06465758

1

Abstract--The implementation of Reliability Centered

Maintenance (RCM) method can assist in achieving cost

effectiveness, and allow a greater understanding of the risk level that organizations presently manage. This systematic method

identifies the most applicable and effective maintenance plan to

avoid each failure mode of the equipment. The work presented

here describes how the RCM methodology can be applied to a digital protective relay installed in a primary distribution

substation. RCM can estimate failure before it occurs with

predictive testing and inspection tools, as well as organize maintenance plans according to the impacts of protective relay

failure on system reliability.

Keywords--Reliability Centered Maintenance (RCM),

Reliability, Distribution Power System, Periodic Preventive

Maintenance.

I. INTRODUCTION

ISTRIBUTION power system operators are increasingly

required to supply customers with high quality and

continuous service. The protection scheme’s responsibility is

to quickly clear faults and limit any damage to distribution

power equipment [2]. It keeps the power system stable by

isolating only the components that are under fault, whilst

leaving as much of the network as possible still in operation.

The technology utilized in protection schemes must be reliable

throughout their life span. Reliability can be improved by an

adequate maintenance plan. A maintenance plan should ensure

high reliability of protection scheme with minimized cost.

Nowadays, digital protective relays are used in protection,

supervision and control of electrical equipment. These

Intelligent Electronic Devices can manage several functions of

protection and automation, measurements of currents and

voltages with great accuracy. Normally, the distribution

operators have implemented a periodic preventive

maintenance practice in these digital protective relays, which

is essential in preventing any unusual performance of the

system where these protection relays operate.

H. Tavares is with FE/UP (Faculty of Engineering, University of Porto),

Porto, Portugal (e-mail: [email protected]).

H. Leite is with INESC TEC and FE/UP (Faculty of Engineering,

University of Porto), Porto, Portugal (e-mail: [email protected]).

A. Pinto, P. Vidal and J. Santos are with the Portuguese Network

Operator, EDP Distribuição - Energia S.A., Porto, Portugal (e-mail:

[email protected], [email protected] and [email protected]).

The authors would like to acknowledge the cooperation of the Laboratory

of Systems Protection at DEEC/FEUP (University of Porto).

In general, preventive maintenance programs are

established according to different criteria, depending on the

Utility or even on the Maintenance Department to which the

protective relay belongs to. These options are based on

manufacturer’s recommendations, analysis of fault statistics,

the relative importance of the equipment protected, and

maintenance experience. Utilities try with periodic preventive

maintenance to decrease failures rates of protection schemes.

However, its periodicity of the maintenance plan leads to high

costs. For the sake of cost management, a maintenance plan

must ensure that the maintenance is at the right time and with

the right activity [6]. Thus, RCM methodology identifies

functions, functional failures, failure modes, their effects and

also consequences of the failures for each component in its

operating context. This approach recognizes that the value of

every technical maintenance activity and provides rules to

decide the most adequate maintenance task to avoid each

failure mode [3].

The work here presented suggests a plan of maintenance

based on RCM methodology for a digital protective relay

installed in a distribution substation protecting a medium

voltage cable.

II. RCM METHODOLOGY AND CHALLENGES TO IMPLEMENT ON

A DIGITAL PROTECTIVE RELAY

The RCM process entails answering seven questions about

the protective relay [3]. They are:

1. What are the functions and associated performance

standards of the relay in its present operating context?

2. In what way does it fail to fulfill its functions?

3. What causes each functional failure?

4. What happens when each failure occurs?

5. In what way does each failure matter?

6. What can be done to predict or prevent each failure?

7. What should be done if a suitable proactive task cannot

be found?

By answering the first five questions, it is possible to obtain

detail information about the protective relay in its operating

context and identify the dominant causes of failures. Question

six and seven determine the maintenance procedure for each

failure mode of the relay [5].

The maintenance plan can be either or both proactive

and/or reactive. Proactive plans are undertaken before a failure

occurs, whereas reactive plan deals with failed state which

prevents equipment from getting into a failed state. The

maintenance tasks considered in this work are the following:

Applying Reliability Centered Maintenance to a

Digital Protective Relay H. Tavares, H. Leite, A. Pinto, P. Vidal and J. Santos

D

2012 3rd IEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe), Berlin

978-1-4673-2597-4/12/$31.00 ©2012 IEEE

2

- T1: Preventive condition-based maintenance;

- T2: Periodic preventive maintenance (overhauling

of components);

- T3: Periodic preventive maintenance (replace

components);

- T4: Corrective maintenance;

- T5: Redesign of functional modes;

- T6: Finding hidden failures;

- T7: Combination of T1, T2 and T3;

To establish the most appropriate maintenance task in order

to avoid each failure mode of the digital protective relay

decision criterion is applied to each of the failure mode of the

digital protective relay. This criterion is based on the

consequences of the digital protective relay failures, reliability

of the equipment, technical, financial resources and risk

tolerance. The Maintenance Department at the Portuguese

Distribution Network Operator (DSO) has lately

registered the causes of protective relays failures and their

respective repair costs. Therefore, this work applies the RCM

method with the currently available information, to address

functional needs of the protection relay with an adequate

maintenance plan.

III. DIGITAL PROTECTIVE RELAYS INSTALLED AND ACTUAL

MAINTENANCE POLICY

A digital relay consists of the following main parts:

processor, analogue input system, digital output system,

independent power supply, communications and user

interface. These are to be assessed as possible functional

failures.

Normally at the Maintenance Department in DSO the

protective relays are subjected to a five year planned

maintenance programme where all the functional functions are

tested and visual inspection is made.

The length of “effective age” in year of the digital

protective relays installed at the primary substation which has

not failed yet is showed in Fig. 1.

Fig. 1. The “effective age” in years of the installed digital protective relays

that never failed.

So that, 121 digital protective relays have been put under a

planned maintenance and 51 of them already had two planned

maintenance. Fig. 2 shows the number of digital protective

relays that failed versus the “effective age” of its failure. Note

that, 4 of them fail before the first five year maintenance plan

and 12 of them fail even after the planned maintenance.

Fig. 2. The number of the digital protective relays failure by their “effective

age”.

An analysis was then performed to establish a correlation

between a condition index and the relays “effective age” and

no survivor curve was identified, i.e. no relationship was

found between the relays age and the percentage of the

equipment that has failed or retired.

IV. RCM METHODOLOGY SUGGESTED

In order to find an optimal maintenance plan, it is necessary

to identify maintenance tasks in order to avoid various failure

modes from the digital protective relay. The maintenance plan

has the objective of ensuring high reliability at a minimum

cost.

Firstly, the RCM methodology identifies the “failure

modes” which are reasonable and likely cause of the digital

protective relay failure. A failure mode is any event which

causes a functional failure. Secondly, an “effect analysis” is

performed for each of those identified functional failure. A

description of the failure effects outlines what happens when a

failure mode occurs. It should include all the information

needed to support evaluating the consequences of those

failures. The effects of the failure are described in terms of

non-evidence or evidence that the failure has occurred, threat

to safety, environmental hazards, operational or non-

operational effects [7]. In this work the cost analysis were

quantified and present in Table I. It is necessary to include

when operational or non-operational failure modes occur.

TABLE I

COSTS CONSIDERED

Costs Quantified Variables

Scheduled maintenance

Staff required

Repair time (h)

Staff per hour (€/h)

Staff transportation, repair equipment, etc...

Failure repair Digital protective relay repair cost

Operational consequences Expected energy not supplied

A. The RCM Decision Process

The RCM decision worksheet is illustrated in Fig. 3, [4].

The reliability indicators of the digital protective relay failure

modes are the Mean Time between Failures (MTBF) and

Mean Time to Repair (MTTR), [8].

9

28

91

12

4855

11 7

39

12

0

20

40

60

80

100

2 3 4 5 6 7 8 9 11 13

1

2

1

6

1

3

2

1

0

1

2

3

4

5

6

7

2 3 5 6 7 8 9 11

3

Is a task to detect

whether the failure is

occurring or about to

occur applicable?

Is a scheduled

restoration task to

reduce the failure

rate applicable?

T1

Is a scheduled discard

task to reduce the

failure rate

applicable?

Is a failure-finding

task to detect the

failure applicable?

Could the multiple

failure affect safety or

the environment?

Is a combination of tasks

to avoid failures

aplicable?

T2

T3

T6

T5 T4

Does the failure mode

have safety

consequences?


have environmental

consequences?


have operational

consequences?

Will the loss of

function caused by this

failure mode on its

own become evident

to the operating crew

under normal

circunstances?

Yes

No

No

Yes

No

No

No

No

No

No

Yes

Yes

Yes

Yes

Yes

Is a task to detect



occur applicable?

Is a scheduled

restoration task to

avoid failures

applicable?

T1


task to avoid failures

applicable?

T2

T3

No

No

No

Yes

Yes

Yes

T7 T4

NoYes

Is a task to detect



occur applicable?

Is a scheduled

restoration task to

reduce the failure

rate applicable?

T1


task to reduce the

failure rate

applicable?

T2

No

No

Yes

Yes

Yes No

T3 T4

NoYes

Fig. 3. The RCM decision logic diagram.

V. APPLYING RELIABILITY CENTERED MAINTENANCE TO A

DIGITAL PROTECTIVE RELAY

As an example of all function from the digital protective

relay, Table II select seven of those functions.

TABLE II

SEVEN DIGITAL PROTECTIVE RELAY FUNCTIONS

Protective elements

1 Instantaneous overcurrent (50)

2 Instantaneous overcurrent for neutral (50N)

3 Under voltage (27)

Automation element

4 Reclosing relay (79)

5 High voltage automatic reset overload relay

Additional elements

6 Disturbance recorders

7 Events log

From Table II, three relay functions were selected (i.e. 1, 4

and 7) as an example to identify the functional failures [9].

TABLE III

EXAMPLE OF THE FUNCTIONAL FAILURES

Functions

(F) Functional Failures (FF)

1

A The protective relay does not operate when a

network fault occurs (incorrect operation)

B

The protective relay does operate when a

network fault does not occurs(unintentional

operation)

4 A Incorrect reclosing operations (acts erroneously)

B Digital output system does not work

7 A Events log or state-change event are not

recorded or stored

From the functional failure, Table IV shows the identified

failure modes from the protective relay.

TABLE IV

FAILURE MODES

F FF Failure Modes (FM)

1 A

1 Microprocessor does not perform the

protection algorithm

2 Power supply unit burned

3 Digital output system does not work, the

contacts are interlocked

4

A / D converter does not correctly convert

the analogue signals measures, so that the

protection does not operate accordingly

5 Input digital signals system damaged

4 A 1

Microprocessor does not perform the

protection algorithm

2 Digital output system does not work, the

contacts are interlocked during the reclosing

process

3 Input digital signals system damaged,

circuit breaker change-

7 A 1

Microprocessor does not perform the voltage

values

2 Memory damaged

Table V lists the failure effects of each failure mode.

TABLE V

FAILURE EFFECTS ANALYSIS

F FF FE Failure effects

1 A

1

-Signalling from watchdog

- Possibility of failing the protected equipment and

incrementing the collateral damage

-Increment outage durations well as the

energy non-supplied

2


- Possibility of failing the protected equipment and

incrementing the collateral damage

- Increment outage duration as well as the

energy non-supplied

3

-Possibility of damaging the

-Increment the collateral damage

-Increment outage duration and

non-supplied

4

-Protective relay mal

-Increase the expected energy

customers

5

-Protective relay mal operate


non-supplied

4 A

1


-Protective relay mal


non-supplied

2

- Possibility of failing the protected

cable in this example) and increment

collateral damage


non-supplied

3 -Protective relay mal

7 A 1 -Signalling from

2 -Signalling from

Table VI shows the failure consequence

determined from the RCM worksheet showed in Fig. 3

TABLE VI

FAILURE CONSEQUENCES & MAINTENANCE

F FF MF Failure

Consequence

1 A

1 Safety

Operational

2 Safety

Operational

3 Non-operational

4 Non-operational

5 Non-operational

4 A

1 Safety

2 Safety

3 Hidden function

7 A 1 Non-operational

2 Non-operational

A. Comments on Failure Consequences

Twenty nine different operational modes

together responsible for different functional

shows the results of the failure consequences

13.79% are hidden and 86.21% are evident functions

during the reclosing

ocess

Input digital signals system damaged,

-state wrong read

Microprocessor does not perform the voltage

values

Memory damaged

each failure mode.

NALYSIS

Failure effects

Signalling from watchdog

the protected equipment and

the collateral damage

well as the expected

supplied


Possibility of failing the protected equipment and

the collateral damage

Increment outage duration as well as the expected

supplied

Possibility of damaging the protected equipment

ncrement the collateral damage

Increment outage duration and the expected energy

supplied

Protective relay mal-operate

energy non-supplied to

mers

relay mal operate

and the expected energy

supplied



Increment outage duration and the expected energy

supplied

Possibility of failing the protected equipment (MV

and incrementing the

collateral damage

Increment outage duration and expected energy

supplied


from watchdog

from watchdog

failure consequences maintenance task,

from the RCM worksheet showed in Fig. 3.

AINTENANCE TASK

Maintenance task number

T4

T4

T6

T6

T6

T4

T6

T6

T4

T4

different operational modes were identified all

functional failures. Fig. 4

shows the results of the failure consequences. Note that

evident functions.

Fig. 4. The failure consequences.

The evident functions

maintenance task, while the

plan of finding hidden failures task.

relationship between equipment age and the percentage of the

equipment population that has failed or retired

preventive maintenance may not be

apply. A finding hidden technique should be carried o

Scheduled failure-finding entails checking hidden function

regular intervals. A failure finding task on a digital protective

relay is very important for a maintenance plan as digital

protective relay are characterized for its high reliability. The

failure finding task consists o

situations of faults, allow

protective relay works when required. To determine failure

finding intervals, the failure rates of digital protective relay

functions must be known, as well as the occasions when

failures of the protected equipment

protective relays are in a failed state

determine the failure-finding intervals

used [3]:

2

MFM

TEDM

TIVEM

FFI

××

=

Where: TIVE

M is the mean time between failures

protective device in years; M

failures of the protected function in years;

time between multiple failures in years

As aforementioned, multiple

protected equipment fails and

a failed state. The failure of the protected equipment and the

protective relay happens independently.

was considered that the protected equipment is a medium

voltage underground cable with a conditional probability of

0.00613failures/year [11] and a conditional probability of

failure of digital protective relay of

As it is known the conditional probability of both

relay and protected equipment, with the help of the

Chronological Monte-Carlo technique

probability of multiple failure of 0.000026 of one year, after

placing 70680 random points

4

functions should have a corrective

, while the hidden functions should have a

finding hidden failures task. As there was no

equipment age and the percentage of the

equipment population that has failed or retired, periodic

maintenance may not be a suitable technique to

apply. A finding hidden technique should be carried out.

finding entails checking hidden functions at

A failure finding task on a digital protective

relay is very important for a maintenance plan as digital

protective relay are characterized for its high reliability. The

consists of testing and simulating different

, allows examining whether digital

protective relay works when required. To determine failure

finding intervals, the failure rates of digital protective relay

functions must be known, as well as the occasions when

the protected equipment occurs and digital

in a failed state [10]. In order to

finding intervals (FFI) expression (1) is

)1(

is the mean time between failures of the

TEDM is the mean time between

ures of the protected function in years; MF

M is the mean

failures in years;

ultiple failures occur when the

and the digital protective relay is in

failure of the protected equipment and the

independently. In this example, it

was considered that the protected equipment is a medium

derground cable with a conditional probability of

] and a conditional probability of

digital protective relay of 0.00333 failures/year [1].

As it is known the conditional probability of both protective

relay and protected equipment, with the help of the Non-

Carlo technique it was determining the

probability of multiple failure of 0.000026 of one year, after

placing 70680 random points. The estimated probability for

5

multiple failure is within the variance of 0.01% and 95% of

confidence interval.

VI. FINAL REMARKS & CONTRIBUTION FROM THIS WORK

This work shows the application of the RCM methodology

to a digital protective relay which protects a medium voltage

underground cable. Nineteen functions from the protective

relay are evaluated, where functional failures, effects and their

consequences are identified. With this analysis, the

maintenance staff may spot functional failures quicker and

reduce the repair time and its associated costs. With RCM

methodology it is possible to understand better the impacts of

its functional failures on the entire network.

Through signalling from the protective relay “watchdog

function”, it is possible to identify 86% of all failures. Thus

the corrective maintenance task should be applied. Based on

the available source data, when a protective relay is protecting

an underground cable at a substation, the planned maintenance

programme should be of 2.5 years for hidden failures.

Nevertheless it would be necessary to analyse a larger sample

of data in order to make this study conclusive.

Applying RCM to a digital protective relay from

distribution substation, the Maintenance department from an

Electrical Utility can:

1. Obtaining the component failure rate of each digital

protective relay, which can help to establish a more

efficient maintenance plan;

2. Improving protection reliability and making ideal

decisions to improve a protection system, as the

maintenance criteria is based on the implication of

power system dependability and security.

