JJMIE

Volume 12 Number 1, June. 2018

ISSN 1995-6665
Pages 59 - 67
Jordan Journal of Mechanical and Industrial Engineering

A Comprehensive Model of Reliability, Availability, and

Maintainability (RAM) for Industrial Systems Evaluations
Mohamed F. Alya, Islam H. Afefya, Ragab K. Abdel-Magiedb, Eman K. Abd Elhalimc*
a
Industrial Engineering Department, Faculty of Engineering, Fayoum University, Egypt.
b
Mechanical Engineering Department, Faculty of Engineering, Beni-Suef University, Egypt.
c
Industrial Engineering Department, Faculty of Engineering, Misr University for Science and Technology (MUST), Egypt.

Received 29 April 2018 Accepted 21 June 2018

Abstract

Reliability, availability, and maintainability are considered as a crucial metrics that are used to evaluate the performance
of the industrial systems. In this work, an integrated reliability, availability, and maintainability (RAM) model of the 3-out-
of-4 system was proposed to quantify the values of RAM indices and to identify the most critical equipment which mainly
affects the system performance. The Markovian approach was adopted to model the system behavior. A transition diagram
for the proposed model was constructed and differential equations of the proposed model were formulated to obtain the state
probability. The availability at steady state, reliability at transient state and maintainability were analyzed and investigated.
The proposed model was verified and validated. A real data of industrial system in Oil and Gas Egyptian Company was
applied to validate the proposed model and the effect of failure and repair rates at different mission time was presented and
discussed. The results of the applied proposed system revealed that the system availability at steady state is 99 %, the system
reliability is 0.59%, and the system maintainability is 0.99%. On the other hand Turbine no. three was found the most
critical item in the system and need more attention to improve the system performance. It could be said that the proposed
model is considered an excellent tool for industrial systems performance evaluations.

© 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved
Keywords: Reliability; Availability; Maintainability; Markov approach.

RAM model applied to a process industry using the

1. Introduction Markovian approach in steady state. Parametric
computations and indices of RAM to assess system
RAM is an important performance metrics in system performance in repairable industrial systems using Genetic
analysis and considered a good starting point for system Algorithm (GA) and the Markovian approach were
improvements [1]. The Oil and Gas industry has presented [6, 7].
consistently increased its requirements, combined with the Evidently, the integration of reliability, availability and
rise in technological systems, and increased maintainability of investigation tends to good results.
competitiveness of service providers to implement However, reliability, Availability, and maintainability
adequate management strategies for these systems to were investigated individually or two of them in the other
improve their availability and productivity to comply with industries. Much effort has been made by the researchers
those most demanding standards. One important point in providing performance modelling and availability analysis
this regard is to have knowledge about the RAM of the applied on different industrial systems as Paper Plant,
main equipment in this industry [2]. Evidentially, a fault- Paint, and thermal power plant Industry [8-10].
based (Breakdown) maintenance system in the Oil and Gas Aggarwal, et. al. [11] presented a model using Markov
industry is a costly and time-consuming process, resulting birth-death process, the concept of fuzzy reliability and
in a substantial and intangible loss to system operators. availability. A numerical method with the assumption that
Corvaro, et al. [3], assesses operational performance of the failure and repair rates of each subsystem follow the
reciprocating compressors used in Gas and Oil industry exponential distribution has been developed. In which a
using RAM model. The study aimed to evaluate the effect mathematical modeling of the system is carried out using
of different factors related to RAM and devoted to the mnemonic rule to derive Chapman–Kolmogorov
collaborating with the private sector aiming continuous differential equations and solved it by Runge-Kutta fourth-
quality improvement. Aoudia, et al. [4], studied the order method Which is considered one of the most
economic impact of maintenance management common methods that are used in solving the differential
ineffectiveness of one of the main industrial plants of a equations as well as it is considered one of the oldest and
major Oil and Gas group. Sharma and Kumar, [5] built a the best method in numerical analysis moreover; it

* Corresponding author e-mail: [email protected].

