International Journal of Scientific Research & Engineering Trends

Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

Analysis of Major Cenrifugal Pumps Failures (With

Application to Irrigation Pumps in Sudan)
Mohamed Yagoub Adam, Hassan Khalifa Osman, M. I. Shukri
Faculty of Post Graduates, Nile Valley University, Atbara, Sudan

Abstract- Centrifugal pumps are one of the most widely used pumps in the world, with a wide range of applications, from the
petroleum industry to the transportation of irrigation water. Despite the extensive use of pumps, few component failures cause
severe degradation of pump performance and increase downtime, which adversely affect production. Therefore, in order to
minimize downtime and improve pump reliability and availability, these pumps must be carefully controlled, diagnosed,
maintained or replaced before a catastrophic pump failure occurs. This study investigated the major faults found in
centrifugal pumps, especially in Sudan's irrigation pumps. In this study, we analyzed the problems faced by pump failures
causing yield loss at the level of all privatization and parastatal pump schemes operating under government license,
government pump stations in the four national government owner agricultural schemes (New Halfa, Rahad, Suki and Gezira
& Managil), by using statistical methods. Results show that, at the technical level of the system, a probabilistic method based
on the evolution of the pump state enables us to carry out preventive interventions; when the component reaches the
degradation zone, it leads to an increase in periodic system intervention. For equipment, whose system schedule set in advance
by the manufacturer cannot effectively meet the maintenance requirements; the mathematical model based on failure
linearization can correct or even optimize the maintenance plan. Through our own investigation, we have concluded that it is
necessary to change the structure of the maintenance cycle of the irrigation pumps under consideration. In this study, the
feasibility of preventive maintenance based on reliability and conditional maintenance was verified. The results obtained are
contributions to meet the reliability goals pursued by irrigation pumps.

Keywords- Major Failures; Centrifugal Pumps; Probabilistic Method; Reliability.

This analyzes lead to determine the failures that greatly

I. INTRODUCTION influence the condition of an item of equipment, and to
implement a suitable maintenance policy in order to limit
Pumps are devices that move fluids or slurries, by the interventions and increase the hours of operation.
mechanical action. Pumps operate by different
mechanisms, and consume energy to perform mechanical Sudan’s irrigation pumps are more strategic at the level of
work moving the fluid. Pumps operate via many energy agricultural schemes, so special attention and analysis are
sources including manual efforts, electrical, thermal, solar, needed to reliably ensure food security for humans and
and wind. Pumps are available in different design and animals, and export crops that can bring hard currency,
sizes to accommodate all uses and to suit all conditions. such as cotton, sesame and other grains.
Pumps are the core of most industrial processes, and the
second most uses machine on the world. This study aims to analyze major mechanical failures and
maintenance methods to improve centrifugal pumps
Since they are so common, pumps are over looked as a availability, in Sudan, by reducing downtime. The main
possible wellspring of improved profitability or a reason objective of this paper is to analyze the causes of irrigation
for overabundance costs if not worked appropriately. Plant pumps failures in the Sudanese government schemes,
chiefs, maintenance specialists, and production supervisor privatization schemes and parastatals schemes, and find
ought to hold best practices and comprehend the solutions to reduce such failures.
operational do's and don'ts for utilizing pumps
appropriately in manufacturing and industrial applications The methodology adopted in this paper included office
[1, 2]. and field works. The data obtained from office and field
was analyzed then accordingly the pumps failures were
In order to improve the availability of production defined, their causes were determined and
equipment in general, it is very important to analyze the recommendations for elimination of these causes were
Reliability, Maintainability, and Availability indicators to given. The statistical Pearson correlation coefficient is
know more about their conditions across their life cycle. used, and the statistical package for social studies (SPSS)

