Вы находитесь на странице: 1из 19

Asset Management.

The state of the Art in Europe from a Life Cycle Perspective

Van der Lei, Telli; Herder, Paulien; Wijnia, Ype (Eds.)

2012, 2012, XIV, 172 p.

ISBN 978-94-007-2723-6

Chapter 6

(pages 81-99)

Life Cicle Cost Analysis (LCCA) consideration within the built and in-use assets maintenance management.

A. Crespo Márquez, **C. Parra Márquez, J.F. Gómez Fernández, M. López Campos & V. González Díaz

Dept. Industrial Management. University of Seville

School of Engineering, University of Seville, Spain

Abstract

The chapter presents a generic model for assets maintenance management. This model inte- grates other models found in the literature for built and in-use assets, and consists of se- quential management building blocks. More precisely we want to show the reader the im- portance of selecting an appropriate method when considering the estimation of the non- reliability cost of an asset. By doing so, we show the impact of maintenance in life cycle costing and provide arguments to claim about the needs for proper assets maintenance con- trol.

1. Introduction

In this chapter, in the first part, we illustrate a process (Section 2) for built and in-use assets maintenance management and to characterize maintenance engineering techniques within that process. This has become a research topic and a fundamental question to reach the effective- ness and efficiency of maintenance management and to fulfill enterprise objectives [15]. We review a model/process proposed in this chapter tries somehow to integrate other models found in the literature (see for instance [6,7]) and presents a total of eight sequential manage-

81

ment building blocks. Each block, as will be discussed, is a key decision area for asset main- tenance and life cycle management.

In the second part of the chapter (Section 3), among referred decision areas and according to the editorial team of this project, we have selected to explore methods and models that may be used to do a suitable asset life cycle cost analysis. More precisely we want to show the reader the importance of selecting an appropriate method when considering the estimation of the non reliability cost of an asset. By doing so, we somehow show the impact of maintenance in life cycle costing and provide arguments to claim about the needs for proper assets mainte- nance control.

2. Characterizing the Maintenance Management Process

The maintenance management process can be divided into two parts: the definition of the strategy, and the strategy implementation. The first part, conditions the success of mainte- nance in an organization, determines the effectiveness of maintenance. Maintenance effec- tiveness allows the minimization of the maintenance indirect costs [3] associated with produc- tion losses and customer dissatisfaction [4], reduces the overall company cost, obtained because production capacity is available when needed [5]. The second part of the process, the implementation of the strategy will allow us to mini- mize the maintenance direct cost (labour and other maintenance required resources). Effi- ciency is acting or producing with minimum waste, expense, or unnecessary effort.

Effectiveness Phase 1: Phase 3: Definition of the Phase 2: Immediate maintenance Assets priority intervention
Effectiveness
Phase 1:
Phase 3:
Definition of the
Phase 2:
Immediate
maintenance
Assets priority
intervention
objectives and
and maintenance
on high impact
KPI’s
strategy definition
weak points
Phase 4:
Improvement
Phase 8:
Design of
Continuous
the preventive
Improvement
maintenance
and new
plans and
techniques
resources
utilization
Phase 7:
Phase 6:
Phase 5:
Asset life cycle
Maintenance
Preventive plan,
analysis
execution
schedule
and replacement
assessment
and resources
optimization
and control
optimization
Efficiency
Assessment

Figure 1. Maintenance management model (Adapted from [2])

Our model for maintenance management consists of eight sequential management building blocks, as presented in Figure 1. At the same time, our idea is that there are maintenance engi- neering tools that may be used to improve each building block decision making process (see Figure 2).

82

Effectiveness   Phase 3: Phase 3: Immediate Failure Root intervention Cause Analysis on high impact

Effectiveness

 
Phase 3: Phase 3: Immediate Failure Root intervention Cause Analysis on high impact (FRCA) weak
Phase 3:
Phase 3:
Immediate
Failure Root
intervention
Cause Analysis
on high impact
(FRCA)
weak points
   
on high impact (FRCA) weak points     Phase 4: Improvement Phase 8: Phase 8: Design
Phase 4: Improvement Phase 8: Phase 8: Design of Phase 4: Continuous Total Productive the
Phase 4:
Improvement
Phase 8:
Phase 8:
Design of
Phase 4:
Continuous
Total Productive
the preventive
Reliability-
Improvement
Maintenance
maintenance
Centred
and new
(TPM),
plans and
Maintenance
techniques
e-maintenance
resources
(RCM)
utilization
Phase 5:
Preventive Phase plan, 5:
Risk―Cost schedule
and Optimization resources
optimization (RCO)
Efficiency
and Optimization resources optimization (RCO) Efficiency [10]. Phase 7: Asset Phase life cycle 7: Life
and Optimization resources optimization (RCO) Efficiency [10]. Phase 7: Asset Phase life cycle 7: Life
and Optimization resources optimization (RCO) Efficiency [10]. Phase 7: Asset Phase life cycle 7: Life
[10]. Phase 7: Asset Phase life cycle 7: Life analysis Cycle and Cost replacement Analysis

[10].

[10].
[10].

Phase 7:

Asset Phase life cycle 7:

Life analysis Cycle

and Cost replacement Analysis

optimization (LCCA)

Phase Phase 6: 6:

Maintenance Reliability

Analysis execution (RA)

& assessment Critical Path

and Method control

(CPM)

Assessment Maintenance Maintenance Cost Effectiveness Cost Effectiveness Maintenance cost (%) Maintenance cost (%)
Assessment
Maintenance
Maintenance
Cost Effectiveness
Cost Effectiveness
Maintenance cost (%)
Maintenance cost (%)
per unit produced (7%)
per unit produced (7%)
Maintenance
Maintenance
planning
planning
Quality
Quality
and scheduling
and scheduling
Accomplishment
Accomplishment
PM
PM
of criticality analysis
of criticality analysis
Compliance
Compliance
(Every 6 months)
(Every 6 months)
(98%)
(98%)
   
   
   
   
                Figure 2. Sample of techniques within the maintenance
                Figure 2. Sample of techniques within the maintenance

Figure 2. Sample of techniques within the maintenance management framework (Adapted from [2])

Phase 1 tries to avoid that the maintenance objectives and strategy could be inconsistent with the declared overall business strategy [8]. This can indeed be done by introducing the Balanced Scorecard (BSC) [9]. The BSC is specific for the organization for which it is devel- oped and allows the creation of key performance indicators (KPIs) for measuring maintenance management performance which are aligned to the organization’s strategic objectives (See Figure 3).

the organization’s strategic objectives (See Figure 3). Learning Learning Data integrity Data integrity (95%) (95%)

Learning

Learning

strategic objectives (See Figure 3). Learning Learning Data integrity Data integrity (95%) (95%) Figure 3. A

Data integrity

Data integrity

(95%)

(95%)

Figure 3. A KPI and its functional indicators (Adapted from [2])

Unlike conventional measures which are control oriented, the Balanced Scorecard puts overall strategy and vision at the centre and emphasizes on achieving performance targets

Once the Maintenance Objectives and Strategy are defined, there are a large number of quantitative and qualitative techniques which attempt to provide a systematic basis for deciding what assets should have priority within a maintenance management process (Phase 2). Most of the quantitative techniques use a variation of a concept known as the “probability/risk number” (PRN) [11]. In professional risk assessments, risk combines the

83

probability of an event occurring with the impact that event would cause R=PxC, where P is probability and C is consequence (Figure 4). Risk assessment techniques can be used to prioritize assets and to align maintenance actions to business targets at any time.

44 44 FF FF rr rr CriticalCritical CriticalCritical 1 1 2 2 1 1 3
44 44
FF FF
rr rr
CriticalCritical
CriticalCritical
1 1
2 2
1 1
3 3
ee ee
33 33
qq qq
SemiSemi--criticalcritical
SemiSemi--criticalcritical
uu uu
ee ee
22 22
nn nn
4 4
2 2
NonNon--criticalcritical
NonNon--criticalcritical
cc cc
yy yy
11 11
3 3
1010 1010
2020
2020
3030
3030
4040
4040
5050
5050
ConsequenceConsequence
ConsequenceConsequence
Figure 4. Generic criticality matrix and assets location

As mentioned above, once there is a certain ranking of assets priority, we have to set up the strategy to follow with each category of assets. Of course, this strategy will be adjusted over time, and will consist of a course of action to address specific issues for the emerging critical items under the new business conditions (see Figure 5).

items under the new business conditions (see Figure 5). Reach optimal reliability, A A maintainability and

Reach optimal reliability,

A

A

maintainability and

A A maintainability and  
A A maintainability and  
 

Asset category

B

B

availability levels

Ensure certain

equipment availability

levels

Sustain

Maintenance strategy

Maintenance strategy

C

C

improve

current situation

Figure 5. Example of maintenance strategy definition for different category assets [2]

An example of detailed maintenance actions for category A assets where we try to reach optimal reliability, maintainability and availability levels could be: 1) Apply FMECA for critical failure mode analysis; 2) Apply RCM for optimal maintenance task selection; 3) Standardise maintenance tasks; 4) Analyse design weaknesses and 5) Continue review FMECA and RCM. Phase 3 deals with finding and eliminating, if possible, the causes of certain repetitive fail- ures that take place in high priority items. There are different methods developed to carry out this weak point analysis, one of the most well known being root-cause failure analysis (RCFA). This method consists of a series of actions taken to find out why a particular failure or problem exists and to correct those causes. Phase 4 is devoted to the design of the preventive maintenance plan for a certain system and this requires identifying its functions, the way these functions may fail and then establish

84

a set of applicable and effective preventive maintenance tasks, based on considerations of sys- tem safety and economy. A formal method to do this is the Reliability Centred Maintenance (RCM), as in Figure 6.

Initial RCM Phase Implementation phase RCM team Operational Functional conformation context Function Failure
Initial
RCM
Phase
Implementation phase
RCM
team
Operational
Functional
conformation
context
Function
Failure modes
failures
definition
and asset
Criticality
selection
Analysis
FMEA
(level?)
Failure Mode and
Effect of
failure modes
Effects Analysis
Tool to answer the first 5
RCM Questions
Tool to answer
Application of
the last 2
the RCM
RCM Questions
logic
Final
Maintenance
plan
Phase
documentation

Figure 6. RCM implementation process

Optimization of maintenance planning and scheduling (Phase 5) can be carried out to en- hance the effectiveness and efficiency of the maintenance policies resulting from an initial preventive maintenance plan and program design. Models to optimize maintenance plan and schedules will vary depending on the time horizon of the analysis [13]. Phase 6 deals with the execution of the maintenance activities ― once designed planned and scheduled using techniques described for previous building blocks . This execution has to be evaluated and deviations controlled to continuously pursue business targets and ap- proach stretch values for key maintenance performance indicators as selected by the organiza- tion. A life cycle cost analysis (Phase 7) calculates the cost of an asset for its entire life span (see Figure 7). The analysis of a typical asset could include costs for planning, research and devel- opment, production, operation, maintenance and disposal. A life cycle cost analysis is impor- tant when making decisions about capital equipment (replacement or new acquisition) [12], it reinforces the importance of locked in costs, such as R&D, and it offers important benefits. We concentrate on techniques for LCCA in Section 3 of this Chapter.