60 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665)

provides a popular way to solve the differential equations. analyzed and investigated. The differential equations are
When the system includes a large numbers of differential solved using Rung-Kutta method with aided of MATLAB
equations, MATLAB program could be used to solve these software to get the system availability at steady state (Ass)
large equations. For this, the MATLAB software is and solved by Laplace transform to get reliability at
considered one of the multi programs that could be used transient state. The second is to apply the proposed model
for numerical computations. for the performance measure of a real case of Oil and Gas
Lin, et al. [12] presented a reliability study using both industrial system. This model provides results for a
classical and Bayesian semi-parametric frame-works, they complete reliability, availability, and maintainability
illustrated how a wheel- set’s degradation data can be (RAM) analysis utilizing data sets from a production
modeled and analyzed to ease the calculation of system system in an Oil and Gas plant. A parametric investigation
reliability during applying preventive maintenance. Singh, of various values of system failure rates and repair rates on
and Goyal, [13] developed methodology to study the system reliability (Rs), availability, and maintainability and
transient behaviour of repairable mechanical biscuit their effects on the system performance are presented. The
shaping system on a biscuit manufacturing plant for results of that analysis help the designers/engineers and
determining the availability of the system based on managers to quantify and measure the system
Markov modelling. The differential equations have been performance; conversely, suitable maintenance
solved using Laplace Transforms. Laplace Transform policies/strategies can be selected to enhance the
commonly used in the transient state to obtain the state productivity of the plant.
probabilities, in which the differential equations are
converted to algebraic equations to simplify the system Start
solution.
The K-out-of-N system is the most important type for
the repairable system according to reliability theory and, it Construct the transition diagram of
is used in many applications such as petroleum industry. the 3-out-of-4 system
An investigation of a 2-out-of-3 system has been presented
recently in published work in which the reliability and
availability have been evaluated and analyzed for the
Write the system availability,
system using Kolmogorov’s equations and applied on Part.1 reliability, and maintainability
some particular cases. In this analysis, mean time to (RAM) equations
system failure (MTSF), steady-state availability, busy
period and profit function were derived to evaluate the
system reliability and availability [14-15]. Preeti, [16] Solve the system equations by
presented an analysis which considered as a powerful tool software program and get the values
to analyze reliability of a linear consecutive 2-out-of-3-F of RAM
system with common cause shock failure in which the
transient equations of the reliability and steady equations
Implement the proposed model on a
of the availability have been investigated. Yusuf, [17]
real case study
evaluate the system reliability indices of a repairable 3-
out-of-4 system with preventive maintenance involving Part.2
four types of failures using Kolmogorov equations.
Apparently, the literature review -up to our reading- Present a parametric investigation to
revealed the following points: evaluate the system criticality
 The researchers concentrated, to more extent, on the
investigation of availability and reliability of the
industrial systems.
 Oil and Gas industry need more attention to improve End
and maintain its system performance.
 A little attention is paid to investigate the integration of Figuer1: Steps of the presented work.
RAM for different industrial application.
 No more work on studying of RAM analysis for a 2. System Description
multi-component system such as 3-out-of-4 system.
Moreover, the applications of RAM analysis as an In this section, the 3-out-of-4 system is described. The
adopted approach for maintenance policies for Oil and Gas system consists of four units in which one unite is standby
Industrial systems could be proposed and applied for (sb) and the other three units must be in the operating state
increasing customer satisfaction, reduce the frequency of (o) for the system to work. The system failed (F) when two
failures and maintenance costs. This is a motivation of the units failed and the other two units are in good state (g).
present work. Based on Markov assumption [18], differential equations
The aim of this work is to develop a comprehensive that describe the proposed system are written to analyze
RAM model for industrial systems evaluation. This study the probability for each state. These equations are further
has two main parts, as presented in Fig.1; the first is to solved for determining the RAM indices. The states of the
develop 3-out-of-4 system RAM model based on the system according to Markov are shown below in Table (1),
Markovian approach. Availability at steady state, and the transition diagram in Fig.2 depicts a model
reliability at transient state and maintainability are showing all the possible states of the system.
 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665) 61

Table 1: System states

State Component state System condition
Available and standby Failed
S0 T1, T2, T3, T4 - Working
S1 T2, T3,T4 T1 Working
S2 T1, T3, T4 T2 Working
S3 T1, T2, T4 T3 Working
S4 T3, T4 T1, T2 Failed
S5 T2, T4 T1, T3 Failed
S6 T2, T3 T1, T4 Failed
S7 T1, T4 T2, T3 Failed
S8 T1, T3 T2, T4 Failed
S9 T1, T2 T3, T4 Failed

Figure 2: State transition diagram.