3. Understanding better the digital protective relay

behavior in its operational context which can help

selecting the most effective maintenance activity that

would prevent each failure mode;

4. Revising the current maintenance procedures during a

digital protective relay life span and adjusting these

procedures according to the relay’s operational

performance.

VII. REFERENCES

[1] R. Kirby and R. Schwartz, "Microprocessor-based protective

relays," Industry Applications Magazine, IEEE, vol.15, no.5, pp.43-50,

September-October 2009.

[2] Juan M. Gers and Edward J. Holmes, “Protection of electricity of

distribution networks”, 2nd ed., Stevenage: Institution of Electrical

Engineers, 2004, ISBN: 0-86341-357-9, pp. 1-43.

[3] J. Moubray, “Reliability centered maintenance”, 2nd ed., Amsterdam:

Elsevier Butterworth-Heinemann, 1997, ISBN: 0-7506-3358-1, pp.18-

180.

[4] J. Moubray, "Reliability centered maintenance”, 2nd ed., Amsterdam:

Elsevier Butterworth-Heinemann, 1997, ISBN: 0-7506-3358-1, pp. 200-

201.

[5] Ting Chan, Chen-Ching Liu and Jong-Woong Choe, "Implementation of

reliability-centered maintenance for circuit breakers," Power

Engineering Society General Meeting, 2005 IEEE, pp. 684- 690.

[6] L. Bertling, R. Eriksson and R.N. Allan, "Relation between preventive

maintenance and reliability for a cost-effective distribution

system," Power Tech Proceedings, 2001 IEEE Porto, pp. 6.

[7] Zao-jie Kong and Ying-xin Meng, "Research on Probability Reliability-

Centered Maintenance model based on Requirement-Oriented

Maintenance," Industrial Engineering and Engineering Management

(IE&EM), 2011 IEEE 18Th International Conference , pp. 1155-1158.

[8] M. Fotuhi-Firuzabad, S. Afshar, D. Farrokhzad, and Jaeseok Choi,

“Reliability centered maintenance program initiation on electric

distribution networks," Transmission & Distribution Conference &

Exposition: Asia and Pacific, 2009, pp.1-4.

[9] Ding Maosheng, Wang Gang and Li Xiaohua, "Reliability analysis of

digital relay," Developments in Power System Protection, 2004, Eighth

IEE International Conference, pp. 268- 271.

[10] I.P. Siqueira, "Optimum reliability-centered maintenance task

frequencies for power system equipments," Probabilistic Methods

Applied to Power Systems, 2004 International Conference, pp. 162-167.

[11] IEEE Recommended Practice for the Design of Reliable Industrial and

Commercial Power Systems, IEEE Standard 493-1990, 1991, pp. 114.

VIII. BIOGRAPHIES

Helder Tavares was born in Vale de Cambra, Portugal, on February 3, 1989.

He received the M.Sc. degree in electrical engineering from the Faculty of

Engineering of the University of Porto (FEUP), Portugal, in 2012.

Helder Leite received his Electrical Engineering degree from University of

Porto, Portugal in 2000 and the PhD degree in Electrical Engineering from

The University of Manchester, UK, in 2004. H. Leite joined university of

Porto as a Lecturer in 2005. His research interests include Distribution

Generation Integration, Electric Power Systems and Power Systems

Protection.

Alberto Pinto received the M.Sc. degree from the Faculty of Engineering of

the University of Porto (FEUP), Porto, Portugal, in 2008, in electrical

engineering. Since 2009, he is a specialist engineer in the Operation and

Maintenance Department at EDP Distribuição-Energia S.A, the Portuguese

Distribution System Operator.

Pedro Borges Vidal received his Electrical Engineering and Computers

degree from Faculty of Engineering of the University of Porto (FEUP), in

1998 and has a master degree in business administration (MBA). Since June,

2011, he is responsible for OMPRT (Operation and Maintenance Department-

Porto) in EDP Distribuição-Energia S.A.

José Miguel Santos received his Electrical Engineering and Computers

degree from Faculty of Engineering of the University of Porto (FEUP), in

1995 and has a MBA in business administration. He is now a subdirector for

Automation and Telecontrol in EDP Distribuição- Energia S.A.

06465758

Documents

Transcript of 06465758