85

CAPEX CAPEX OPEX OPEX Capital Costs Capital Costs Operational Costs Operational Costs Development Development
CAPEX
CAPEX
OPEX
OPEX
Capital Costs
Capital Costs
Operational Costs
Operational Costs
Development
Development
Investment
Investment
Operation
Operation
costs
costs
costs
costs
costs
costs
Acquisition
Acquisition
Corrective Maintenance + Security, Environment, Production =
Corrective Maintenance + Security, Environment, Production =
NonNon ReliabilityReliability CostsCosts == RiskRisk
NonNon ReliabilityReliability CostsCosts == RiskRisk
Design
Design
Operation + Planned Maintenance Costs.
Operation + Planned Maintenance Costs.
Investigation
Investigation
Construction
Construction
Remove
Remove
Time (years)
Time (years)

Figure 7. Life cycle cost analysis

Finally, continuous improvement of maintenance management (Phase 8) will be possible due to the utilization of emerging techniques and technologies in areas that are considered to be of higher impact as a result of the previous steps of our management process. Regarding the application of new technologies to maintenance, the “e-maintenance” concept (Figure 8) is put forward as a component of the e-manufacturing concept [14], which profits from the emerging information and communication technologies to implement a cooperative and dis- tributed multi-user environment. E-Maintenance can be defined [10] as a maintenance support which includes the resources, services and management necessary to enable proactive decision process execution.

Conventional Maintenance Conventional Maintenance E-maintenance Top Management Top Management Top Management Top
Conventional Maintenance
Conventional Maintenance
E-maintenance
Top Management
Top Management
Top Management
Top Management
Reports
Reports
Middle Management
Middle Management
Middle Management
Middle Management
Precise &
Precise &
Login to
Login to
Reports
Reports
Concise
Concise
iScada
iScada
Information
Information
Maintenance Dept
Maintenance Dept
Maintenance Dept
Maintenance Dept
Inspections/Complaints
Inspections/Complaints
Assets /
Assets /
Assets /
Assets /
Information Source
Information Source
Information Source
Information Source
Figure 8. Implementing e-maintenance (http://www.devicesworld.net)

3. Evaluating the economic impact of the failure in the LCCA

Life cycle costing is a well-established methodology that takes into account all costs arising during the life cycle of the asset. These costs can be classified as the ‘capital expenditure’ (CAPEX) incurred when the asset is purchased and the ‘operating expenditure’ (OPEX) in- curred throughout the asset’s life. LCCA is a method that can be used, for instance, to evalu- ate alternative asset options [1], or/and assets maintenance management strategies [2]. For all these potential purposes, a key aspect to introduce in a LCCA is the failure costs. In order to model that cost we will now introduce the Non-homogeneous Poisson Process model (NHPP, repairable systems). With the NHPP model we can estimate the frequency of failures and the impact that could cause the diverse failures in the total cost of ownership of a produc-

86

tion asset. This section also contains a case of study to illustrate the above mentioned con- cepts.

3.1. Characterizing the total costs of failures (non reliability)

Life cycle cost analysis (LCCA) can be defined [14] as a systematic process of technical- economical evaluation that considers, in a simultaneous way, economic and reliability aspects of an asset, quantifying their real impact along its life cycle cost. Reliability is related to op- erational continuity. We normally say that a production system is "reliable" when it is able to accomplish its function in a secure and efficient way along its life cycle. Low reliability causes normally high costs, mainly associated to the asset function recovery (direct costs) be- sides the corresponding escalated impact in the production process (penalization costs). The totals costs of non reliability can be then classified as follows ([20], [21] and [22]):

Costs for penalization: Downtime, opportunity losses/deferred production, production losses (unavailability), operational losses, impact in the quality, impact in security and en- vironment. Costs for corrective maintenance: Manpower, direct costs related with the manpower (own or hired) in the event of a non planned action; and materials and replacement parts, direct costs related with the consumable parts and the replacements used in the event of an unplanned action.

3.2. Using NHPP for Reliability Analysis

NHPP is a stochastic discrete process where, in its initial formulation, we assume that the equipment is “as bad as old” (ABAO) operating condition after a repair (this is also referred as minimal repair in the maintenance modelling literature [24, 25]). In this process the probabil- ity of occurrence of n failures in any interval [t1, t2] has a Poisson distribution with the mean

[24]:

1

2

t

t

t dt

( )

Where (t)

Therefore, according to the Poisson process:

is the rate of occurrence of failures (ROCOF).

Pr[

N

(

t

2

)

N

(

t

1

)

n

]

t

t

1

2

t dt

( )

n

exp

t

t

1

2

t dt

( )

n !