3. Model Proposed Where:

d /dt: derivative with respect to t.,P0 (t): probability that
The differential equations associated with the transition
diagram are derived on the basis of Markov birth-death the system is working at full capacity at time t, Pi (t): state
process. Various probability considerations generate the
probability that the system is in the ith state at time t, λi:
following sets of differential equations:
failure rate for unit i, μi: repair rate for unit i.
[(d/dt) + λ1+ λ2 +λ3] P0 (t) = μ1P1 (t) + μ2P2 (t) + μ3P3 (t) (1)
[(d/dt) + μ1 +λ2 + λ3 + λ4] P1 (t) = μ2 P4 (t) + μ3 P5 (t) + μ4
3.1. Availability Equations
P6 (t) + λ1 P0 (t) (2)
[(d/dt) + λ1 + λ3 + λ4 + μ2] P2 (t) = μ1 P4 (t) + μ3 P7 (t) + μ4 To get the steady state availability of the system (ASS),
(i.e., time independent performance behavior) which is
P8 (t) + λ2 P0 (t) (3) mean d/dt =0 and t → ∞, the above equations (eq. 1 to 10)
[(d/dt) + λ1+ λ2 + λ4 + μ3] P3 (t) = μ1 P5 (t) + μ2 P7 (t) + μ4 become:
(λ1 + λ2 + λ3) P0 = μ1P1 + μ2P2 + μ3P3 (12)
P9 (t) + λ3 P0 (t). (4)
(μ1 + λ2 + λ3 + λ4) P1 = μ2 P4 + μ3 P5 + μ4 P6 + λ1 P0 (13)
[(d/dt) + μ1 + μ2] P4 (t) = λ1P2 (t) +λ2 P1 (t) (5)
[(d/dt) + μ1 + μ3] P5 (t) = λ1 p3 (t) + λ3 P1 (t) (6) (λ1 + λ3 + λ4 + μ2) P2= μ1 P4 + μ3 P7 + μ4 P8 + λ2 P0 (14)
[(d/dt) + μ4] P6 (t) = λ4 P1 (t) (7) (λ1+ λ2 + λ4 + μ3) P3 = μ1 P5 + μ2 P7 + μ4 P9 + λ3 P0 (15)
[(d/dt) + μ2 +μ3] P7 (t) = λ2 P3 (t) + λ3 P2 (t) (8) (μ1 + μ2) P4 = λ1 P2+ λ2 P1 (16)
[(d/dt) + μ4] p8 (t) = λ4P2 (t) (9) (μ1+ μ3) P5 = λ1 p3 + λ3 P1 (17)
[(d/dt) + μ4] P9 (t) = λ4 P3 (t) (10) (μ4) P6= λ4 P1 (18)
Where, the initial conditions at time t = 0 are: (μ2+μ3) P7 = λ2 P3 + λ3 P2 (19)
1, ifi = 0 (μ4) p8 (t) = λ4 P2 (20)
pi (t) = { (11)
0, ifi ≠ 0 (μ4) P9 = λ4 P3 (21)
62 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665)

These equations were solved using Rung-Kutta Forth After solving these equations, the probabilities of
order method and MATLAB, the values of steady state operating states for the system, under consideration, are
probabilities are as follows: calculated as follows:
P1 = (λ1/μ1) P0, P2 = (λ2 / μ2) P0, P3 = (λ3 / μ3) P0m, P0 = (C A B) / D (35)
P4 = (λ1λ2) / (μ1μ2) P0, P5 = (λ1λ3) / (μ1μ3) P0, P1= (λ1 A B) / D (36)
P6 = (λ1λ4) / (μ1μ4) P0, P7 = (λ2λ3) / (μ2μ3) P0, P2 = (λ2 C B) / D (37)
P8 = (λ2λ4) / (μ2μ4) P0, P9 = (λ3 λ4) / (μ3 μ4) P0 P3 = (λ3 C A) / D (38)
The probability of full working capacity (P 0) is Where:
determined using normalizing conditions (i.e.,∑9𝑖=0 Pi= 1) A= μ2+ S +λ1 +λ3 +λ4 (39)
as follows:
B= μ3+ S +λ1 +λ2+λ4 (40)
P0= [(μ1 μ2 μ3 μ4) / (λ1 μ2 μ3 μ4 + λ2μ1μ3 μ4 + λ3 μ1 μ2 μ4 + λ1λ2
C= μ1+ S +λ2 +λ3 +λ4 (41)
μ3 μ4 + λ1λ3 μ2 μ4 + λ1λ4 μ2 μ3 + λ2λ3 μ1 μ4 + λ2λ4 μ1 μ3 + λ3λ4
D= see appendix A (42)
μ1 μ2)] (22)
Having the values of probabilities of working states
Having the values of probabilities (P 0-P9) determined,
determined, RS is calculated as follows:
ASS is calculated as a summation of all working state
probabilities as follows: Rs(S) = P0(S) + P1(S) + P2(S) + P3(S) (43)
ASS= P0 + P1 + P2 + P3 (23) Taking the inverse of Laplace transforms then, P0, P1,
P2, and P3, are calculated as F(t) and the system reliability
3.2. Reliability Equations. is calculated at time t as follows:
R (t) = P0 (t) + P1 (t) + P2 (t) + P3 (t) (44)
To get the reliability (RS) of the system under
consideration at any time, the equations (1 to10) are solved
taking Laplace transform and the probability transform are 3.3. Maintainability Equations
as follows:
For any system, the system maintainability (MS) is
[S + λ1 + λ2 +λ3] P0(S) = μ1P1(S) + μ2P2(S) + μ3P3(S) (24) calculated as follows:
[S + μ1 +λ2 + λ3 + λ4] P1(S) = μ2 P4(S) + μ3 P5(S) + μ4 MS (t) = 1- e (-μt) (45)
P6(S) + λ1 P0(S) (25) Where(μ) is the repair rate (μ= 1 ⁄ MTTR), MTTRs is a
[S + λ1 + λ3 + λ4 + μ2] P2(S) = μ1 P4(S) + μ3 P7(S) + μ4 mean time to repair of the system and is calculated as a
function in mean time to repair (MTTR) and mean time
P8(S) + λ2 P0(S) (26) between failure (MTBF) of system component i where:
[S +λ1+λ2 +λ4 + μ3] P3(S) = μ1 P5(S) + μ2 P7(S) + μ4 P9(S) ∑𝑛
𝑖=1 𝑀𝑇𝑇𝑅𝑖/𝑀𝑇𝐵𝐹𝑖
MTTRS = ∑𝑛
(46)
+ λ3 P0(S). (27) 𝑖=1 1/𝑀𝑇𝐵𝐹𝑖

