ArticlePDF Available

Reliability Centered Maintenance Plan for the Utility Section of a Fertilizer Industry: A Case Study

Authors:
  • COMSATS Institute of Informaton Technology, Wah Catt, Pakistan

Abstract and Figures

This paper provides an insight to the application of Reliability Centered Maintenance (RCM) and life data analysis. The objective is to formulate a Failure Management Policy for the utility section of a fertilizer industry. It focuses on how key elements of the RCM process can be combined to select appropriate policies for managing a system's failure modes and their consequences, using an RCM decision algorithm. The foremost objective of Reliability Centered Maintenance plan is to optimize reliability of physical assets while being cost effective. Furthermore in addition to traditional RCM, life data of the machinery is analyzed using empirical analysis of Weibull distribution. The estimates of the parameters of Weibull distribution are found using Median Rank Regression (MRR) method. Each machine is then matched with one of the six most widely accepted patterns of failure. In the end these two techniques i.e. RCM and Weibull Analysis are integrated in a manner to formulate an optimized Failure Management Policy. The plan is in compliance with RCM SAE JA1011 standard.
Content may be subject to copyright.
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
9
Reliability Centered Maintenance Plan for the Utility
Section of a Fertilizer Industry: A Case Study
Muhammad Abid*, Suleman Ayub, Humza Wali, Muhammad Najam Tariq
Faculty of Mechanical Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences & Technology, Topi, 23640,
Swabi, KPK, Pakistan
*abid@giki.edu.pk
AbstractThis paper provides an insight to the application of
Reliability Centered Maintenance (RCM) and life data
analysis. The objective is to formulate a Failure Management
Policy for the utility section of a fertilizer industry. It focuses
on how key elements of the RCM process can be combined to
select appropriate policies for managing a system’s failure
modes and their consequences, using an RCM decision
algorithm. The foremost objective of Reliability Centered
Maintenance plan is to optimize reliability of physical assets
while being cost effective. Furthermore in addition to
traditional RCM, life data of the machinery is analyzed using
empirical analysis of Weibull distribution. The estimates of
the parameters of Weibull distribution are found using
Median Rank Regression (MRR) method. Each machine is
then matched with one of the six most widely accepted
patterns of failure. In the end these two techniques i.e. RCM
and Weibull Analysis are integrated in a manner to formulate
an optimized Failure Management Policy. The plan is in
compliance with RCM SAE JA1011 standard.
Keywords: RCM, Reliability Engineering, Failure
Management Policy, Life data analysis.
I. INTRODUCTION
Utility plant is a very important and critical section of a
Fertilizer Industry. The utility plant provides services to the
ammonia/urea complex and other areas. The main products
of the company are ammonia and urea. This paper aims at
formulating a maintenance plan based on RCM
methodology for the utility plant machinery. Application of
RCM can improve reliability and availability while
minimizing the downtime and maintenance cost. Reliability
centered Maintenance is “a process used to determine what
must be done to ensure that any physical asset continues to
do what its users wanted it to do in its present operating
context” [1]. RCM philosophy employs preventive
maintenance, predictive maintenance (PdM), real-time
monitoring (RTM), run-to-failure (RTF) and proactive
maintenance techniques, in an integrated manner to increase
the probability that a machine or component will function
in the required manner over its design life cycle with a
minimum of maintenance [1]. Although it has been
established that we need to implement either of the above
mentioned five maintenance techniques but the real
challenge is to find the correct and the most effective
combination of each, for different equipment, with different
failure curves. In this paper we have taken Condensate
recovery and boiler feed water system as a working
example, deductions and results obtained can be applied to
the all the systems of ammonia/urea complex plant.
II. THEORETICAL BACKGROUND
A. Failure Curves
Failure rate or hazard rate is defined as probability
density function divided by reliability [2]. There are a total
of six failure curves associated with most of the equipment
installed in an Industry which are shown in Fig. 1. The
shape of the failure curve allows us to identify whether the
failure mode was an early life or infant mortality stage
failure, a randomly induced failure or due to wear out and
aging. These curves are formed by plotting failure rate of
equipment with respect to time.
Figure 1 Six Failure Patterns as Identified by Nolan and Heap. [3]
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
10
B. Weibull Distribution
Weibull Distribution is one of the most widely
incorporated distribution which is used to carry out life data
analyses in reliability engineering. .It’s probability density
function (PDF) or f(t) is given by [4]:
 
(1)
Where:

“β” is the Weibull shape parameter which determines
the shape of the Weibull plot. “α” is the scale parameter
which represents the characteristic life at which 63.2% of
the population is expected to fail. "γ" is the location
parameter or failure free life. “t” is the time.
The Weibull Distribution is said to be two-parameter
distribution if γ=0 [5].
Keeping all the parameters constant and varying β can give
different failure curves [6].
β<1 Early life or infant mortality stage failure.
β=1 Constant failure rate curve.
β>1 Wear out or failure due to age curve.
III. RELIABILITY CENTERED MAINTENANCE
For an effective RCM plan the following tools and
techniques were applied.
A. Selection of system
System selected for the implementation of RCM plan is
Utility Section of a Urea Plant.
B. Selection of subsystem
Following Subsystems from the plant are selected:
1) Condensate Recovery and boiler feed water
system.
2) Electricity Generation System.
3) Fire Water System.
4) Plant and Instrument Air system.
Condensate Recovery and boiler feed water system is
selected as a case study for this paper. Same steps and
maintenance decisions combinations can be applied to the
rest of the systems. Fig. 2 shows the functional block
diagram for condensate recovery and boiler feed water
system.
C. Acquiring the failure data
Complete failure data was acquired from the plant using
the Enterprise Resource Planning (ERP) software AS 400,
produced by IBM®.
D. Identification of functions
The function of the Condensate recovery and feed water
system is to supply high-pressure water to the boiler. The
feed water is physically and chemically de-aerated, further
preheated in steam Boiler feed water heater and distributed
to the process steam Boilers and miscellaneous users.
Under normal operating conditions two of the three
50% capacity High pressure Feed water Pumps (P402 A B
and C) take suction from the de-aerator storage section and
pump feed water to the Steam Boiler Feed water Heater (E-
403 , which is not a part of this system).
The System automatically maintains the proper flow to
the boiler drum. A minimum head of 1850 ft. is required
from the pumps rated at 615 gpm, these pumps have an
initial capability of 1950 ft., and 100 ft. is the margin for
deterioration after which the pump will be repaired or
replaced.
Figure 2 Condensate Recovery and Boiler Feed water system Schematic diagram [7].
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
11
The boiler feed water system is a flow process in
compliance with ISO 9001. This system operates for 24 hrs.
per day and has a redundant pump (P402 C). All spare
parts of a pump are available in the warehouse [7].
E. Selection of Critical Equipment
Following critical equipment were selected.
a) Pump (P-402A)
b) Pump (P-402B)
c) Pump (P-402C)
d) Turbine (TP-402AB)
e) Motor (MP-402C)
Criticality of the equipment was based on the
functionality and importance of the functions performed by
the above mentioned equipment. P-402 A and B are run by
TP 402AB which is also as critical as P402 A and B. Pump
P402 AB and TP 402AB are the most critical and work all
the time to deliver the feed water, while the P 402C is a
standby pump and is not as critical as P402 A/B. MP 402C
is used to run P 402C in standby operations.
F. Actuarial Analysis of the failure Data
The Weibull cumulative distribution function (CDF),
denoted by Fw (t) is given by:

(2)
Since for our case the failure free life γ=0 [5] therefore the
analysis will become two-parameter Weibull analysis:

(3)
And the Reliability is given by:


(4)
The linear form of CDF equation is given by:



(5)
For the estimation of Weibull parameters there are two
widely used methods to calculate Cumulative Density
Function which are given below:
a) Median Rank Method using Benard’s Formula [8].
b) Kaplan-Meier Estimation.
Since Kaplan-Meier estimation requires a large data size
to give a useful plot therefore Median rank method using
Benard’s Formula is utilized in our analyses [9].
Fw(t) can be estimated by using Benard’s Formula [8]
for median rank estimator which is given by:


(6)
Weibull parameters can be estimated by the following
methods:
a) Maximum likely hood estimation (MLE).
b) Rank regression on X.
c) Rank regression on Y.
Rank regression on X is used for estimation since MLE
is used for censored data whereas the failure data obtained
for critical equipment is complete data and regression
generally works best for small data sizes (10-11 samples in
our case). Rank regression on Y is not used because
uncertainty is in time to failure which changes with sample
to sample and it is on x-axis of the plots.
Figure 3 Condensate Recovery and Boiler Feed water system functional block diagram
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
12
Therefore Regression on X is used which is given by
[9]:



(7)
Comparing the equation with the equation of Straight
line i.e.
; 
;
; 









Hence

As β (shape parameter) is less than 1 it suggest an early
failure i.e. infant mortality. Hazard rate decreases
exponentially right from the start as shown in Fig. 5.
G. P-402A and P-402B Analysis
For pump 402A β (Shape parameter) is 0.78 and α
(Scale parameter) is 577 days which is shown in Fig. 4.
Hazard rate is very high in the beginning due to infant
mortality and decreases with time. P402B follows the same
distribution and model as of P402 A and has a β = 0.85 and
α =338 days.
Similar results were found of P402B.
H. TP-402A/B Analysis
For Turbine TP402A/B β is 3.58 and α is 1662 days as
shown in Fig. 6. As β is greater than 1 it suggest failure due
to wear out. Hazard rate increases exponentially, at
approximately 500 days as shown in Fig. 7. The reliability
of TP 402A/B decreases exponentially after 400 days,
hence scheduled restoration task at approximately after 400
days is recommended as shown in reliability curve Fig. 8.
I. P402-C and MP402-C Analysis (Standby System):
For Standby systems we perform failure finding tasks.
For these tasks failure finding interval (FFI) [10] is
determined which is given below:
For Standby Systems an availability of 90% is required
therefore unavailability is 10%.
From the given data MTBF = 100 days

(8)

Figure 4 shows the Probability-Weibull plot. Shape and scale parameters for P-402A are estimated
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
13
Figure 5 shows the failure rate or hazard rate plot for P-402A
Figure 6 shows the Probability-Weibull plot. Shape and scale parameters for TP-402A/B are estimated.
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
14
Figure 7 shows the failure rate or hazard rate plot for TP-402AB.
Figure 8 shows the reliability plot for TP-402A/B.
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
15
TABLE I. FAILURE MODES AND EFFECT ANALYSIS.
J. Failure Modes and Effect Analysis
An RCM process that conforms to SAE JA1012 states,
“All the failed states associated with each failure shall be
identified” [10].
A failure mode could be defined as any event which is
likely to cause an asset (or system or process) to fail. A
system can fail for many reasons, for the complete plant
these failure modes or reasons can rise into thousands. For
our system we have listed down the following main causes
that prevent the system from performing its complete
function. These causes could be drilled down further into
the function failures of sub systems, but we have to decide
on the level which is practically feasible for a particular
industry.
Once the failure modes are identified it then becomes
possible to assess the root cause of these failures and decide
on the relative maintenance techniques and intervals.
Function failure that represents total failure can be
identified easily, for our system a total failure will take
place when either of pumps P402 A/B or P402 C fails.
Partial failures may be many for example if only one
pump fails, turbine fails of motor fails which is shown in
Table 1.
K. RCM tree Diagram and Decision Worksheet:
RCM decision diagram (Fig. 9) integrates all the
decision processes into a single framework. The RCM
decision worksheet, Table 2 is filled with the help of RCM
decision diagram and Table I (Failure modes and effect
analysis). Furthermore the proposed tasks and interval are
selected by considering Actuarial analysis of RCM
methodology.
From Table 1 it can be seen that equipment with early
life failure curve are put on increased monitoring directly
after commissioning and keeping a spare nearby may be
helpful until the cause of failure is found and resolved. Root
cause analysis of the equipment itself is recommended to
find out the cause and eliminate it. For equipment with
wear out failure a Scheduled restoration task is suggested so
that potential failure could be avoided and hence making
the whole maintenance cost effective. For Standby
equipment a failure finding interval of 20 days is
determined.
IV. CONCLUSION
All the techniques of maintenance including preventive,
predictive and proactive techniques are currently being
applied in the fertilizer plant under study. A new approach
has been used in which RCM is integrated with life data
analysis in order to accurately estimate the failure mode
followed by each component of the system. Using this
technique a better failure management policy is developed
keeping in view the health of each equipment. This RCM
plan helps to optimize reliability of the system while being
cost effective and decreasing the system downtime. In
addition determination of maintenance tasks and their
intervals was achieved by the use RCM as shown in Table
II.
Function (F)
Functional Failure (FF)
Failure Mode (FM)
(Loss of Function)
(Cause of Failure)
1
Condensate
recovery system
collects
condensate and
returns it to feed
water system,
which distributes
feed water to
process steam
boilers and misc.
users.
A
Failed to supply
high pressure
water to the
boiler during
startup, normal
& emergency
operations by
achieving a
differential head
of 1950ft.
1
P402A fails
2
P402B Fails
3
TP402 A/B
Fails
4
P402C Fails
(Standby)
5
MP402C
Fails
(Standby)
Figure 9 shows RCM Decision Diagram
International Journal of Science and Advanced Technology (ISSN 2221-8386) Volume 4 No 3 March 2014
http://www.ijsat.com
16
V. ACKNOWLEDGMENT
The authors acknowledge Ikram.Ahmad Khan, Dawood
Hercules Fertilizers Limited for all cooperation for
completing this study.
VI. REFERENCES
[1] J. Moubray, Reliability-centered maintenance., Vols. ISBN 978-
0831131463., New York: Industrial Press, 1997.
[2] P. MacDiarmid and S. Morris, Reliability Toolkit (Commercial
Practices ed.), Rome; New York: Reliability Analysis Center and
Rome Laboratory..
[3] F. S. Nowlan and H. F. Heap, Reliability-Centered Maintenance,
San Francisco: Dolby Access Press, 1978.
[4] R. B. Abernathy, The New Weibull Handbook, Houston: Dr. Robert
B. Abernethy, 2004.
[5] C.-c. Liu, "A Comparison between the Weibull and Lognormal
Models used to Analyze Reliability Data. P.hd thesis," University of
Nottingham, UK, Nottingham, 1997.
[6] J. F. Lawless, Statistical Models and Methods of life, New York: J
WILEY & SONS, 1982.
[7] Dawood Hercules Fertilizers Ltd., Utility Plant Training Manual,
Lahore: Technical Training Department, 2009.
[8] M. S. K. A. H. P. G.R Pasha, "Empirical Analysis of The Weibull
Distribution," Journal of Statistics, vol. 13, no. 1, pp. 33-45, 2006.
[9] NIST/SEMATECH, e-Handbook of Statistical Methods, 2005.
[10] S. International, Evaluation Criteria for Reliability-Centered
Maintenance (Rcm) Processes- General requirements. SAEJA1012,
2002.
Failed to supply high pressure water to the boiler during startup, normal & emergency operations by achieving a differential
head of 1950ft
p
Consequence Evaluation
H1
S1
O1
N1
H2
S2
O2
N2
H3
S3
O3
N3
Default Action
Proposed Task
Interval
Can be done by
F
FF
FM
H
S
E
O
H4
H5
S4
1
A
1
Y
N
N
N
Y
N/A
N/A
N/A
N/A
N/A
Scheduled On-condition task
1 day
Technician
1
A
2
Y
N
N
Y
Y
N/A
N/A
N/A
N/A
N/A
Scheduled On-condition task
1 day
Technician
1
A
3
Y
N
N
N
N
Y
N/A
N/A
N/A
N/A
Scheduled restoration task
400 days
Technician
1
A
4
N
N/A
N/A
N/A
N
N
N
Y
N/A
N/A
Schedule Failure finding task
20 days
Technician
1
A
5
N
N/A
N/A
N/A
N
N
N
Y
N/A
N/A
Schedule Failure finding task
20 days
Technician
TABLE II. RCM DECISION WORKSHEET TABLE
... We expect the current framework can be applied specifically in emergencies, considering also that a high percentage (89% according to [17], [18] ) of equipment failure occurs due to random reasons. ...
Article
Full-text available
In manufacturing, biologicalisation defines the analysis of biological patterns as a source of inspiration to model intelligent manufacturing systems. This analysis is highly desirable as an answer to the increasing complexity of modeling current engineering solutions which are required to be self-organized, cooperative, and autonomous. Building on this line of research, our paper introduces a framework inspired by the notion of "reciprocal altruism" observed in species like vampire bats. The goal of this framework is twofold; first, to showcase the bio-inspired methodological guidelines of altruism in a manufacturing context. Second, to foster cooperative behaviors among its constitutive resources (equipment, machine, or robot that is capable of autonomously performing a manufacturing task). The core idea revolves around the assignment of two roles: altruistic/donors and recipients. In this context, altruistic/donors are individuals willing to share their resources or capabilities with those in need, even at the potential cost of their own fitness. We believe that this concept has versatile applications in various manufacturing scenarios, ranging from peer-to-peer energy sharing among mobile robots to load sharing and even tool sharing. In our work, we instantiate the control logic, functionalities, and proof of concept of this altruistic approach, focusing on tool-sharing as an illustrative application. Our preliminary results demonstrate an improvement when comparing the altruistic approach with a manual one. This improvement is particularly evident when considering potential production downtime and production rate fluctuations caused by tool defects. These findings underscore the tangible benefits that bio-inspired solutions can offer in addressing the ongoing challenges of smart manufacturing, especially in terms of engineering design.
... They state that "using this technique a better failure management policy is developed keeping in view the health of each equipment. This RCM plan helps to optimize reliability of the system while being cost-effective and decreasing the system downtime" [8]. ...
Article
Full-text available
The period from the 1990s has seen a large body of work in the areas of Plant Maintenance and Reliability Management. While strategies like RCM and TPM are popular and well known, there is also a large body of work that has attempted to move forward from these methodologies. The main intent of these approaches has been to address the issues faced during implementation of these methods. The newer methods of Plant Maintenance and Reliability Management are not as well known as the conventional methods. This paper attempts to collate and organize this body of knowledge in the form of a comprehensive review.
Conference Paper
This paper is presenting a deep understanding of the application of Reliability Centered Maintenance (RCM) and cost effective data analysis. The objective is to provide a reliability assessment of the four water process pump. It concentrates on how key elements of the RCM process can be amalgamated with appropriate strategies and policies for managing a system's failure modes and their consequences, using an RCM decision making procedure. The most prominent objective of Reliability Centered Maintenance is to optimize reliability of assets with cost effective. Furthermore, life cycle of the system is analyzed using Weibull distribution algorithm. The Median Rank Regression method is used to estimate the parameters of Weibull distribution. The system is then matched with one of the six most widely accepted patterns of failure. The plan is in compliance with RCM SAE JA1011 standard.
Article
Full-text available
In this paper we present the comprehensive analysis for complete failure data. The aim of this research is the Empirical Analysis of the Weibull Distribution for Failure Data. We access the Weibull distribution assumptions of a data set. Median rank regression (MRR) for data-fitting method is described and goodness-of-fit using correlation coefficient is applied. We use the simulation technique to present confidence bound.
Article
Reliability-centered maintenance is a process used to determine - systematically and scientifically - what must be done to ensure that physical assets continue to do what their users want them to do. Widely recognized by maintenance professionals as the most cost-effective way to develop world-class maintenance strategies, RCM leads to rapid, sustained and substantial improvements in plant availability and reliability, product quality, safety and environmental integrity. The author and his associates have helped users apply RCM and its more modern derivative, RCM2, on more than 700 sites in 34 countries. These sites include all types of manufacturing (especially automobile, steel, paper, petrochemical, pharmaceutical, and food manufacturing), utilities (water, gas, and electricity), armed forces, building services, mining, telecommunications, and transport. This book summarizes this experience in the form of an authoritative and practical description of what RCM2 is and how it should be applied. This book will be of value to maintenance managers, and to anyone else concerned with the reliability, productivity, safety, and environmental integrity of physical assets. Its straightforward, plant-based approach makes the book especially well suited to use in centers of higher education.
Article
This book explains basic concepts, principles, definitions, and applications of a logical discipline for development of efficient scheduled (preventive) maintenance programs for complex equipment, and the on-going management of such programs. Such programs are called reliability-centered maintenance (RCM) programs because they are centered on achieving the inherent safety and reliability capabilities of equipment at a minimum cost. A U.S. Department of Defense objective in sponsoring preparation of this document was that it serve as a guide for application to a wide range of different types of military equipment. There are essentially only four types of tasks in a scheduled mainenance program: (1) Inspect an item to detect a potential failure; (2) Rework an item before a maximum permissible age is exceeded; (3) Discard an item before a maximum permissible age is exceeded; (4) Inspect an item to find failures that have already occurred but were not evident to the equipment operating crew. A central problem addressed in this book is how to determine which types of scheduled maintenance tasks, if any, should be applied to an item and how frequently assigned tasks should be accomplished. The use of a decision diagram as an aid in this analysis is illustrated. The net result is a structured, systematic blend of experience, judgment, and operational data/ information to identify and anlayze which type of maintnence task is both applicable and effective for each significant item as it relates to a particular type of equipment.
A Comparison between the Weibull and Lognormal Models used to Analyze Reliability Data. P.hd thesis
  • C.-C Liu
C.-c. Liu, "A Comparison between the Weibull and Lognormal Models used to Analyze Reliability Data. P.hd thesis," University of Nottingham, UK, Nottingham, 1997.
Utility Plant Training Manual
Dawood Hercules Fertilizers Ltd., Utility Plant Training Manual, Lahore: Technical Training Department, 2009.