Definition

DefinitionDefinition

Reliability is a general quality of an object – an ability to perform a desired function, sustaining the values of rated operational indicators in given limits and time according to given technical conditions.

Reliability is probability that an activity of an appliance in given time and given operation conditions will be adequate to its purpose.

EIA (Electronic Industry Association, USA)

The Basic Reliability CalculationsThe Basic Reliability Calculations

Reliability Calculations

1. Reliability of single parts of networks in the time of production of project documentation

2. Reliability of already operated networks

3. Reliability in the area of control of electric power system operation

failure rate [ year-1]

mean time of failure [ h ]

probability of failure-free run R [ - ]

probability of failure Q [ - ]

mean time between failures tS [ h ]

Restored x Not restored objectsMean time between failures x Mean time to failure

Numerical Representation of Reliability

(Classical - reliability of elements)

Global Indices of Reliability

– Outage rate - SAIFI average system outage rate

(number of outages/year/consumer)

– Total time of all outages - SAIDI average system outage time

(min/year/consumer)

– Time of one outage - CAIDI average outage time at a customer

(min/outage)

(Reliability of electric energy supply)

Bathtub curve

I II III t

Early failure period Constant failure rate period

Wear-out failure period

The relation between the function of reliability and failure rate is:

For failure rate it is valid:

Division of Probability of Failure

Exponential division

Exponential rule of failure 

Poisson’s division

Weibull’s deal

If k = 0, there is probability of no failure, therefore probability of failure-free running.

Obtaining of input values for reliability calculations

Calculation of Reliability in Electricity Industry

A priori reliability – determination of reliability quantities from data of a producer.  Empirical reliability – monitoring of failures in electricity industry. 

The empirical method is mostly used for obtaining the input values for reliability calculations, because an application of a priori reliability method requires different attitude to every element of electricity system.

Analysis of Distribution Network Failure Databases

• Exclusive outage databases had been on the rise since 1975 in the former Czechoslovakia.

• Unfortunately, database building stopped in1990 because of political and social changes.

• Thanks to the expert group CIRED Czech distributors opted for unified monitoring of global reliability indices and the reliability of selected pieces of equipment in 1999 again.

• Data for the reliability computation is centrally processed and analyzed at the VSB - Technical University of Ostrava since the year 2000.

• Collected data are often heterogeneous.• It is necessary to solve the storage, indexing, and also transformation

of such data.• We need to create a common relational scheme for the storage of the

data, a new relation makes the querying and analysis possible.

• Database range

Region1 Region2 Region3 Region4 Region5 Region6 Region7 Region82000 - - - 1 - 12 - - 1 - 12 1 - 122001 1 - 12 - 1 - 12 1 - 12 - 1 - 12 1 - 12 1 - 122002 1 - 12 - 1 - 12 1 - 12 - 1 - 12 1 - 12 1 - 122003 1 - 12 - 1 - 12 1 - 12 - 1 - 12 1 - 12 1 - 122004 1 - 12 - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 122005 - - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 122006 - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 122007 - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 122008 - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 122009 - 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12 1 - 12

Heterogeneous Data

• We developed a common relational schema.Order Attribute Description

1 Distribution company Unique number of a distributor2 Outage identification Unique number of an outage3 Outage type Accidental, planned or forced4 Distribution point Type of substation: single, double busbar substation, . . .5 Distribution area Specification of the location6 Network type Insulated system, resonant grounded neutral system, . ..7 Network voltage 0.4 kV, 22 kV, . . .8 Equipment voltage 0.4 kV, 22 kV, . . .9 Original outage identification Unique number of the outage cause

10 Outage cause Foreign influences, causes before starting operation, . . .11 Equipment type Overhead line, underground line, . . .12 Failed equipment Specific equipment: conductor, switch, pole, fuse, . . .13 Type of equipment Further specification: wooden pole, steel pole, . . .14 Amount of failed equipment15 Short circuit type One-line-to-ground fault, ground fault, line-to-line grounded fault, . . .16 Producer Siemens, ABB, . . .17 Production year Production year of the equipment18 Outage start time19 First manipulation Failure limitation start time20 End of manipulations Failure limitation time21 End of outage Supply renewal time for all consumers22 End of equipment failure Equipment reparation end time23 Time of dead earth24 Unsupplied power at the outage start25 Unsupplied power at the end of manipulations26 Unsupplied distribution transformers at the outage start27 Unsupplied distribution transformers at the end of manipulations28 Unsupplied customers at the outage start29 Unsupplied customers at the end of manipulations30 Number of unsupplied customers multiplied by time of their outage31 Failure type With or without equipment fail

Failure rate

(year-

1)