[S +μ1 + μ2] P4(S) = λ1 P2(S) + λ2 P1(S) (28)

4. Implementation of Proposed Model
[S + μ1+μ3] P5(S) = λ1 p3(S) + λ3 P1(S) (29)
[S + μ4] P6(S) = λ4 P1(S) (30) Fig. 3 depicts the block diagram of the real industrial
system in Egyptian Petrol Company. This system is
[S + μ2 +μ3] P7(S) = λ2 P3(S) + λ3 P2(S) (31)
consists of two unites of fuel supply connected in parallel
[S + μ4] p8 (S) = λ4 P2(S) (32) with each other; in which the natural gas unit is active and
[S + μ4] P9(S) = λ4 P3(S) (33) the diesel unit is standby. The fuel supply units are
connected in series with four turbines which are connected
Where S is the Laplace transform variable. in parallel with each other (3-out-of-4); three of them are
To determine RS, a verification model is applied. It in operating state while the fourth is in the standby state
could be noted that the probabilities of the failed states and the system fails when two components (Turbines) fail.
haven’t any effect on the system reliability, so the system The turbines are considered the most critical items in
reliability could be calculated considering only the that system because it is used to operate four plants as
working states (i.e., excluding the failed states). Based on shown in the block diagram. Whereas the actual output for
this result, the probability of failed states could be each turbine is 3.8 MW and the total power required must
neglected during solution of the complex systems and the be not lower than 11.4 MW; otherwise, the system will
probabilities of operating states are only considered. stop working which tends to big losses.
The previous equations (24 to 33) are solved using The transition diagram of this system is built based on
MATLAB 2015 at the following initial conditions at time the Markov, as explained previously (in section 2). Table
t= 0 where, (2) illustrates the number of failures, the repair time, and
1, if i = 0 the operating time of this real system that were collected
pi (t) = { (34)
0, if i ≠ 0 from historical data during year 2015. These data are
applied to validate the proposed model. The RAM analysis
and discussions are presented in this section.
 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665) 63

Output
Input
Turbine (T1)
Loading plant

Natural gas (supply)

Turbine (T2)
Gas injection
Electrical loads
Turbine (T3) LPG
Diesel generator
(supply)
Turbine (T4) Oil plant