© 2021 IJSRET
632
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

software is used to measure the relationship between the 2.1 Root Cause Analysis (RCA) is: a structured
studied variables. At the same time, a statistical test is investigation designed to determine the real cause of
performed; the SPSS is used to perform a Chi-square test the problem and the necessary measures to eliminate
to test the nature of the probability distribution of it [5]. From a technical point of view, it is usually a
operating time between failures. In addition, Pareto multidisciplinary problem and the necessary measures
diagrams, Weibull analysis, and Ishikawa diagrams are to eliminate it. From a technical point of view, it is
also used to analyze major pump failures. usually a multidisciplinary problem, usually focusing
on traditional engineering fields such as chemistry,
II. PUMP MAINTENANCE AND FAILURE physics, materials, statics, dynamics, and fluids. They
ANALYSIS are easily placed in engineering environments, and the
real root cause may lie in human tissues [4]. There are
several methods available to perform a structured
Maintenance as a process is simply all activities and analysis of the failure. The main goal of all these
resources required to maintain the specified performance methods is to prevent failures, especially those with
and condition of the item within a given period. The serious consequences. The method is divided into five
technicalmeaning of maintenance involves functional tools.
inspections, services, repair or replacement of necessary
devices, equipment, machinery, building infrastructure, 2.2 Pareto Chart: Pareto analysis is an easy-to-use
and supporting utilities in industrial, commercial, technique that helps to choose the most effective
government, and residential facilities. Over time, it has changes. It uses the Pareto principle, that is, by
included a variety of wording describing various cost- completing 20% of the work, you can generate an
effective practices for keeping equipment running; these 80% advantage to complete the entire work. Pareto
activities occurred before or after failure. analysis is a formal technique used to discover the
changes that will bring the greater benefit. This
1. Maintenance Types: method is useful in situations where many possible
Generally, there are two main strategies for performing actions are competing for attention [6]. The Pareto
maintenance functions. The first is the preventive method; chart is a chart that combines a histogram or bar chart
that is, maintenance is performed to prevent malfunctions with a line chart. It groups the frequency or cost of
or breakdowns. The second is the corrective method. different problems to show their relative importance.
Perform maintenance after failure or obvious failure. The bar graph displays the frequency in descending
These two types of maintenance can be further divided, as order, while the line graph displays the cumulative
shown in Figure (1) [3]. percentage or total moving from left to right.

2.3 The Five Whys: The "5 Whys" sometimes called

GembaGembutsu (literally means location and
information in Japanese, much of today's RCA
methods taken from the Japanese automobile
industry)[6]. Five Whys usually refers to the practice
of asking five questions. Each question starts with
why the failure occurred, to find the root cause of the
problem. No special techniques or forms are required,
but the results should be recorded in the worksheet.
Five Whys is it best to use when solving simple RCA.

2.4 Cause and Effect Analysis: Cause and effect

analysis is a graphical method used for failure
Fig 1. Different Maintenance Types.
analysis, called fishbone analysis. The name is
derived from fish-shaped patterns and is used to plot
2. Failure Analysis:
the relationship between various factors that
Despite the best efforts to avoid them, failures are still a
contribute to a specific event [7]. Fishbone analysis
common occurrence in every industry. Failure not only
usually plots the four main categories of potential
means that the technical attempts to design and operate
causes (ie, human, machine, material, and method),
complex systems are imperfect, but it also means that the
but can include any combination of categories.
organization cannot successfully manage the competing
interests of time, quality and money [4]. Therefore, in
2.5 Fault Tree Analysis (FTA): FTA is a top-down
order to continuously improve and learn all the abilities
deductive failure analysis, which uses Boolean logic
from these failures, it is our best interest, which allows us
to combine a series of lower-level events to analyze
to avoid making the same mistakes.
the undesired state of the system. This analysis

© 2021 IJSRET
633
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

method is mainly used in safety engineering and that it is very flexible and can be adapted to different
reliability engineering to understand how the system experimental results [8].
fails, to determine the best way to reduce risks, and to
determine (or perceive) the incidence of safety 1. The Bathtub curve and product failure behavior:
accidents or events at a specific system level failure. Reliability experts usually use a graphical representation
FTA is a graphical technique that provides a system called a bathtub curve to describe the life of a group of
description of the combination of situations that may products. The bathtub curve includes three periods: the
occur in the system, and combines the system and infant mortality period at which the failure rate decreases,
human faults. The fault tree analysis can be displayed followed by the normal life period (also known as the
from left to right or top to bottom. Many professional useful life), the failure rate is relatively low and relatively
software products can perform fault tree analysis, but constant, and concluding with a wear-out period that
both Microsoft Power Point and Visio can be used to exhibits an increasing failure rate. The Bathtub curve and
complete these diagrams [6]. Figure (2) shows fault product failure behavior shown in Figure (3) [9, 10].
tree diagram.

Fig 3. Bathtub curve of component reliability.

2. Determining the failure frequencies:

By classifying all defective parts in ascending order
Fig 2. A Fault Tree Diagram.
according to their life characteristics, Equation (1), can be
2.6 Failure Mode and Effects Analysis (FMEA): Failure used to determine the corresponding failure probability
Mode and Effects Analysis (FMEA) is a design F(i) in a very simple form for the case N < 20, assign a
evaluation process used to identify potential failure rank "Ni" (approximate value of the middle rank) to each
failure[11]:
modes and determine the impact of each mode on system
performance. This process formally records standard
𝑵𝒊 + 𝟎. 𝟑
practices, generates historical records, and serves as the 𝑭(𝒊) = ≈ 𝑭(𝒕) … … … … . . (𝟏)
basis for future improvements. The (FMEA) process is a 𝑵 + 𝟎. 𝟒
series of logical steps, starting with the analysis of lower- Where;
level subsystems or components. The analysis starts from i: Ordinal for sorted defective parts,
the point of view of failure and identifies potential failure N: Reference quantity, e.g. production quantity,
modes and failure mechanisms. Then, trace the impact of
each failure mode back to the system level. According to 3. F (t) defines the failure failing or failure probability,
the probability of occurrence, severity and gratitude, a and can be obtained by Equation (2) [12].
severity level is assigned to each failure mode and
impact. For failures with a higher score on the critical 𝒕−𝜸 𝜷
−[ ]
level, design changes are recommended to reduce the 𝑭(𝒕) = 𝟏 − 𝒆 𝜼 … … … … . … … … . (𝟐)
failure rate [7].
From which;
III. WEIBULL DISTRIBUTION 𝒕 𝜷
−[ ] 1
𝑡 𝛽
( ) 1
1 − 𝐹(𝑡) = 𝒆 𝜼 , =𝑒 𝜂 , 𝑙𝑛𝑙𝑛 ( ) = 𝛽lnt −
1−𝐹(𝑡) 1−𝐹(𝑡)
Weibull distribution is one of the widely used distributions βlnη that is similar to: 𝑌 = 𝛽𝑋 + 𝐴
in technology practice. Swedish scientist Walodi Weibull
(1887-1979) used it in reliability theory and introduced The expression Y = βX + A is the familiar formula for a
this distribution for the first time. Today, the Weibull straight line. By choosing ln (t) as X, the scale on the
distribution is also used in the field of mechanics such as 1
abscissa, and lnln 𝑙𝑛𝑙𝑛 ( ) as y, the scale on the
determining the wind speed distribution in the design 1−𝐹(𝑡)

layout of wind power plants. The advantage of this law is ordinate, the cumulative Weibull distribution can
represented as a straight line.

