before tests. Max lift: .008
Seal condition
Pump
Condition (0 60%): 60%
Replaced?
Y/N
y
Replaced?
Y/N
n
Replaced?
Y/N
y
Max: 0.060 Dependent on end float
Adjusted: 0.060
Inboard: 0.003
Outboard: 0.001
Inboard: new style seals fitted to both inboard and
Replaced?
outboard
Y/N
Replaced?
Y/N
Gland Configuration
Please make a note of correct gland
configuration.
Packing Type used
Shaft sleeve
Stuffing Box condition
Outboard:
Stuffing Box
No. of rings2 l/ring 4
Lantern Ring
No. of rings6
1333-g
Replaced? Y/N
y
Size: 124mm
Condition: new
good
82
Stuffing Box Bore size
Size:
Specification: Greater than 1mm on

requires replacement.
Stuffing Box Concentricity about
Shaft Sleeve
Condition: good
A: 20mm
B: 20mm
D
C: 20.5mm
D: 20mm
Gland Water Pressure
10 atu
Strainer Condition
good
Flow restrictor size

Gland Follower condition
12l/min
good
Gland Studs
good
Gland Water
Condition
Supply
Line
new
Scale Muncher
Attrition Plate to Cutter Clearance.

To be checked before pumps pulled
down
Recommended: 5.0mm
Actual:..
Adjusted:
Attrition Plate condition
Replaced? Y/N
Cutter Tool condition
Replaced? Y/N
Stub Shaft condition
Replaced? Y/N
Belt Drive
Pulley Driven
Pulley Drive
Belts condition
Replaced? Y/N
n
Replaced? Y/N
n
Replaced? Y/N
good
good
Retention setting:
Condition: new
y
Guard Condition
83
RIO TINTO ALCAN GOVE PUMP CREW
Pump Repack Report
Equipment No.
Type
Checked by
Date
Failure
Failure date
Operating history
P633-4A
6X8 K&L
R WHITHAM
14/04/10
GLAND
14/04/10
Root Cause
Pump Criticality
Pump run time (from
last failure)
Gland configuration
Packing type
Shaft sleeve dia
Shaft sleeve surface
roughness
Scale build-up in stuffing
box bore
Stuffing box bore dia
Stuffing box
concentricity about shaft
sleeve
PACKING
B
Manufacturers / Alcan data

2 L/RING 4
1333-G
126mm
Measurement / observation
GROOVE UNDER W/PLATE

GOOD
NIL
A: 19.5
B: 19.5
C: 20.00
D:20.00
Water sealing
requirement
Sealing water flow past
throat bush
Gland water pressure
Pump discharge
pressure
Gland water flow rate at
pressure
Strainer condition
Restrictor flow rate
Gland packing run-in
12 L/MIN
10 ATU
GOOD
12 L /MIN
84
period
st
1 hour
nd
2 hour
3rd hour
4th hour
Pump handover
Comments
SHAFT SLEEVE STARTING TO GET GROOVE UNDER WEAR

PLATE,PROBABLY NEED CHANGING OUT IN NEAR
FUTURE,(4-6 WEEKS)
Recommendations
PLAN TO C/OUT SLEEVE
85
Appendix G
Reality Chart Summaries
87
88
89
90
91
92
Appendix H
Letter of Offer concerning Stiffened Shaft Assemblies
KSB Australia Pty Ltd A.C.N. 006 414 642 A.B.N. 29 006 414 642 KSB Australia Pty Ltd
Mobile: 0408 202 585 Email brett.lewis@ksbajax.com.au 22th October,
2009
Rio Tinto Alcan

Gove Operations
Melville Bay Road,
Nhulumbuy NT 0880 Australia
ATTENTION: Damon Bishop
Rotating Equipment Engineer
Dear Sir,
Subject: Mill Injection Pumps
Site: Gove Operations
KSB ref: BL 090013
We thank you for your valued enquiry and take the opportunity to offer a quotation for the above and
have pleasure in providing our proposal in accordance with KSB Australia Pty Ltd Standard Terms and
Conditions of Sale.
1. Introduction
KSB Australia P/L and Georgia Iron Works, or GIW, are both wholly owned subsidiaries of the
international KSB AG Group. KSB Australia P/L/ source product as well as manufacture under licence to
GIW. The success of our organization is the practiced philosophy in that we strive to build partnerships
with our many clients throughout the world, ensuring that productivity is maximized, and in turn
generating increasing profits.
2. Executive Summary / Slurry Data