(1)

(2)

Where n = 0, 1, 2,… are the total expected number of failures in the time interval [t1,t2]. Let us represent with (t) the expected number of failures in a time interval [0, t], then

( )

t

t

0

( )

t dt

(3)

One of the most common forms of ROCOF used in reliability analysis of repairable systems is the Power Law Model ([24] and [25]), that estimates the failure rate as follows:

( )

t

t

 

1

(4)

This form comes from the assumption that the inter-arrival times between successive fail- ures follow a conditional Weibull probability density function, with parameters α and β. The Weibull distribution is typically used in maintenance area due to its flexibility and applicabil- ity to various failure processes (however, solutions to Gamma and Log-normal distributions are also possible). As we know by reliability theory, λ(t) is a conditional probability for which we can consider the following definition (see Figure 10):

87

Figure 10. Conditional probability of occurrence of failure [26] P T (  t T

Figure 10. Conditional probability of occurrence of failure [26]

P T

(

t

T

t

1

)

( )

F t

(

F t

1

)

1

( )

R t

 

1

( )

R t

(

R t

1

)

(

R t

1

)

 

1

( )

R t

(

R t

1

)

(5)

where F(t) and R(t) are the probability of failure and the reliability at the respective times. Assuming a Weibull distribution, Eq. (5) yields:

F

(

t

i

)

 

1

exp

 

t

i 1

 

t

i

 

(6)

Therefore, the conditional Weibull density function is:

f

(

t

i

)

t

i

 

1

.exp

t

i 1

t

i

(7)

Now we will use this function in order to obtain the maximum likelihood (ML) estimators of the parameters of the Power Law model. For the case of the NHPP, different expressions for the likelihood function may be obtained. We will use expression based on estimations at a time t after the occurrence of the last failure and before the occurrence of the next failure, see details on these expressions in [27].

3.2.1. Time terminated NHPP maximum likelihood estimators

In the case of time terminated repairable components, the maximum likelihood function L can be expressed as:

n

L i   1

f

(

t

i

) f

Therefore:

L

 

 

 

 

 

t

1

1

n 1

n t

i

2

1

( t

1 )

n

f

i 2

(

t

i

)

R ( t

exp

1

t

1

exp

 

n

i

2


     t

i

1

n

t )

 

t

i

  

 

(8)

(9)

Then the ML estimators for the parameters are calculated. The results are ([24] and [25]):

88

ˆ

ˆ

t

n

1

n

n

n

ln i 1

t n  


i

t

(10)

(11)

Where ti is the time at which the ith failure occurs, tn is the total time where the last fail- ure occurred, and n is the total number of failures. The total expected number of failures in the time interval [tn, tn+s] by the Weibull cumulative intensity function is [27]:

(

t

n

,

t

n

s

)

1

 

t

 n

t

s

t

n



(12)

Where t s is the time after the last failure occurred in the one which needs to be considered the number of failures and tn is:

t

n

  t

n i 1

i

(13)

3.3. A NHPP MODEL FOR FAILURE COST ASSESSMENT IN LCCA

Our previous NHPP model structure can be used for the quantification of the costs of fail-

ures in the LCCA (cost of non reliability [28]). With this model we propose to assess the im- pact of main failures on a production system LCC structure by following the next procedure:

1. Identify for each alternative to evaluate the main types of failures. This way for certain equipment there will be f = 1F types of failures.

2. Determine for the n (total of failures), the times to failures t f . This information will be

gathered by the designer based on records of failures, databases and/or experience of maintenance and operations personnel.

3. Calculate the Costs for failures C f ($/failure). These costs include: costs of penalization

for production loss and operational impact Cp ($/hour), costs of maintenance corrective Cc ($/hour) and the mean time to repair MTTR (hours). The expression used to estimate

the

C

f

C

(

f

is shown next:

Cp

Cc

)

MTTR

(14)

4. Define the expected frequency of failures per year (tn ,tns ) . This frequency is assumed as a constant value per year for the expected cycle of useful life. The (tn,tns) is calcu-

lated starting from the expression (12). This process is carried out starting from the times

to failures registered

from the following expressions (10) and (11). In the expression (12), ts it will be a year (1 year) or equivalent units (8760 hours, 365 days, 12 months, etc.). This time ts represents the value for estimate de frequency of failures per year.

t f by failure type (step 2). The parameters and , are set starting

5. Calculate the total costs per failures per year TCPf , generated by the different events of

stops in the production, operations, environment and security, with the following expres- sion:

89

6.

TCP

f

F

 

f

t

n

, t

n

s

C

f

(15)

The obtained equivalent annual total cost, represents the probable value of money that will be needed every year to pay the problems of reliability caused by the event of failure, during the years of expected useful life.

Calculate the total costs per failures in present value

. Given a yearly value

, the quantity of money in the present (today) that needs to be saved, to be able to

pay this annuity for the expected number of years of useful life (T), for a discount rate (i).

The expression used to estimate the

PTCP

f

TCP

f

PTCP

f

is shown next:

PTCP

f

TCP

f

1

i

T

1

i

1

i

T

(16)

Once this cost is estimated, it is added to the rest of the evaluated costs (investment, planned maintenance, operations, etc.). Finally, the total cost is calculated in present value for the selected discount rate and the expected years of useful life. Different results can be ob- tained, for instance for different assets options or/and maintenance strategy options.