Figure 3: Block diagram of the industrial real system of Egyptian Petrol Company
(Turbines of power generation).
The value of the failure rate and repair rate for each
turbine is illustrated in Table (3) where: 4.2. System Reliability
Failure rate = number of failures / operating time (47)
RS is calculated as follows:
Repair rate = number of failures/ repair time (48)  Substitute about λ and µ from Table (3) in equations
As an example for T1; (39 to 42), the variables A, B, C and D could be
calculated as follows:
Failure rate (λ) = 11/ 6471= 0.0017 (49)
A = 10000.0S + 1541.0 (51)
Repair rate (μ) = 11 / 88 = 0.125 (50)
B = 10000.0S + 1543.0 (52)
C = 2500.0(5000.0S + 643.0) (53)
Table 2: The collected historical data of the considered system for D = (1.25e15S4 + 5.5112e14S3 + 8.075e13S2 +
year 2015.
3.9329e12S + 4.1045e8) (54)
Number of Repair time Operating
failures (hours) time (hours)  Substitute about the variables A, B, C, and D in (eq.,
T1 11 88 6471 35-38) to get the probability of the working states.
T2 8 53 6509  Taking the inverse Laplace of system reliability (eq., 35
T3 6 38 6526
- 38) to get :
T4 9 105 6456 P0 (t) =0.97e (-0.0001t) + 0.015e (-0.15t) + 0.00003e (-0.15t) +
0.01e (-0.13t) (55)
Table 3: Failure and repair rates of the considered system. (-0.0001t) (-0.15t) -6 (-0.15t)
P1 (t) =0.01e - 0.0001 e –2e e
T1 T2 T3 T4 (-0.13t)
– 0.012e (56)
Failure rate (λ) 0.0017 0.0012 0.001 0.0014
(-0.0001t) (-0.15t) (-0.15t)
P2 (t) =0.0075e - 0.007e - 0.0003e +
Repair rate ( μ) 0.125 0.15 0.15 0.08
0.0005e (-0.13t) (57)
(-0.0001t) (-0.15t) (-0.15t)
4.1. System Availability P3 (t) =0.006e - 0.007e + 0.0003e +
(-0.13t)
0.0004e (58)
By substituting values of failure and repair rates in
system equations from (Eq.12 to 21) the state probabilities  Substitute about P0, P1, P2 and P3in equation (44) to
are calculated as a function in P0 as follows: get RS as follows:
P1= 0.0136 P0, P2= 0.008 P0, P3= 0.006 P0, Rs= 0.9935 e (-0.0001t) + 0.0009 e (-0.15t) + 0.000028 e (-0.15t)
P4= 0.000011 P0, P5 =0.00001P0, P6 =0.0002 P0
- 0.0011e (-0.13t) (59)
P7= 0.000005 P0, P8 = 0.00014 P0, P9 = 0.00001 P0
Where, P0 and Ass are calculated from equations (22)
and (23) and equal 0.97 and 0.9995 respectively.
64 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665)

4.3. System maintainability

MTTRs is calculated from equation (46) based on the

historical data of the system components which is
illustrated in Table.2, (i.e., μs= 0.1196, MTTRS= 8.36 hr)
where,
MTTR= repair time / number of failure (60)
MTBF= 1/ failure rate (61)
4
So, MTTRS= ∑ i=1
((88/11+53/8+38/6+105/9)/(1/0.0017+1/0.0012+1/0.001+1
/0.0014)) = 8.36 h

∑4i=1(0.0017+0.0012+0.001+0.0014) (62)
Then the system maintainability at time t is as follows:
Figure 5: Effect of turbines repair rate (μ) on system availability
M (t) = 1- e (-0.1196 t) (63) at steady state.

5. Results Discussion 5.2. Reliability Analysis

An analysis of system performance has been carried out Fig. (6), depicts RS of the system under consideration at
at different values of failure and repair rates of system real data of λ and µ mentioned previously in Table (3),
components. The effects of these values on the system along operating time and it is concluded that the system
availability, reliability, and maintainability are discussed in reliability decreases with time and the system reliability
the following sections. after 5000 running hours is 0.59%.

5.1. Availability analysis

The ASS was calculated at different values of failure

rate and repair rate as follows:
Fig. (4) Shows the effect of failure rate for each turbine
on the system availability at different values of failure rate
(i.e., λ=0.001, 0.002, 0.003, 0.004, 0.005, 0.006) without
changing repair rate values mentioned in Table (3). It
could be seen from this Figure that the system availability
decreases, slightly, with increasing failure rate of the
system component. Moreover, it is observed that turbine
(T4) has the lower availability than the other turbines.
On the other hand Fig. (5) depicts the effect of repair
rate for each turbine on system availability at different
values of repair rate (i.e., μ=0.06, 0.09, 0.12, 0.15, 0.18, Figure 6: Effect of time on reliability of the system.
0.21) without changing failure rate values mentioned in
Table (3). It is observed that the system availability To investigate the effect of the failure rate of system
increases, slightly, with increasing repair rate of turbines, components individually on RS at different mission time,
and (T4) has the higher availability than the other turbines the failure rate of each turbine is changed within range
this because it has the lower value of repair rate. This (0.001 to 0.006 with incremental value 0.001) while, the
revealed that increasing the failure rate reduces the values of rapier rates for all turbines and failure rates of
availability while increasing the repair rate leads to the other turbines are the same real data.
increasing the availability. Fig. (7), Illustrates the criticality of the system
components, i.e., which turbine decreases RS. It could be
noted with comparing the cases of (λ = 0.001) for each
turbine that T3 has the higher effect on the system
reliability and therefore it is the critical component as
shown in Fig. (7-a), furthermore, with increasing λ to
0.006 for the same cases, it is also still the critical one as
shown in Fig (7-b).

Figure 4: Effect of turbines failure rate (λ) on system availability

at steady state.
 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665) 65

Figure 7:(a) and (b) Effect of failure rate (λ) on the system reliability at transient state.

6. Conclusion
5.3. Maintainability Analysis
A RAM model of a 3-out-of-4 system has been
Fig. (8) Illustrates the maintainability of each turbine proposed based on the Markovian approach. Availability at
as well as Ms of the overall system at real data along steady state, reliability at transient state and
first 100 operating hours. It could be seen that the maintainability equations have been formulated. A real
maintainability of T4 is lower than the data of Oil Gas Egyptian Company was applied to obtain
maintainability of the other components; this is due the system reliability, availability, and maintainability. A
to the lower value of its repair rate than the others. parametric investigation of various values of system
failure and repair rates on system reliability, availability,
and maintainability, as well as their effects on the system
performance, are presented. The finding of this study could
be concluded as follows:
 The proposed RAM model could be used as an
integrated model, to investigate system reliability,
availability, and maintainability of 3-out-of-4 system.
 It could be also used to determine the most critical
component of the system.
 The proposed model helps maintenance engineers and
designers to evaluate the system performance and
carried out modification.
Figure 8: Maintainability of each turbine and the overall system
versus time.  The implementation of the proposed model revealed
To investigate the effect of repair rate of overall system that the system availability at steady state is 99%, and
on Ms along the first 100 hrs of operating time, the repair the system maintainability is 0.99% but the system
rate of the system is assumed within range (0.0 to 0.21 reliability after 5000 running hours is 0.59%. This
with incremental value 0.03) as shown in Fig. (9). It could means that, an enhancement required improving the
be seen that the increase in the system repair rate increases system reliability and reducing the system down time.
Ms. It is observed that T3 is the most critical component in
the system and need special attention with careful
observation to reduce it’s down time and increase the
system performance.

References

[1] S. Saraswat and G.S. Yadava, “An overview on reliability,

availability, maintainability and supportability (RAMS)
engineering”, International Journal of Quality & Reliability
Management Vol. 25 No. 3, 2008, 330-344.
[2] D. Sanctis, C. Paciarotti, and O. D. Giovine, “Integration
between RCM and RAM: a case study”, International Journal
of Quality & Reliability Management, Vol. 33 Issue: 6, 2016,
852-880.
Figure 9: Effect of Repair Rate on Maintainability of the Overall [3] F. Corvaro, G. Giacchetta, B. Marchetti, and M. Recanati,
System at Different Mission Time. “Reliability, Availability, Maintainability (RAM) study, on
reciprocating compressors API 618 Petroleum”, Vol. 3, 2017,
266-272.
66 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665)

[4] M. Aoudia, O. Belmokhtar, G. Zwingelstein, "Economic [11] Aggarwal, S. Kumar, V. Singh, “Mathematical modelling
impact of maintenance management ineffectiveness of an oil and fuzzy availability analysis for serial processes in the
and gas company", Journal of Quality in Maintenance crystallization system of a sugar plant”, Journal of Industrial
Engineering, Vol. 14, 2008, 237-261. Engineering International, Vol.13, 2017, 47–58.
[5] R. K. Sharma, and S. Kumar, “Performance modelling in [12] J. Lin, J. Pulido, M. Asplund, “ Reliability analysis for
critical engineering systems using RAM analysis”, Reliability preventive maintenance based on classical and Bayesian
Engineering & System Safety, Vol.93, Issue 6, 2008, 913– semi-parametric degradation approaches using locomotive
919. wheel-sets as a case study”, Reliability Engineering and
[6] Komala, S.P. Sharma, and D. Kumar, “RAM analysis of System Safety, Vol. 134, 2015, 143–156.
repairable industrial systems utilizing uncertain data’, [13] P. Singh and A. Goyal,” Behaviour Analysis of a Biscuit
Applied Soft Computing, Vol. 10, 2010, 1208–1221. Making Plant using Markov Regenerative Modelling”,
[7] S. K. Rajiv, and S. Pooja, “Computing RAM Indices For International Journal on Theoretical and Applied Research in
Reliable Operation of Production Systems”, Advances in Mechanical Engineering (IJTARME), Volume-2, 2013, 2319
Production Engineering & Management, Vol. 7, 2012, 245- – 3182.
254. [14] I.Yusuf, and N. Hussaini,’ Evaluation of Reliability and
[8] R. Khanduja, P.C. Tewari, R.S.Chauhan, and D. Kumar, Availability Characteristics of 2-out of -3 Standby System
“Mathematical Modeling and Performance Optimization for under a Perfect Repair Condition”, American Journal of
the Paper Making System of a Paper Plant”, Jordan Journal Mathematics and Statistics, Vol. 2, 2012, 14-119.
of Mechanical and Industrial Engineering, Vol. 4 No 4, 2010, [15] Yusuf, and F. Koki,” Analysis of Stochastic Reliability
487 – 494. Characteristics of a Repairable 2-out-of-3 System with
[9] M. Kumar, V. Modgil, and V.Singla, “Mathematical Minimal Repair at Failure”, Applied Mathematics, Vol. 4,
Modelling and Availability Analysis of Packaging Section in 2013, 1115-1124.
a Paint Industry: A Case Study”, International Journal of [16] Preeti,” Analysis of a Linear Consecutive k-out-of-N System
Engineering Technology Science and Research, Vol. 2, 2015, with Common Cause Shock Failure”, International Journal of
179 – 185. Technology Innovations and Research (IJTIR), Vol.16, 2015.
[10] S. Gupta, P. C. Tewari, and A. Sharma, ”A probabilistic [17] Yusuf, ”Availability and Profit Analysis of 3-out-of-4
model for performance evaluation of steam and water system Repairable System with Preventive Maintenance”,
of a thermal power plant”, International Journal of International Journal of Applied Mathematical Research,
Management Science and Engineering Management, Vol. 4 Vol. 1(4), 2012, 510-519.
No.3, 2009,177-187. [18] Elsayed A. Elsayed, “Reliability Engineering 2nd Edition”,
John Wiley & Sons, Hoboken, United States, 2012.
 © 2018 Jordan Journal of Mechanical and Industrial Engineering. All rights reserved - Volume 12, Number 1 (ISSN 1995-6665) 67

Appendix (A)

D= (μ 1S3 + μ 2S3 + μ 3S3 + S λ13 + 3.0S3 λ1 + S λ23 + 3.0S3 λ2 + S λ33 + 3.0S3 λ3 + S λ43 + 3.0S3 λ4 + λ1 λ23 + λ13
λ2 + λ1 λ33 + λ13 λ3 + λ1 λ43 + λ2 λ33 + λ13 λ4 + λ23 λ3 + λ2 λ43 + h23 λ4 + λ3 λ43 + λ33 λ4 + S4 + 3.0S2 λ12 + 3.0S2 λ22
+ 3.0S2 λ32 + 3.0S2 λ42 + 2.0 λ12 λ22 + 2.0 λ12 λ32 +2.0 λ12 λ42 + 2.0 λ22h32 + 2.0 λ22 λ42 + 2.0 λ32 λ42 + μ1μ2S2 +
μ1μ3S2 + μ2μ3S2 + μ1S λ12 + 2.0 μ1S2 λ1+ μ1S λ22 + 2.0 μ1S2 λ2 + μ2S λ12 + 2.0 μ2S2 λ1 + μ1S λ32 + 2.0 μ1S2 λ3 +
μ2S λ22 + 2.0 μ2S2λ2 + μ3S λ12 + 2.0 μ3S2 λ1 + μ1S λ42 + 2.0 μ1S2 λ4 + μ2S λ32 + 2.0 μ2S2 λ3 + μ3S λ22 + 2.0 μ3S2 λ2
+ μ2S λ42 + 2.0μ2S2 λ4 + μ3S λ32 + 2.0 μ3S2 λ3 + μ3S λ42 + 2.0 μ3S2 λ1+ μ1 λ1 λ22 + μ1 λ12 λ2 + μ1 λ1 λ32 + μ1 λ12 λ3
+ μ2 λ1 λ22 + μ2 λ12 λ2 + μ1 λ2 λ32 + μ1 λ22 λ3 + μ2 λ1 λ32 + μ2 λ12 λ3 + μ3 λ1 λ22 + μ3 λ12 λ2 + μ1 λ2 λ42 + μ1 λ22 λ4 +
μ2 λ1 λ42 + μ2 λ2 λ32 + μ2 λ12 λ4+ μ2 λ22 λ3 + μ3 λ1 λ32 + μ3λ12 λ3 + μ1 λ3 λ42 + μ1 λ32 λ4 + μ3 λ1 λ42 + μ3 λ2 λ32 + μ3
λ12 λ4 + μ3 λ22 λ3 + μ2 λ3 λ42 + μ2λ32 λ4 + μ3λ2 λ42 + μ3 λ22 λ4 + 5.0S λ1 λ22 + 5.0S λ12 λ2 + 7.0S2 λ1 λ2 + 5.0S λ1 λ32
+ 5.0S λ12 λ3 + 7.0S2 λ1 λ3 + 5.0S λ1 λ42 + 5.0S λ2 λ32 + 5.0S λ12 λ4 + 5.0S λ22 λ3 + 7.0S2 λ1 λ4 + 7.0S2 λ2 λ3+ 5.0S
λ2 λ42 + 5.0S λ22 λ4+ 7.0S2 λ2 λ4+ 5.0S λ3 λ42 + 5.0S λ32 λ4+ 7.0S2 λ3 λ4+ 4.0 λ1 λ2 λ32 + 4.0 λ1 λ22 λ3 + 4.0 λ12 λ2
λ3 + 4.0 λ1 λ2 λ42 + 4.0 λ1 λ22 λ4+ 4.0 λ12 λ2 λ4 + 4.0 λ1 λ3 λ42 + 4.0 λ1 λ32 λ4+ 4.0 λ12 λ3 λ4 + 4.0 λ2 λ3 λ42 + 4.0 λ2
λ32 λ4+ 4.0 λ22 λ3 λ4 + μ1 μ2 μ3S + μ1μ2S λ1+ μ1 μ2S λ1 + μ1 μ3S λ1+ μ1 μ2S λ3+ μ1 μ3S λ2+ μ2 μ3S λ1+ μ1 μ2S λ4+
μ1 μ3S λ3+ μ2 μ3S λ2+ μ1 μ3S λ4+ μ2 μ3S λ3 + μ2 μ3S λ4+ μ1 μ2λ1 λ3 + μ1 μ3 λ1 λ2+ μ1 μ2 λ2 λ3+ μ2 μ3 λ1 λ2+ μ1 μ3 λ2
λ3+ μ2 μ3 λ1 λ3+ μ1 μ2 λ3 λ4+ μ1 μ3 λ2 λ4 + μ2 μ3 λ1 λ4 + 3.0 μ1S λ1 λ2+ 3.0 μ1S λ1 λ3 + 3.0 μ2 S λ1 λ2 + 2.0 μ1S λ1
λ4+ 3.0 μ1S λ2 λ3+ 3.0 μ2S λ1 λ3 + 3.0 μ3 S λ1 λ2+ 3.0 μ1S λ2 λ4+ 3.0 μ2S λ1 λ4+ 3.0 μ2S λ2 λ3+ 3.0 μ3 S λ1 λ3+ 3.0
μ1S λ3 λ4+ 2.0 μ2S λ2 λ4+ 3.0 μ3 S λ1 λ4+ 3.0 μ3S λ2 λ3+ 3.0 μ2S λ3 λ4 + 3.0 μ3S λ2 λ4+ 2.0 μ3 S λ3 λ4+ 2.0 μ1 λ1 λ2
λ3+ 2.0 μ1 λ1 λ2 λ4+ 2.0 μ2 λ1 λ2 λ3+ 2.0 μ1 λ1 λ3 λ4+ 2.0 μ2 λ1 λ2 λ4+ 2.0 μ3 λ1 λ2 λ3+ 2.0 μ1 λ2 λ3 λ4 + 2.0 μ2 λ1 λ3
λ4+ 2.0 μ3 λ1 λ2 λ4+ 2.0 μ2 λ2 λ3 λ4+ 2.0 μ3 λ1 λ3 λ4+ 2.0 μ3 λ2 λ3 λ4+ 11.0S λ1 λ2 λ3+ 11.0S λ1 λ2 λ4+ 11.0S λ1 λ3
λ4+ 11.0S λ2 λ3 λ4+ 9.0 λ1 λ2λ3 λ4)