Current gland life of the LSA 6 x 8/25 Mill Injection Pumps is creating production losses. After looking at
the slurry data / performance operating conditions, we were able to reselect the pumps and start
looking at shaft deflections and gland water pressure etc
The performance and slurry data is follows;
The shaft deflection at the three duties is listed below for the standard 3 15/16 shaft which you are
using. I have also included the shaft deflection of the 3 15/16 stiffened shaft design and you will note
that there is a difference of at least 0.5mm.
366m3/hr Max
340m3/hr Normal
206m3/hr Min
LSA 6x8/25 315/16 Standard

Shaft
LSA 6x8/25 35/16 Stiffened
Shaft Design
0.777mm
0.898mm
0.946mm
0.277mm
0.290mm
0.292mm
KSB Australia Pty Ltd A.C.N. 006 414 642 A.B.N. 29 006 414 642 KSB Australia Pty Ltd
Mobile: 0408 202 585 Email brett.lewis@ksbajax.com.au Although
a shaft deflection of 1mm may be alright in

certain applications, it is not a desirable deflection and we would under normal circumstances be looking
for a better figure. It may also be worth while checking gland water pressures and flow rates to see if
they are as per required. Under Section: 5.2.3.1 Stuffing box of the attached LSATech Book, it states
the required flush and pressures for gland packing. I have included below a general outlined;
Forward Flush Gland Assembly 1.3l/s @ 10PSI Above Maximum Discharge Pressure
KE Low Flow Gland Assembly 0.3l/s @ 10PSI Above Maximum Discharge Pressure
3. Scope of Supply
One (1) LSA 315/16 Shaft Bearing Assembly Stiffened Shaft Design to suit the original pedestal. Part
No 9283D Assembly.
Price $27890.00 each nett, excluding GST / Ex works KSB Tottenham Victoria
Prices quoted are net sell and exclusive of GST. Prices are not subject to rise and fall due to labour and
materials if order placed within the Sixty (60) day validity period.
4. Delivery
We would be able to affect delivery of the equipment within 13 working weeks.
Delivery commences after KSB has received the complete technically and commercially clear written
order.
Should the above delivery be in conflict with your program, we would be happy to discuss your
requirement with our manufacturing group to see if improvement can be made to accommodate your
request.
5. Validity
Our proposal is open for acceptance for a period of sixty ( 60 ) days then subject to our written
confirmation.
6. Payment
100% payable on presentation of invoice thirty (30) days from delivery
7. Warranty
94
Our equipment is guaranteed for twelve ( 12 ) months operation against any manufacturing failure. This
guarantee is limited to eighteen ( 18 ) months after shipping of the equipment, the shortest period being
retained.
We guarantee that the goods manufactured / supplied by KSB Australia shall be of first class materials
and sound workmanship. KSB Australia will make good or replace any defects or defective parts therein,
which under proper use may appear.
8. Quality Assurance
KSB Australia Pty Ltd (Australia) has been accredited with Quality Assurance to Australian Standard 3901
and International Standard 9001. Our Quality manual is available for viewing at our Head Office in
Melbourne, Victoria.
9. Documentation
One (1) complete set of the following documentation will be supplied after official order has been given
to proceed with pump upgrade;
Certified GA Drawing including parts listing
Bill of materials
Operation and maintenance manuals
10. General Comments

Please note that KSB Australia will only comply with customer specifications which have been
commented upon. No other specifications or standards even if mentioned in the body of reviewed
specifications have been taken into account in this offer.
We trust the above and enclosed meets with your approval and we look forward to discussing your
requirements in further detail with you.
We remain at your service.
KSB Australia Pty Ltd
Brett Lewis
Business Development / Technical Support
95
Appendix I
Power Point Presentation to Management Concerning Mill Injection Shaft Upgrade
Mill Injection Pumps

Consist of 4 groups. A, B, C and D Group
A, B and C Group utilise TKL pumps (GIW). These Pumps
have been in service since plant commissioning.
D Group Pumps were installed as part of G3 and utilise 6/8
Warman pumps.
6 August 2010
Presentation Title
Current Pump Status
Days
Days
Isolated incident
not indicative of
the
MTBF
140
180
MTBF
MTBF(Based
(Basedon
onRun
RunTime)
Time)
160
120
140
100
120
80
100
80
60
60
40
40
20
20
0
MTBF
MTBF(Days)
(Days)
Average
Average
Target
Target
P633- P633P633- P633P633- P633P633- P633P633- P633P633- P633P633- P633P6334B

305D
4A
104C
5A
5B
304D
4A
104C
5A
5B
304D
105C 105C
305D
Production Losses due to Gland failure for the Last Year =

1900 Tonnes. (Already 1500 tonnes for this year)
Highest contributor to Production Losses for Pumps
(Second goes to Stage 3 ISC Pumps)
Second lowest MTBF for Glands (First goes to Pressure
Decanter Under Flow Pumps).
Second largest gland water uses 170 l/min (First goes to
Mud Injection in HTD)
6 August 2010
Presentation Title
96
Our current Utilisation of Mill

injection
Mill Injection Utalisation
We need 75% availability from each group to meet plant

requirements!!
85
80
75
70
That means that each individual pump requires 87%

availability to meet its groups target. (.87 X .87 = .75)
65
60
55
50
6 August 2010
Presentation Title
What is the issue? (A, B and C

Group)
We have long suspected Shaft deflection as the root cause

and consequently it has limited our ability to improve the gland
reliability. Mechanical Seals, Crown Bushes and alternate
Packing arrangements have all been trialled with limited or no
success.
This cyclic shaft deflection will

deform the packing which will compromises the stuffing
boxes ability to seal.
Introduce axial loading on Mechanical seals quickly
leading to failure.
Cause the shaft sleeve to contact the pump wear plate or
flow restrictor bush.
Discussion with GIW (pump supplier) they offered to quantify

the deflection by performing a dynamic analysis of the
pumping system.
6 August 2010
Presentation Title
97
Calculations Involved
6 August 2010
Presentation Title
Dynamic analysis results
Under current pumping conditions the impeller exerts a cyclic

radial load of 14122 N on the end of the shaft.
The consequence of this Radial load is a .95mm deflection in

the pumps 104mm shaft.
1.4 Tonnes
.95mm
6 August 2010
Presentation Title
98
Solution (A, B and C Group)

Stiffened shaft which will have greater load carrying capacity then
the existing shaft due to
Utilisation of parallel bore bearings instead of tapered adaptor
sleeves this will enable larger diameter shaft.
Increased diameter mid section which will reduce flexing
Larger diameter sealing journal face
This will reduce deflection to 0.3mm in worst case scenario.
6 August 2010
Presentation Title
10
Whats the pain?
$28,000 per complete Bearing Assembly
12 week lead time
366m3/hr Max
340m3/hr Normal
206m3/hr Min
LSA 6x8/25 3
15/16 Standard
Shaft
0.777mm
0.898mm
0.946mm
LSA 6x8/25 3
5/16 Stiffened Shaft
Design
0.277mm
0.290mm
0.292mm
6 August 2010
Presentation Title
11
Appendix J
99
Capital Project Forms for Mill Injection Shaft Upgrade
100
This page (p. 101) removed at request of author.

17/8/2012
Request for Authorization
Fields with " * " = mandatory depending on the step
RFA No.
Status
506H529
Project in Progress
Title * Mill Injection Pump Shaft Upgrade

RFA Phase
Syspro Number
Current Intervening
EXE Execution
party
Nigel Manwaring
401-423-605
Compiler
Paying Plant
Damon Bishop/Alcan
506
Gove Aluminium Limited
Plant Responsible
Paying Organization *
Paying Department *
01 Refinery Operations
012 Redside - Digestion
Org. Responsible *
Department Responsible *
Manager Paying *
Promoter (Org. Manager) *
Mark Briggs/Alcan
Bob Gordon/Alcan
Manager Responsible *
Manager Org. Responsible
Promoter Representative *
Plant Priority
Mark Briggs/Alcan
1 Critical project (within 12 months)
*
User Representative *
Organization Priority
Program
Project manager
Digestion
Nigel Manwaring/Alcan
Fixed capital
170.3 K AUD
Total
Program Responsible
506 Gove Aluminium

Limited
CPD Capital Projects
010 Capital Projects
Department
John Waller/Alcan
Scott Savage/Alcan
Project forecast
Expense
US $ equivalent
amount
0.0 K AUD
Required if the RFA's

amount is greater than
the plant limit
170.3 K AUD
Fixed Expense
capital
170.3 0.0 K AUD
K AUD
170.3 K AUD
K USD
Non Routine
Sustaining
RFA type *
Sustaining
Project Type *
Target start date *
Mar 31, 2010
Oct 29, 2010
Actual start date
Apr 6, 2010
Target completion
date *
Actual completion
Offensive / Defensive *
2010
Mark Briggs/Alcan
1 Urgent
Nigel Manwaring/Alcan
Total RFA Amount

In thousands
(,000)
1st Year Cost

Summary *
date
102
Attachments
Project nature (What) *
Idea Initiation Form

This project will upgrade the shafts of the mill
injection pumps (A, B and C group) with stiffer
shafts.
CEA
This stiffer shaft will reduce the current shaft
deflection from 1mm to .3mm (refer KSB pumps letter
of offer).
Capital Estimate
This will have benefits of reducing production
losses, maintenance costs and gland water
reduction.
This RFA requests execution funds for purchase of
Chart Of Accounts
shaft assemblies only.