3.4. CASE STUDY

The following case study proposes the evaluation of the economic impact of the failures us- ing the method NHPP. The analysis was developed for the oil company PETRONOX (con- tractor of Petróleos of Venezuela), located in the field of gas and petroleum Naricual II, in Monagas, Venezuela. In general terms, it is requires to install a compression system to man- age a flow average of 20 millions of cubic feet of gas per day. The organization PETRONOX, evaluates the information of two suppliers of compressors. Next, are shown the data of costs of: initial investment, operation and maintenance for the two options to evaluate (value esti- mated by the suppliers, see Table 1):

Option A:

Reciprocant Compressor, 2900-3200 hp, caudal: 20 millions of feet cubic per day

Option B:

Reciprocant Compressor, 2810-3130 hp, caudal: 20 millions of feet cubic per day

Data

Option A

Option B

I: Investment

1.100.000 $

900.000 $

OPC: opera-

100.000 $/year

120.000 $/year

tionals costs

PRC: preven-

60.000 $/year

40.000 $/year

tive costs

OVC: overhauls

costs

100.000 $ every 5 years

80.000 $ every 5 years

i: interest

10%

10%

T: expected

15 years

15 years

useful life

Table 1. Economical data

With this information the organization PETRONOX carried out a first economic LCCA and a comparison made among the two alternatives, in this first evaluation, no failure cost analysis was considered and results are presented in Table 2:

90

 

Results

Option A

Option B

1) I: Invesment

1.100.000 $

900.000 $

2) OPC(P):

760.607,951 $

912.729,541 $

operationals costs in present value

 

3) PRC(P): preventive costs in present value

456.364,77 $

304.243,18 $

4) OVC(P): overhauls costs in present value,

62.092, 1323 $

49.673,7058 $

t

= 5 years

5) OVC(P): overhauls

38.554,3289 $

30.843,4632 $

costs in present value,

 

t

= 10 years

6) OVC(P): overhauls

23.939,2049 $

19.151,3639 $

costs in present value,

 

t

= 15 years

TLCC(P): Total Life Cycle Costs in pre- sent value, i: 10%, T:

2.441.558,387 $

2.216.641,254 $

15 years (Sum 1…6)

Table 2. Economical results without to evaluate the costs per failures

In Table 2, the oil company doesn't consider the possible costs of failures events. The op- tion B results to be the best economic alternative (more economic alternative for a lifespan pe- riod of 15 years). There is a difference of approximately: 224.917,133 $ between the two al- ternatives (this quantity would be the potential saving to select the option B, without considering the possible costs for failures). Later on, a proposal consisting on the evaluation of the same figures taking into considera- tion now the failure costs was made to the organization. It was suggested using a NHPP model

for this evaluation, the total expected number of failures the interval of time [tn, tn+s] is es- timated by the NHPP stochastic model (Weibull cumulative intensity function) [27]. Next, are shown the data of costs and times of failures to be used inside the NHPP model (the data of

were gathered by PETRONOX of two similar compression systems that

operate under very similar conditions in those that will work the compressor to be selected):

times to failures

t

f

91

 

Data

Option A

Option B

Cp

($/hour)

6.000

6.000

Cc

($/hour)

700

400

MTTR (hours)

9

8

t

f

(months)

5, 7, 3, 7, 2, 4, 3, 5, 8, 9, 2, 4, 6, 3, 4, 2, 4, 3, 8, 9

2, 3, 3, 5, 6, 6, 5, 6, 5, 6, 4, 3, 2, 2, 2, 2, 3, 2, 2, 3, 2, 2, 3, 3

t

n (total of

98

82

months)

 

n (total of fail- ures)

20

24

Table 3. Failure costs and maintainability/reliability data

With the information of the Table 3, the equation (16) was used to calculate the frequency

of failures per year

tained in the equation (16) were calculated from the equations (14) and (15). The total costs

TCP f were calculated from the equations (18) and (19); these costs are

. The parameters and of the Distribution of Weibull con-

(

t

n

,

t

n

s

)

for failures per year

with the equation (20). Next, are shown the results of the

frequency of failures and the total costs for failures for year obtained starting from the NHPP model, for the two evaluated options:

converted to present value

PTCP

f

Results

Option A

Option B

6,97832

6,13985

1,13382

1,22614

(

t

n

,

t

n

s

)

= fail-

ures/year

TCPf

= $/year

PTCP

f

=

$

(i=10%, T=15 years)

2,7987 = 2,8

168.840

1.284.210,46

4,3751=4,38

224.256

1.705.708,97

Table 4. Results from NHPP model

Later on, a second LCC economic evaluation was carried out including the results of costs of failures obtained from the NHPP model. The results are presented in Table 5:

92

 

Results

Option A

Option B

1) I: Invesment

1.100.000 $

900.000 $

2) OPC(P):

760.607,951 $

912.729,541 $

operationals costs in present value

 

3) PRC(P): preventives costs in present value

456.364,77 $

304.243,18 $

4) OVC(P): overhauls costs in present value,

62.092, 1323 $

49.673,7058 $

t

= 5 years

5) OVC(P): overhauls costs in present value,

38.554,3289 $

30.843,4632 $

t

= 10 years

6) OVC(P): overhauls

23.939,2049 $

19.151,3639 $

costs in present value,

 

t

= 15 years

7)PTCPf: total costs

1.284.210,46 $

1.705.708,97 $

per failures in present

 

value

TLCC(P): Total Life Cycle Costs in present value, i: 10%, T: 15 years (Sum 1…7)

3.725.768,851 $

3.922.350,22 $

PTCPf / TLCC(P) = %

34,46%

43.48%

(total costs per failures

/ total life cycle costs)

Table 5. Economical results with the costs per failures

In the results of this second evaluation (see Table 5), the total costs for failures are included in present value PTCPf. Notice that now Option A turns out to be the best economic alterna- tive, with a difference of approximately: 196.581,368 $ (this quantity would be the potential saving if selecting the option A instead of B). An important aspect to be considered in this analysis, is that PTCPf category of cost turns out to be the highest economic factor, with more weight, inside the process of the two alternatives comparison. Specifically, this category of costs represents the 43,48% (Option B) and the 34,46% (Option A) of the total LCC of the- se two assets (with an interest rate of 10% and a prospective cycle of life of 15 years). Finally, as per previous results discussion, PETRONOX decided to consider failures cost analysis in their LCCA. Additionally, the organization PETRONOX decided to develop an in- ternal procedure allowing the evaluation of reliability opportunity cost, this procedure would be used in a continuous and obligatory basis every time different options are analyzed inside the processes of: design, selection, substitution and/or purchase of assets.

3.4.1. Limitations of the NHPP proposed model

The analysis of the failure is an important facet in the development of maintenance strategy in the life cycle cost analysis of the asset. Only by properly understanding the mechanism of failure, through the modeling of failure data, can a proper maintenance plan and an analysis of costs be developed [47]. This is normally done by means of probabilistic analysis of the fail- ure data. From this, conclusions can be reached regarding the effectiveness and efficiency of preventive replacement (and overhaul) as well as that of predictive maintenance. The optimal

93

frequency of maintenance can also be established by using well developed optimization mod- els. These optimize outputs, such as profit, cost and availability. The problem with this ap- proach is that it assumes that all repairable systems are repaired to the ‘good-as-new’ condi- tion at each repair occasion. Maintenance practice has learnt, however, that in many cases equipment slowly degrades even while being properly maintained (including part replacement and periodic overhaul). The result of this is that failure data sets often display degradation. This renders conventional probabilistic analysis useless. The NHPP model has proved to provide good results even for realistic situations with bet- ter-than-old but worse-than-new repairs [29]. Based on this, and given its conservative nature and manageable mathematical expressions, the NHPP was selected for this particular work. The NHPP models can be considered as simple curve-fitting approach that can be easily un- derstood and implemented by software engineers and developers [47]. It is also this type of models that have been used by practitioners in most cases. On the other hand, without an in- depth understanding, the models and analysis are more likely to be misused and further analy- sis, which could been possible are not carried out. There is a need for more in depth study of NHPP model and their effectiveness in predicting future failure behaviour. Most of current re- search focuses on developing more complex models, see other models found in the literature ([48] and [49]). However more research is needed with regard to model selection. When com- paring models, the focus should be on the prediction rather than fitting as a model can fit the past data correctly, but has a poor predictive ability. Knafl and Morgan [50] provides some initial discussion on this area. The model described above has advantages and limitations. In general, the more realistic is the model, the more complex are the mathematical expression involved. The main strengths and weakness of this model are summarized next:

Strengths:

It is a useful and quite simple model to represent equipment under aging (deterioration).

Involves relatively simple mathematical expressions.

It is a conservative approach and in most cases provides results very similar to those of more complex models like Generalized Renewal Process [29]. Weakness:

Is not adequate to simulate repair actions that restore the unit to conditions better than new or worse than old.

4. Conclusions

The orientation of this chapter is towards maintenance management models, and within them, to the presentation of techniques to consider LCCA within the process (Phase) of assets maintenance assessment, control and improvement. We have shown how the reliability factor and its impact on costs can be critical for LCCA and may influence in final results produced with this analysis for assets options and/or for maintenance management strategy alternatives. Prevision of unexpected failure events and their cost is crucial for correct decision making and profitability of production process. Improvements of process reliability (quality of the de- sign, used technology, technical complexity, frequency of failures, costs of preven- tive/corrective maintenance, maintainability levels and accessibility) may have a great impact on the total cost of the life cycle of the asset, and on the possible expectations to extend the useful life of the assets to reasonable costs.

94

5. Future trends

We believe that, within LCCA techniques, there is a potential area of research related to the optimization of the reliability impact evaluation techniques on LCC. Some interesting trends that we have identified are as follows:

Stochastic methods see ([30], [31] and [32]). Table 6 shows the stochastic processes used in reliability investigations of repairable systems, with their possibilities and limits [27].

Advanced maintenance optimization using genetic algorithms see ([33] and [34]).

Monte Carlo simulation techniques see ([35], [36] and [37]).

Advanced Reliability distribution analysis see ([38], [39], [40], [41] and [42]).

Markov simulation methods see ([43], [44], [45] and [46]).

Reliability methods in phase of design see ([51] and [52]).