© 2021 IJSRET
634
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

MTBF is the mean time between failures; MTTR is the

4. The characteristics of Weibull distribution mean time to repair.
parameter:
The most general Weibull distribution or function has V. FAILURE ANALYSIS IN SUDAN
three parameters: IRRIGATION PUMPS
4.1 Shape parameter (β) – it defines the shape of the
distribution [8]. Failures, maintenance methods and management vary
4.2 Scale parameter (ƞ) – it defines the spread of the from site to site. The Blue Nile, White Nile and Main Nile
distribution.It can be obtained by equations (3) and (4) have the same pump types, so although the maintenance
[13]. and management methods are different, their failures and
𝒕−𝜸 𝜷 effects are similar. The Guneid pumping station, located
−( )
𝑹(𝒕) = 𝒆 𝜼 … … … … . … . (𝟑) east of the Blue Nile Bank, north of the city of Roffaa,
irrigates Guneid’s sugar scheme. The pump type is 605/
𝜷 𝒕 − 𝜸 𝜷−𝟏 WPA 2/0023, installed in 1962. The technical
𝝀(𝒕) = [ ] … … … … . … … (𝟒)
𝜼 𝜼 specifications shown in Table (1).
Where;
Table 1. Technical specifications of Guneid units.
λ (t): Failure rate,
R (t): Reliability value
Quantity Discharge Head Motor Pump
4.3 The location parameter (γ): Usually, the location m3/s m power(kw) power
parameter is not used, and the value of this parameter can (kw)
be set to zero. In this case, the Weibull distribution 4 3.5 24 1000 900
parameter will be reduced to only two parameters.
Table (2) shows the recorded failures of pump components
5. Performing a Weibull analysis: from 2010 to 2020.
Weibull plotting is a graphical method used to informally
check the assumptions of the Weibull distribution model Table 2. Number of failures per pump from 2010 to 2020.
and estimate of the two Weibull parameters. In the
Weibull analysis, η is defined as the time when 63.2% of
Cause No. of failures DT
the analyzed systems or components fail (the intersection
of maintainability and Weibull curve, when M (t) = Bearing 65 4680
63.2%) [11, 14]. Shaft sleeve 63 4536
Wear ring 30 1500
IV. AVAILABILITY
Impeller 3 1200
Availability (A) can be described as the ability of the item Gland packing 25 200
to be in a certain state, in this state, it can perform a given None return valve 4 3
function under given conditions at a given point in time or
within a given time interval, assuming the required Table (3) shows continuous wear failures of bearings and
external resources need to provide. Instantaneous sleeves from 2019 to 2020.
availability refers to the probability that an item can
perform the required function under the given conditions Table 3. Failure history of bearings and shaft sleeves.
within a given time, that is, provide the required external
resources, or the probability that a system will be available n Date TBF TTR
for use at any random time (t) after the start of operation
1 25-7-2019 600 11
[15, 16].
2 10-8-2019 360 25
Clearly, availability represents the possibility that a system 3 7-9-2019 640 30
can perform its required functions when called without 4 21-10-2019 1032 45
failure or maintenance. Therefore, availability is not only a 5 30-11-2019 930 80
function of reliability, but also a function of 6 15-3-2020 900 100
maintainability. Inherent availability (Ai), express by
7 20-4-2020 720 48
equation (5) [17]:
𝑴𝑻𝑩𝑭 8 30-5-2020 700 120
(𝑨𝒊) = … … … … … (𝟓) 9 3-7-2020 680 200
𝑴𝑻𝑩𝑭 + 𝑴𝑻𝑻𝑹
Where; 10 5-8-2020 620 110

© 2021 IJSRET
635
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

1. Analysis of the root cause of the failure: 1.1 Results of ABC Curve:
The practical application of the ABC (Pareto) predictive 1.1.1Zone A: As a result, the cumulative number of
analysis method to the type 605 / WPA 2/0023 pumps in failures corresponding to 67.37% accounted for
the failure history record. Use the original data in the 76.04% of the downtime. This is the area of the front
Table (3) to find the major cause of the failure. For the and rear bearings and the shaft protection sleeve, it
application of the ABC method, it is first necessary to has a lot of downtime, so this area is very important.
classify the failures in descending order of the hours of the Therefore, in order to reduce downtime, preventive
failures, as shown in Table (4), and then proceed to the maintenance becomes necessary or even mandatory.
establishment of a Pareto chart, as shown in Figure (4). Maintenance must be carried out in short time
intervals; it includes monitoring of detectable
Table 4. Classification of failures in descending order. parameters such as vibration, temperature, flow, and
pressure. In addition, condition monitoring also
No. of failures

% Cumulative

% Cumulative
Cumulative

Cumulative
Down time

down time allows us to plan and prepare parts that may fail.

down time
Failure
failure
Cause

1.1.2Zone B: This area is not as important as area A, the

number of failures is 15.79%, accounting for 12.37%
of the downtime. This area of replacement of the wear
ring, so preventive maintenance is recommended.
1.1.3Zone C: In this area, 16.84% of failures accounted
34.2105

38.6171

for only 11.59% of downtime. In this case, corrective

4680

4680
65

65

Bearing maintenance is sufficient, because a large amount of

repair time is not required to indicate a failure.
67.3684

76.0459

2. Weibull distribution for the reliability of bearings

4536

9216

Shaft
128
63

sleeve and sleeves (zone A):

According to the Table (3) and using Equation (1),
(because in our case N = 10 , which mean N < 20), the
83.1579

88.4231
10716

calculations and results shown in the Table (5), which

1500

158
30

Wear ring contain TBF classified in ascending order, the rank for
each TBF, and the F (i).
84.7368

98.3249

Table 5. Reliability calculation results.

11916
1200

161

Impeller
3

n TBF TTR N Σni F(i) F(i)%

1 360 25 1 1 0.067 6.731
97.8947

99.9753
12116

Gland
200

186
25

2 600 11 1 2 0.163 16.346

packing
3 620 110 1 3 0.259 25.962
None 4 640 30 1 4 0.356 35.577
12119
190

100

100

return
4

5 680 200 1 5 0.452 45.1923

valve
6 700 120 1 6 0.548 54.8077
7 720 48 1 7 0.644 64.4231
8 900 100 1 8 0.740 74.0385
9 930 80 1 9 0.836 83.6539
10 1032 45 1 10 0.932 93.2692

2.1 Determine the parameters β and ŋ:

𝑡 𝛽
1 ( ) 1
According to Equation (2), =𝑒 𝜂 , 𝑙𝑛𝑙𝑛 ( )=
1−𝐹(𝑡) 1−𝐹(𝑡)
𝛽lnTBF − βlnη, is similar to Y=BX+A.

Select from Table (5) ln (TBF) as X axis, and

1
𝑙𝑛𝑙𝑛 ( ) as y axis, as shown in Table (6), the
1−𝐹(𝑡)
Fig 4. ABC analysis (Pareto).

© 2021 IJSRET
636
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

cumulative Weibull distribution can be expressed as a 720.635−0 3.63

)
straight line, as shown in Figure (2). 𝑅(𝑡 = 𝑀𝑇𝐵𝐹) = 𝑒 −( 799.82 = 0.504,

Table 6. Construction of abscissa (t) and ordinate (Y). The reliability function calculation of each TBF shown in
the Table (7), the relationship between TBF and R(t)
Construction of Construction of ordinate shown in Figure (3).
abscissa (t) (Y)
1 Table 7. Calculation of the reliability function of each
TBF Ln (TBF) F (i) (ln (ln ( ))
1−𝐹(𝑡) TBF.
360 5.886104 0.067308 -2.66384 TBF R(t) F(t) f(t) λ(t)
600 6.39693 0.163462 -1.72326 360 0.637564 0.362436 0.000526 0.000556
620 6.429719 0.259615 -1.20202 600 0.472287 0.527713 0.001498 0.002131

640 6.461468 0.355769 -0.82167 620 0.460624 0.539376 0.001562 0.002323

640 0.449248 0.550752 0.001618 0.002525
680 6.522093 0.451923 -0.5086
680 0.427333 0.572667 0.001701 0.002962
700 6.55108 0.548077 -0.23037
700 0.41678 0.58322 0.001726 0.003196
720 6.579251 0.644231 0.032925
720 0.406488 0.593512 0.001739 0.003442
900 6.802395 0.740385 0.299033
900 0.324571 0.675429 0.001334 0.00619
930 6.835185 0.836538 0.593977
930 0.312622 0.687378 0.001198 0.006748
1032 6.939254 0.932692 0.992689 1032 0.275191 0.724809 0.000712 0.008872

Fig 5. Cumulative Weibull distribution.

From Figure (2), Y = 3.6304 and X – 24.267, which mean

β = 3.6304
A = 24.267 = β ln (η)
24.267
∴ 𝜂 = 𝑙𝑛−1 = 799.82 ℎ𝑟𝑠
3.6304 Fig 6. Reliability curve.

2.2 Determination of MTBF: The graph shows the change in reliability over time (TBF).
From the table Distribution of WEIBULL: values of the As (TBF) increases, the reliability of the pump decreases,
coefficients A and B according to the shape parameter, which explains the degradation phenomenon.
when β = 3.63, we find that A = 0.901 and B = 0.277, but
MTBF = A.+= 0.901 x 799.82+ 0 2.4 Determine the distribution function F (t):
∴ 𝑀𝑇𝐵𝐹 = 720.635 ℎ𝑟𝑠 From Equation (2),
𝑤ℎ𝑒𝑛;
2.3 Determine the reliability function: (𝑀𝑇𝐵𝐹) = 720.635,
According to Equation (3), when t = MTBF = 720.635, 720.635−0 3.63
)
𝑡ℎ𝑒𝑛, 𝐹(𝑡 = 𝑀𝑇𝐵𝐹) = 1 − 𝑒 −( 799.82 = 49.6%

© 2021 IJSRET
637
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

Fig 7. Distribution function curve.

Table (7) shows the calculation of F (t) for each (TBF),

Fig 9. Failure rate curve.
and Figure (4) shows the relationship between (TBF) and
F (t).
2.6.1 Curve analysis: The calculation results show that (β
>1), and λ increases with time (TBF). Referring to the
2.4.1 Curve analysis: The failure function increases as the
bathtub curve, we find that the pump is in normal wear out
time between failures (TBF) increases.
stage.
2.5 Determine the probability density f (t):
2.7 Calculation of MTTR:
Combined with Equations (3) and (4) this gives:
𝑡−𝛾 𝛽
From Equation (3.9),
𝛽 𝑡−𝛾 𝛽−1 −[ ]
𝑓(𝑡) = [ ] 𝑒 𝜂 , So𝑓(𝑡) = 𝜆(𝑡). 𝑅(𝑡),
𝜂 𝜂 𝛴𝑇𝑇𝑅 769
The calculation results are shown in Table (7). The 𝑀𝑇𝑇𝑅 = = = 76.9 ℎ𝑟𝑠
𝑁 10
relationship between TBF and f (t) is shown in Figure (5).
2.8 Calculation of availability (Ai):
From Equation (5),
𝑀𝑇𝐵𝐹 720.635
(𝐴𝑖) = = = 90.36%
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅 720.635+76.9

The study proved in the previous results: β = 3.63 >1, and

the failure rate λ (t) increases with time, and then we infer
that the centrifugal pump studied is in the wear phase. It is
also proved that when the time is equal to MTBF, the
availability rate of the pump is equal to 90.36%, which is
very beneficial to operation. This can be explained by
downtime, which is almost short due to the complexity of
the failure and the use of any material to repair the pump
regardless of the impact on other pump components.

VI. CONCLUSION
Fig 8. Probability density curve.
In this paper the author discusses about the failures that
2.5.1 Curve analysis: According to this curve, we find happen in irrigation pumps, accompanying Sudanese
that the function f(t) increases with time TBF up to time pump station established since the seventies of previous
(TBF = 720 hrs), and after this value, the function f(t) century like Meina and Guneid pumping units. He mainly
decreases with time. concluded that, it is necessary to change the structure of
the maintenance cycle of the irrigation pumps under
2.6 Failure Rate Curve: consideration. It is clear that by using the Bathtub curve to
𝛽 𝑡−𝛾 𝛽−1 forecast the type of failure of the pump type 605/WPA
From Table (5) and using Equation (4), 𝜆(𝑡) = [ ] ,
𝜂 𝜂
2/0023, it is essential to decide the best maintenance
the result of the failure rate corresponds to TBF shown in strategy and estimate the probability of random failure.
Table (7). While Figure (6) represents the failure rate (λ)
as a function of time (TBF).

© 2021 IJSRET
638
 International Journal of Scientific Research & Engineering Trends
Volume 7, Issue 2, March-April-2021, ISSN (Online): 2395-566X

The findings show that Guneid pump type 605/WPA [12] Sagias, N. C., &Karagiannidis, G. K. (2005).
2/0023 reaches the wear stage. In this investigation, the Gaussian class multivariate Weibull distributions:
feasibility of preventive maintenance based on reliability theory and applications in fading channels. IEEE
and conditional maintenance was verified for components, Transactions on Information Theory, 51(10), 3608-
such as; bearings, gland packing and gearbox. The results 3619.
obtained are contributions to meet the reliability goals [13] Stenström, C., Parida, A., & Kumar, U. (2016).
pursued by irrigation pumps. Measuring and monitoring operational availability of
rail infrastructure. Proceedings of the Institution of
The author recommends that, the maintenance department Mechanical Engineers, Part F: Journal of Rail and
should have a database containing computerized files of Rapid Transit, 230(5), 1457-1468.
all pumps and their characteristics in the schemes. This [14] Hoseinie, H., Ghodrati, B., &Hosseini, A. (2015).
allows quick access to pump documentation during Availability analysis of drum shearer machine;: A
failures. Case Study. In International Mining Congress of
Turkey: 14/04/2015-17/04/2015 (pp. 913-918).
REFERENCES [15] TMMOB MadenMuhendisleriOdasi.Kauhanen, J. T.
(2017). Spare parts inventory management for the
[1] Robison, A. B. (1940). Some notes on power pumps. next generation Finnish Defense Force fighter fleet.
University of Melbourne, Victorian Institute of Naval Postgraduate School Monterey United States.
Engineering, Victoria Australia.
[2] Karassik, I. J., Messina, J. P., Cooper, P., &Heald, C.
C. (2001). Pump handbook (Vol. 3). New York:
McGraw-Hill.
[3] Bengtsson, M. (2004). Condition based maintenance
systems: an investigation of technical constituents and
organizational aspects. Eskilstuna, Sweden:
Mälardalen University.
[4] Ransom, D. L. (2007). A Practical Guideline For A
Successful Root Cause Failure Analysis. In
Proceedings of the 36th Turbomachinery Symposium.
Texas A&M University. Turbomachinery
Laboratories.
[5] Andersen, B., &Fagerhaug, T. (2006). Root cause
analysis: simplified tools and techniques. Quality
Press.
[6] Vorley, G. (2008). Mini guide to root cause analysis.
Quality Management and Training, London.
[7] Mobley, R. K. (1999). Root cause failure analysis.
Butterworth-Heinemann.
[8] Pobočíková, I., &Sedliačková, Z. (2014). Comparison
of four methods for estimating the Weibull
distribution parameters. Appl Math Sci, 8, 4137-4149.
[9] Wilkins, D. J. (2002). The bathtub curve and product
failure behavior. Reliability HotWire, 21(NOV).
[10] Vagenas, N., Runciman, N., & Clément, S. R. (1997).
A methodology for maintenance analysis of mining
equipment. International Journal o/Surface Mining,
Reclamation and Environment, 11(1), 33-40.Scholz,
F. (1996). Plotting on Weibull paper. Research and
technology. Boeing Information & Support Services.
[11] Sheikh, A. K., Younas, M., & Al-Anazi, D. M. (2002,
December). Weibull analysis of time between failures
of pumps used in an oil refinery. In The 6th Saudi
Engineering Conference, KFUPM, Dhahran (Vol. 4,
pp. 475-491). Singh, D., &Suhane, A. (2013).
Availability of Centrifugal Pump on the Basis of
Weibull Analysis. International Journal of Science
and Research, 4(6), 378-380.

© 2021 IJSRET
639