H529PROC - Procurement - $170,340
Installation costs have been agreed to be covered by
site maintenance because the shafts are expected to
be changed out within a 3 month period during
normal maintenance activities.
Justifications (Why) *
Refer Idea definition form for more detail.

NPV
Why now *
KSB Pumps Letter of Offer
Alternatives *
Alt 1
Replace the make of pump with a Warman pump.
Warman pumps traditionally have a lower shaft
slenderness ratio which results in increased gland
reliability. This however would cost 4 to 10 times
more than the current proposed solution.
Alt 2
Buy the shafts at $8000 each (x6) and perform our
own upgrades with our facilities onsite with risk of
possible shaft failures and design refinement period.
Summary of risks *
Contingency *
.
.
Comments *
103
Impact on emissions
Complies with Alcan/local commitements
GHG
tons
SO2
tons
N/A
Quant.
measures
(kw,cm,cfm,psi,et
Constraints
Energy
Services
c)
N/A
Quant.
measures (kw,cm,cfm,psi,etc)
Considerations (check list only - details covered in evaluation below)

N/A
Health & safety *
Environment *
Maintenance LCC *
Key Equipment
Attachments
N/A
N/A
No
104

Mantenimiento de Bomba Horizontal de Lodos Warman

Uploaded by

Copyright:

Available Formats

Mantenimiento de Bomba Horizontal de Lodos Warman

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mantenimiento de Bomba Horizontal de Lodos Warman

Uploaded by

Copyright:

Available Formats

University of Southern Queensland

Faculty of Engineering and Surveying

Pump Gland Improvement Project

Damon Thomas Bishop

In fulfilment of the requirements of

Courses ENG4111 and 4112 Research Project

Bachelor of Engineering (Mechanical)

Submitted: 28th of October, 2010

University of Southern Queensland

ENG4111 & ENG4112 Research Project

Prof Frank Bullen

Damon Thomas Bishop

Rio Tinto Alcan Gove (RTAG)

List of Figures ...

List of Tables and Charts ................

List of Appendices ................. 13

1.1 Outline of the Study.......... 15

1.3 The Problem..

1.4 Research Objectives..

1.5 Conclusions: Chapter 1..

Chapter 2 SAFETY REVIEW..

2.1 Safety Introduction...

2.2 Identify Hazards..

2.3 Identifying Risks...

2.4 Controlling the Risks.

Chapter 3 MILL INJECTION PUMP OVERVEIW..

3.1 Aims of the Chapter...

3.2.3 Purpose of Refinery...

3.2.3 Bayer Process..

3.3 Identification of a Reliability Project..... 24

3.4.2 Pump Makes...

Chapter 4 LITERATURE REVIEW...

4.1 Aims of the Chapter.... 31

4.5 Gland Failure Characteristics. 36

Worn Shaft Sleeves..

Poor Packing Material Selection.....

Incorrect Gland Water Supply....

Incorrect Packing Installation.

Wrong Pump Application...

4.6 Industry Methods in Improving Slurry Pump Gland Reliability...

Grooved Flow Restrictor Bushes

Live Loading on the Glands ..

Grease Purging of the Glands..

Stuffing Box Bearings

Stiffened Shaft Assemblies...

4.7 Previous Projects Concerning the Mill Injection Pumps......

Gland Water Differential Project.......

Installation of a Grooved Flow Restrictor..

Mechanical Seal Installation......

Slurry Packing Trial

Chapter 5 DATA COLLECTION and ANALYSIS...

5.1 Aims of the Chapter.......

5.2 Data Required....

Suitability of Shaft Sleeve.......

Suitability of Packing Selection ..

Suitability of Gland Water Supply... 49

Suitability of Pump To Application....

Mechanical Inspection on Gland Failure ..

Chapter 6 APOLLO ROOT CAUSE ANALYSIS..