95

Stochastic pro-

Can be used

Back-

cess

ground/

Difficulty

Renewal process

Spare parts provisioning in the case of arbitrary failure rates and negligi- ble replacement or repair time (Poisson process)

Renewal the-

ory/

Medium

Alternating

One-item repairable (re- newable) structure with arbitrary failure and re- pair rates

Renewal the-

renewal process

ory/

Medium

Markov process

Systems of arbitrary structure whose elements have constant failure and repair rates during the stay time (sojourn time) in every state (not neces- sarily at a state change, e.g. because of load shar- ing)

Differential

(MP)

equations

or integral

equations/

Low

Semi-Markov

Some systems whose el- ements have constant or Erlangian failure rates (Erlang distributed fail- ure-free times) and arbi- trary repair rates

Integral

process (SMP)

equations/

Medium

Semi-Regenerative

Systems with only one repair crew, arbitrary structure, and whose el- ements have constant failure rates and arbitrary repair rates

Integral

process

equations/

High

Nonregenerative

process

Systems of arbitrary structure whose elements have arbitrary failure and repair rates

Partial diff. eq.; case by base sol./ High to very high

Table 6. Stochastic processes used in reliability analysis of repairable systems

Finally, it is not feasible to develop a unique LCCA model, which suits all the require- ments. However, it is possible to develop more elaborate models to address specific needs such as a reliability cost-effective asset development.

6. Acknowledgements

This research is funded by the Spanish Department of Science and Innovation project DPI2008-01012 (Modelling e-maintenance policies for the improvement of production sys- tems dependability and eco-efficiency).

96

References

[1] Blanchard BS. 2001. “Maintenance and support: a critical element in the system life cycle”, Proceedings of the International Conference of Maintenance Societies, paper 003, May, Melbourne. [2] Crespo Márquez, A. 2007. The maintenance management Framework. Models and methods for complex systems maintenance, Springer Verlag, London. [3] Vagliasindi F. 1989. Gestire la manutenzione. Perche e come, Franco Angeli, Milano. [4] Wireman T. 1998. Developing performance indicators for managing maintenance, Industrial Press, New York. [5] Palmer RD. 1999. Maintenance Planning and Scheduling, McGraw-Hill, New York.

[6] Gelders L, Mannaerts P, Maes J, (1994) “Manufacturing strategy, performance indi- cators and improvement programmes”, International Journal of production Research,

32(4):797-805.

[7] Kaplan RS, Norton DP. 1992. “The Balanced Scorecard - measures that drive per- formance”, Harvard Business Review, 70(1): 71-9. [8] Tsang A, Jardine A, Kolodny H. 1999. “Measuring maintenance performance: a holis- tic approach”, International Journal of Operations and Production Management,

19(7):691-715.

[9] Moubray J. 1997. Reliability-Centred Maintenance (2nd ed.), Butterworth- Heinemann, Oxford. [10] Duffuaa SO. 2000. Mathematical models in maintenance planning and scheduling. In Maintenance, Modelling and Optimization. Ben-Daya M, Duffuaa SO, Raouf A, Edi- tors. Kluwer Academic Publishers, Boston. [11] Campbell JD, Jardine AKS. 2001. Maintenance excellence, Marcel Dekker, New York. [12] Lee J. 2003. “E-manufacturing: fundamental, tools, and transformation”, Robotics and Computer-Integrated Manufacturing, 19(6): 501-507. [13] Mackenzie, J. 1997. “Turn your company's strategy into reality”, Manufacturing Management, January, pp. 6-8. [14] Woodhouse, J. 1993. Managing Industrial Risk, Chapman Hill Inc, London. [15] Blanchard, B.S. and Fabrycky, W.J. 1998. Systems Engineering and Analysis, 3rd ed., Prentice-Hall, Upper Saddle River, NJ. [16] Goffin, K. 2000. “Design for supportability: essential component of new product de- velopment”, Research-Technology Management, Vol. 43, No. 2, pp. 40-7. [17] Markeset, T. and Kumar, U. 2001. “R&M and risk analysis tools in product design to reduce life-cycle cost and improve product attractiveness”, Proceedings of the Annual Reliability and Maintainability Symposium, 22-25 January, Philadelphia, pp. 116-122. [18] Smith, C. and Knezevic, J. 1996. “Achieving quality through supportability: part 1:

concepts and principles”, Journal of Quality in Maintenance Engineering, Vol. 2, No. 2, pp. 21-9. [19] Woodward, D. G. 1997. “Life Cycle Costing - Theory, Information Acquisition and Application”, International Journal of Project Management, Vol. 15, No. 6, pp. 335-

344.

[20] Barlow, R.E., Clarotti, C.A. and Spizzichino, F. 1993. Reliability and Decision Mak- ing, Chapman & Hall, London. [21] Ruff, D. N., and Paasch, R. K. 1993. “Consideration of failure diagnosis in conceptual design of mechanical systems”, Design Theory and Methodology, ASME, New York, pp. 175-187.

97

[22] Woodhouse, J. 1991. “Turning engineers into businessmen”, 14th National Mainte- nance Conference, London. [23] Dowlatshahi, S. 1992. “Product design in a concurrent engineering environment: an optimization approach”, Journal of Production Research, Vol. 30 (8), pp. 1803-1818. [24] Ascher, H. and Feingold, H. 1984. “Repairable System Reliability: Modeling, Infer- ence, Misconceptions and their Causes”, New York, Marcel Dekker. [25] Crow, LH. 1974. “Reliability analysis for complex repairable systems”, Reliability and biometry, Proschan F, Serfling RJ, eds., SIAM, Philadelphia, pp. 379-410. [26] Yañez, M., Joglar, F., Mohammad, M. 2002. “Generalized renewal process for analy- sis of repairable systems with limited failure experience”, Reliability Engineering & System Safety, Vol. 77, pp. 167-180. [27] Modarres, M., Kaminskiy, M. and Krivtsov V. 1999. Reliability engineering and risk analysis. Marcel Dekker Inc., New York. [28] Asiedu, Y. and Gu, P. 1998. “Product lifecycle cost analysis: state of art review”, In- ternational Journal of Production Research, Vol. 36 No. 4, pp. 883-908. [29] Hurtado, J.L., Joglar, F.and Modarres, M. 2005. “Generalized Renewal Process: Mod- els, Parameter Estimation and Applications to Maintenance Problems”, International Journal on Performability Engineering, Vol. 1, No. 1, paper 3, pp. 37-50. [30] Tijms, H.C. 1986. Stochastic Modelling and Analysis, Wiley and Sons, New York, NY. [31] Karyagina, M., Wong, W., Vlacic, L. 1998. “Life cycle cost modeling using marked point processes”, Reliability Engineering & System Safety, Vol. 59, pp. 291-298. [32] Vasiliy, V. 2007. “Recent advances in theory and applications of stochastic point pro- cess model in reliability engineering”, Reliability Engineering & System Safety, Vol. 92, No. 5, pp.549-551 [33] Martorell, S., Carlos, S., Sánchez, A. and Serradell, V. 2000. “Constrained optimiza- tion of test intervals using a steady-state genetic algorithm”. Reliability Engineering & System Safety 67, 215 232. [34] Martorell, S., Villanueva, J.F., Nebot, Y., Carlos, S., Sánchez, A., Pitarch, J.L. and Serradell, V. 2005. “RAMS+C informed decision-making with application to multi- objective optimization of technical specifications and maintenance using genetic algo- rithms”. Reliability Engineering & System Safety 87, 65 75. [35] Barringer, H. Paul and David P. Weber. 1997. “Life Cycle Cost & Reliability for Pro- cess Equipment”, 8th Annual ENERGY WEEK Conference & Exhibition, George R. Brown Convention Center, Houston, Texas, Organized by American Petroleum Insti- tute. [36] Barringer, H. Paul and David P. Weber. 1996. “Life Cycle Cost Tutorial”, Fifth Inter- national Conference on Process Plant Reliability, Gulf Publishing Company, Hou- ston, TX. [37] Kaminskiy M and Krivtsov V. 1998. “A Monte Carlo approach to repairable system reliability analysis”, Probabilistic safety assessment and management, Springer, pp.

1063-1068.

[38] Elsayed, E. A. 1996. Reliability Engineering, Addison Wesley Longman INC, New York. [39] Ireson, W. Grant, Clyde F. Coombs Jr., Richard Y. Moss. 1996. Handbook of Relia- bility Engineering and Management, 2nd edition, McGraw-Hill, New York. [40] Scarf, P.A. 1997. “On the application of mathematical models in maintenance”, Euro- pean Journal of Operational Research, Vol. 99, No. 3, pp. 493-506.

98

[41] Ebeling, C. 1997. Reliability and Maintainability Engineering, McGraw Hill Compa- nies, USA. [42] Dhillon, B.S. 1999. Engineering Maintainability: How to Design for Reliability and Easy Maintenance, Gulf, Houston, TX. [43] Roca, J.L. 1987. “An approach in the life cycle costing of an integrated logistic sup- port”, Microelectronics and Reliability, Vol. 27, No. 1, pp. 25-27. [44] Kijima, M., Sumita, N. 1987. “A useful generalization of renewal theory: counting process governed by non-negative Markovian increments”, Journal Appl. Prob., Vol. 23, pp. 71-88. [45] Kijima, M. 1997. “Markov processes for stochastic modeling”, Chapman &Hall, Lon- don. [46] Bloch-Mercier, S. 2000. “Stationary availability of a semi-Markov system with ran- dom maintenance”, Applied Stochastic Models in Business and Industry, 16, pp. 219-

234.

[47] Limnios, N. and Nikulin, M. 2000. Recent advances in reliability theory: methodolo-

gy, practice, and inference, Birkhauser, Boston.

[48] Xie, M. and Ho, S.L. 1999. Analysis of repairable system failure data using time se- ries models”, Journal of Quality in Maintenance Engineering, Vol. 5 Iss: 1, pp.50

61.

[49] Pham, H. 1999. Recent advances in reliability and quality engineering, World Scien- tific Publishing CO, Singapore. [50] Knafl, G.J. and Morgan, J. 1996. “Solving ML Equations for 2-Parameter Poisson- Process Models for Ungrouped Software-Failure Data”, IEEE Transactions on Reli- ability, Vol. 45, No. 1, pp. 42 - 53. [51] Goel, H; Grievink, J. & Weijnen, M. 2003. Integrated optimal reliable design, production, and maintenance planning for multipurpose process plants”, Computers & chemical engineering, 27 (11), pp. 1543-1555. [52] Ajah, A; Grievink, J; Herder, P. and Weijnen, M. 2007. Adaptive control approach in modeling life-cycle maintenance policy selection and optimization during infrastruc- ture systems conceptual design & operation”, Computer Aided Chemical Engineer- ing, Volume 24, pp. 1145-1150.

Corresponding author

A.

Crespo Márquez can be contacted at: adolfo@esi.us.es

C.

Parra Márquez can be contacted at: parrac37@yahoo.com

99