N = number of failures (-) Z = number of elements of the given type

in the network (-) P = the considered period (year)

Results

PZNλ

Mean duration of the failure

(h)

N = number of failures (-) ti = the considered period (h)

Results

N

tτ

N

ii

1

Results

• The value tendency of reliability indices of the 22 kV cable

Cable 22 kV

0

1

2

3

4

5

6

7

200020012002 20032004 200520062007 20082009 Total

Mea

n tim

e to

repa

ir (h

)

00.010.020.030.040.050.060.070.080.090.1

Failu

re ra

te (y

ear

-1)

(h) (year-1)

Results

• Comparison with methodology ČEZ 22/80

ČEZ22/80

22 kV cable (year-1) 14.5 5.480 (h) 215 4.034

22 kV overhead line (year-1) 14 3.018 (h) 3 4.163

110 kV overhead line (year-1) 5.2 0.370 (h) 3.5 3.992

MV/LV transformer (year-1) 0.03 0.007 (h) 2500 4.315

110 kV/MV transformer (year-1) 0.04 0.059 (h) 1300 0.480

22 kV circuit breaker (year-1) 0.015 0.016 (h) 30 64.179

110 kV circuit breaker (year-1) 0.01 0.052 (h) 100 47.425

Equipment 2000 - 2009

Results

• Division of failures according to their causes

Četnosti příčin událostí

1

2

3

4

8

9

Causes before starting operationOperation and maintenance causesForeign influencesForced outageCause not explainedOther causes

The Main Calculation Methods of Reliability

Method of reliabilty schemes

Department of electrical power engineering

Markov‘s processes

Simulative methods

-         make-up of reliability diagram,-         assignment of relevant reliability quantities to single elements,-         simplification of reliability diagram towards one element, 

Advantages:

-         considered systems do not have to really exist as yet,-         procedure of solving is well-arranged and not exacting concerning

mathematics,-         mathematical procedure does not require iterative calculation,-       accuracy of results depends only on the accuracy of input parameters

of calculation. 

Disadvantages:

-         it is impossible to pursue power balance of network,-         „T“ type bay can be modelled only approximately.

Method of Reliability Schemes

Probability of failure-free run:

Rule of Multiplication of Probabilities :

P(A) probability of occurrence of A

P(B) probability of occurrence of B

Series systems

A failure of one element leads to a failure of a system.

A failure of a system occurs when all elements have a failure

Probability of a failure:

Probability of failure-free run: 

Parallel Systems

Simplified

Probability of failure-free run:

Methodology of Calculation of Reliability according to ČEZ 22/80 Regulation

Advantages:

- takes into account maintenance outages,- enables to include manipulation into calculation as well, takes so-called coldreserves into account. The disadvantage is that this methodology does not include so-called coordination of maintenance.

These operating states are considered with calculation:- operation,- failure outage,- maintenance outage. Supposition and simplification: - the effect of weather on failure rate and repair rate is not taken into account,- exponential division of distributive function of time of failures and repairs for all elements of electric network is taken into account,- average data are started from.

 Mean times of outages of a two-elements system:

Series connection of elementsFor this circuit with two elements it holds:

P ... Failure rate [year-1]

U ... Maintenance rate [year-1]

... Outage rate (maintenance + repair) [year-1]

Maintenance outage cannot occur at this connection, because at the failure of one element maintenance of another element will not begin.

Parallel connection of elements – hot reserve

Failure rate:

Mean time of failure:

Parallel circuit of elements – cold reserve

…. Manipulation time [h]

Simulation Methods of Calculation of Reliability

It is necessary to know the intensity of outages and mean time of outages of all the elements of a system.Simulation - numerical method which resides in experimenting with mathematical models of real systems on numerical computers.

Advantages:

- considered systems do not have to really exist as yet,- considered systems can be too complicated for using analytical methods,- simulation makes possible study of behaviour of systems in real, accelerated, or

retarded time. The second possibility is the most important in this case, because the processes of outage of elements and their re-introduction into operation are very slow. It would be very inefficient to study them in any other time but accelerated.- with simulation it is possible to verify results obtained by other independant processes,- possibility of modelling „T“ type bays - simple power balance of a diagram is carried out, outage is always simulated at

overloaded elements.

- construction of a useful simulation model is very time-consuming. Mostly several variants of a model are needed.

- simulation is a numerical method, so a solution of certain problem cannot be generally transferred on analogous problems.

- the results obtained from stochastic simulation models are values of accidental quantities, and it would be very computer time-consuming if their and it would be very computer time-consuming if their

accuracy should be increasedaccuracy should be increased

- precision of results depends on the number of iterations,

- the needed number of iterations depends on the extent of the solved network and on the required precision.

Disadvantages of simulation methods: