Literature Search For Reliability Data of Components in Electric Distribution Networks

Literature Search for Reliability
Data of Components in Electric

Distribution Networks
by
M.H.J. Bollen
EUT Report 93-E-276
ISBN 90-6144-276-1
August 1993
Eindhoven University of Technology Research Reports
ISSN 0167-9708
EINDHOVEN UNIVERSITY OF TECHNOLOGY
Faculty of Electrical Engineering
Eindhoven, The Netherlands
LITERATURE SEARCH FOR RELIABILITY DATA
Coden:TEUEDE
OF COMPONENTS IN ELECTRIC DISTRIBUTION NETWORKS
by
M.H.J. Bollen
EUT Report 93-E-276
ISBN 90-6144-276-1
Eindhoven
August 1993
CIP-DATA KONINKLlJKE BIBLIOTHEEK, DEN HAAG
Bollen, M.H.J.
Literature search for reliab.ility data of components in
electric distribution networks I by M.H.J. Bollen. -
Eindhoven: Eindhoven University of Technology, Faculty
of Electrical Engineering. - Fig., tab. - (EUT report,
ISSN 0167-9708 ; 93-E-276)
With ref.
ISBN 90-6144-276-1
NUGI832
Subject headings: power system reliability I distribution
networks.
Abstract.
This report gives the result of a literature search for component lifetimes for use in
reliability studies of distribution networks. Data are given for power transformers, circuit
breakers and switches, protective equipment, fuses, voltage and current transformers,
generators, "uninterruptable" power supplies, cables and acessories, busbars and large
motors. For each chapter the available lifetime data is divided into recommended values,
data from surveys, data used in reliability studies and ageing data. Also data for duration
of the restore are given. Each chapter results in lifetime values that appear to be
reasonable.
Keywords: power system reliability I distribution networks.
Bollen, M.H.J.
LITERATURE SEARCH FOR RELIABILITY DATA OF COMPONENTS IN ELECTRIC
DISTRIBUTION NETWORKS.
Eindhoven: Faculty of Electrical Engineering, Eindhoven University of Technology, 1993.
EUT Report 93-E-276
Address of the author.
M.H.J. Bollen
University of Manchester Institute of Science and Technolgy
Department of Electrical Engineering
P.O. Box 88
Manchester M60 1 aD
United Kingdom
iii
CONTENTS OF THIS REPORT
pag
1. Introduction 1
1.1. Aim and contents of this report 1
1.2. Terms used in this report 2
1.3. Conclusions 3
2. Power transformers 4
2.1. Recommended values 4
2.2. Data from surveys 6
2.3. Data used in reliability studies 11
2.4. Ageing data 16
2.5. Conclusions 19
3. Circuit breakers and switches 21
3.4. Ageing data 41
3.5. Conclusions 44
4. Protective equipment; general and relays 47
4.4. Ageing data 57
4.5. Conclusions 59
5. Fuses 60
5.4. Ageing data 63
5.5. Conclusions 64
6. Voltage and current transformers 65
6.4. Ageing data 68
6.5. Conclusions 69
7. Generators 70
7.4. Ageing data 81
7.5. Conclusions 84
iv
pag
8. "uninterruptable" power supplies 86
8.4. Ageing data 92
8.5. Conclusions 93
9. Cables and accessories 94
9.4. Ageing data 114
9.5. Conclusions 116
10. Busbars. 120
10.4. Ageing data 126
11. Large motors 129
11 .4. Ageing data 134
12. References 136
v
1. INTRODUCTION
1.1. Aim and contents of this report
This report is the result of an extensive literature search after failure data of components
in distribution systems. It started as part of a project aimed at the incorporation of
reliability aspects in the design of industrial power systems. During the execution of this
project it felt that there was a lack of data on failure rates (lifetimes) of industrial power
system components. It was then decided to perfrom this literature search after
component reliability data.
This report is divided in 10 more chapters, each dealing with one type of power system
components. Every chapter is further divided into five sections: recommended values;
data from surveys; data about ageing; conclusions. Within each of these sections the
references are sorted to year of publication.
Not much information has been found on ageing of components. I therefore present the
available information without drawing any conclusions. Also on repair and replace times
no conclusions are given. These times are too much dependent on the specific situation
to give recommended values. Furthermore, they can easily be assessed in an industrial
environment, so there is less need for recommended values. I do however present the
available data. These interested are able to draw there own conclusions.
-1- Introduction
1.2. Terms used in this report
In this report the ETTF (Expected Time To Failure) has been used to describe the failure
behaviour of a component. For surveys it has been defined as
ETTF = #components * #years
#failures
It is related to the failure rate A through
(1 )
(2)
An ETTF of 1000 years thus does not mean that this kind of component is expected to
last a thousand years but that each year 1 %0 of the population will fail during the next
few years (20 or 30 at most).
From a survey the ETTF can never be determined exactly. Where possible a confidence
interval has been given in this report. The 95% confidence interval (designated as c.L)
is approximated through the following expression (for a survey finding n failures in a
population of N component-years):
. [N N] (3)
c. ~ . = n+2*/Ii' n-2*/Ii .
The repair proces is described through the term "repair time" which refers to the
expected repair time which is again, in case of survey, an average repair time. No
confidence interval is given for the repair time. The values should however be used with
the same caution as the ETTF-values.
For some recommeded values the authors gave a range of values [e.g. IEEE, 1983;
Kloeppel et al., 1990]. This is either a spread in the value for different locations and/or
circumstances or a measure of the uncertainty in the recommended value. As it needs
not to be the same as a confidence interval, the original term "range" has been used in
this report.
Some authors distinguish between active and passive failures. The followingm definition
for this is given by Billinton and Allan [1984].
A passive event is a component failure mode that does not cause operation of protection
breakers and therefore does not have an impact on the remaining healthy components.
Service is restored by repairing or replacing the failed component. Examples are open
circuits and inadvertent opening of breakers.
An active event is a component failure mode that causes the operation of the primary
protection zone around the failed component and can therefore cause the removal of
other healthy components and branches from service.
I often took over the terms as used by the authors of the papers I refere to. This causes
a non-standard use of vocabulary. I hope the reader will except this. It is for instance not
always clear what an author means with repair time. Is it the time needed to repair a
component, or the time needed to restore the service. It the latter case switching time
or replacement time seem more appropriate terms. In case the latter terms were used
I took them over.
-2- Introduction
1.3. Conclusions
From the data found during this survey I would suggest the following ETTF values for
components in distribution networks. For the reasoning behind these values, please refer
to the corresponding chapters further on in this report.
MV /L V transformers
MV/MV transformers
HV /MV transformers
MV and LV circuit breakers
disconnect switches
electromagnetic relays
electronic relays (single function)
electronic relay systems
fuses
voltage and current transformers
standby generators
probability of fail-to-start
continuous generators
UPS invertor
rectifier
underground cables (1000 meters)
cable terminations
cable joints
busbars (one section)
large motors
: 500 - 1000 years
: 75 - 100 years
: 40 - 70 years
: 1000 - 5000 years
: 250 - 1000 years
: 250 - 1000 years
: 100 - 200 years
: 10 - 30 years
: 1000 - 5000 years
: 2000 - 3000 years
: 5 - 20 days
: 0.5 - 2 %
: 1 - 3 years
: 0.5 - 2 years
: 10 - 30 years
: 40 - 75 years
: 1000 - 3000 years
: 500 - 2000 years
: 500 - 2000 years
: 15 - 30 years
-3- Introduction
2. POWER TRANSFORMERS
2.1. Recommended values
2.1.1. Green and Bourne [1972) give average component failure-rates for electrical
components. For power transformers they give:
less than 15 kV
15 kV to 33 kV
33 kV to 132 kV
132 kV to 400 kV
: ETTF = 190 years
: ETTF = 57 years
: ETTF = 29 years
: ETTF = 16 years
2.1.2. IEEE standard 500 [lEEE,1983) gives reliability data for components of nuclear-
power stations. The data has been derived by a Delphi-method combined with the results
from several surveys and data bases.
For station service transformers the following data are, recommended (for catastrophic
failures only)
liquid filled single phase transformers
0-600 Volt: ETTF = 650 years; range = < 200, 1300 >
601 V -15 kV: ETTF = 600 years; range = <250,1500>
15 - 40 kV: ETTF = 225 years; range = < 100,600 >
restore time = 1260 hours; range = < 12.8, 3740>
liquid filled three phase transformers.
601 V- 15 kV: ETTF = 875 years; range = < 350, 1850 >
15 - 40 kV: ETTF = 300 years; range = < 140, 1200 >
restore time = 1300 hours; range = < 12.8, 3700>; based on [lEEE,1980)
Dry type single phase transformers
601 V-15 kV: ETTF = 350 years; range = <160,800>
15 - 40 kV: ETTF = 130 years; range = < 50, 500 >
restore time = 28.0 hours; range = <0.5, 720>; based on [lEEE,1980)
Dry type three phase transformers
601 V-15 kV: ETTF = 340 years; range = <140,750>
15 - 40 kV: ETTF = 105 years; range = <80,290>
restore time = 28.0 hours; range = < 0.5, 720>; based on [lEEE,1980)
For substation transformers the following data are recommended (catastrophic failures
only)
Liquid filled single phase transformers
2-30 kV: ETTF = 340 years; range = < 200, 700 >
31-72 kV: ETTF = 300 years; range = <150,500>
-4-
power transformers
over 550 kV: ETTF = 120 years; range = <60,285>
liquid filled, three phase transformers
146-242 kV: ETTF = 70 years; range = <45, 165>
over 550 kV: ETTF = 80 years; range = <50,250>
2.1.3. Kloeppel et al. [1990) present recommended values of component data for
reliability studies. They were based on data from several industries in Eastern Germany
as well as form the public supply. For transformers the following data are recommended:
MV/LV
30kV/MV
ETTF = 100 years; range < 65, 170 >
repair time = 18 hours; range = < 6, 30 >
ETTF = 50 year; range < 30, 80 >
110 kV/MV ETTF = 30 years; range = <20,45>
220 kV /MV ETTF = 20 years; range = < 16, 25 >
repair time = 180 hours; range = < 12, 360>.
2.1.4. The IEEE gold book [lEEE,1990) gives recommended values for the components
of industrial and commercial power systems. For power transformers it recommends:
liquid filled transformers: ETTF = 161 year
rated power between 300 and 10,000 kVA: ETTF = 169 year
rated powers above 10,000 kVA: ETTF = 65 year.
These values are based on the results of the 1979 IEEE survey [Aquilino, 1980). This
survey is further discussed in Section 2.2.5.
-5-
power transformers
2.2. Data from surveys
2.2.1. Dickinson (1962) reports on an AlEE survey undertaken in 1959. Responses were
recieved from 33 compagnies, covering 85 plants. For power transformers the results
are:
above 15 kV
below 15 kV
above 500 kVA
below 500 kVA
: ETTF = 50 years; c.i. = <25,300>
: ETTF = 130 years; c.i. = <65,260>
: ETTF = 110 years; c.i. = <90,140>
The repair time (average down time) reported was:
above 15 kV
below 15 kV
above 500 kVA
below 500 kVA
: 2000 hours;
: 250 - 300 hours
: 8 - 15 hours
2.2.2. Todd [1964) gives outage rates for transmission and distribution facilities derived
from 5 - 11 years experience at Indianapolis Power & Light Company.
138/34.5 kV transformers:
ETTF = 70 years;
outage duration exceeding 4 hours in all cases;
time of occurrence between 7 AM and 6 PM in all cases.
34.5/4.1 kV transformers:
ETTF = 180 years;
outage duration less than 2 hours
2 - 4 hours
exceeding 4 hours
time of occurence 7 AM to 6 PM
6 PM to 10 PM
:20%
:30%
: 50%
:80%
: 20%
2.2.3. Connor and Parkins [1966) report about a 14-year survey of faults in networks
with nominal voltages between 2 and 33 kV. The results for transformers (excluding
lightning) are shown in the table below. Table 1 gives failure rates per 100 transformers
per year.
system vol tag. 2 5 kV 6.6 kV 11 kV 11 kV 22 kV 33 kV
method of earthing sol id/res sol id/res arc sup cot l sol id/res sol id/res sol id/res
bushing fai lure 0.03 0.04 0.04 0.04 0.06 0.06
wincHna fai lure 0.06 0.17 0.22 0.23 0.22 0.25
oil QUal;tv
. .
0.01
. . .
overload 0.07 0.03 0.08 0.07 0.09
.
taD-chsnqe mechanism 0.01 0.01 0.01 0.02 0.04 0.44
miscellaneous 0.05 0.06 0.04 0.06 0.07 0.65
total 0.25 0.27 0.38 0.39 0.49 1.41
Table 1: failure rate per 100 transformers per year according to a 14-year survey
reported by Connor and Parkins {1966]
-6-
power transformers
The number of transformers connected to the system under study was 14,561 in 1952
and 37,464 in 1965. The following ETTF-values can be derived from the table above:
2-5 kV : ETTF = 400 years;
6.6 kV : ETTF = 370 years;
11 kV arc sup coil : ETTF = 260 years;
11 kV solid/res : ETTF = 260 years;
22 kV : ETTF = 200 years;
all voltages : ETTF = 260 years; c.i. = < 240,270 >.
2.2.4. An IEEE sponsored survey of electrical equipment reliability in industrial plants
was completed during 1972 IIEEE,19741. This survey included a total of 1982
equipment failures that were reported by 30 compagnies covering 68 plants in nine
industries in the United States and Canada. The results for power transformers are given
in Table 2.
average actual estimated restore
down time (hours) time (hours)
ETTF (years) c. i . industry plant repair replace
LfCl\Jid filled
All vol tages 240 [190 325] 529 219 378 73.4
600 15 000 V
All sizes 340 [250 500] 174 49 382 74.3
300 750 kVA 270 [170 750] 61 10.7 49 3.7
751 2 499 kVA 400 [260 870] 217 64 297 39.7
2 500 kVA and above 310 [200 700] 216 60 618 150
Above 15 000 V 77 [55 130] 1076 1260 367 71.5
Dry type
o 15 000 V 275 [190 500] 153 28 67 39.9
Rectifier
above 600 Vol t 34 [23,60] 380 80 300 20.0
Table 2: results of IEEE survey completed during 1972.
The last four columns give: the industry average anf the plant average of the actual
downtime per failure and the average estimated time to fix the failure during 24 hours
work day in case of repair of the failed component and in case of replace with a spare.
2.2.5. The IEEE performed another survey after failures of power transformers in
industrial plants and commercial buildings in 1979 [Aquilino, 1980, 19831. Table 3 gives
the results for liquid filled power transformers and rectifier transformers.
DOwer transfonmers rectifier transformers
ETTF c. i. restore time (hours) ETTF c. i. restore time (hours)
(years) repair replace (year-s) repair replace
All sizes 160 [130 215] 365.1 85.1 50 [30 140] 2316 41.4
300 10 000 kVA 170 [130 230] 297.4 79.3 65 [30 300] 1664 38.7
Above 10 000 kVA 65 [35,400] 1178.5 192 30 [15,111> 2707.2 60
Table 3: Results of IEEE survey of 1979.
-7-
power transformers
The table gives ETTF and confidence interval in years, as well as average estimated time
to fix the failure during 25 hour work day in case of repair of the failed component and
in case of replace with a spare.
For power transformers the number of failures was high enough (about 100) to justify
a further subdivision of the failures.
A subdivision to failure initiating cause yields:
transient overvoltage disturbance :16.4%
overheating : 2.7%
insulation breakdown : 48.1 %
mechanical failure : 12.7%
shorting by external object : 3.6%
malfunction of protection : 4.5%
improper operating procedure : 3.6%
loose connection or termination : 7.3%
others : 0.9%
A subdivision to failure contributing cause yields:
abnormal temperature
exposure of aggressive chemicals or moisture
normal ageing
severe weather conditions
lack or malfunction of protective device
loss or degradation of cooling medium
improper operating procedure or testing error
inadequate maintenance
others
: 5.5%
: 14.4%
: 13.3%
: 4.4%
: 10.0%
: 10.0%
: 3.3%
: 7.8%
: 31.1%
A subdivision to suspected failure responsibility yields:
manufacturer defective component or improper assembly
improper application
inadequate installation and testing prior to start-up
inadequate maintenance
inadequate operation
outside agency
others
A subdivision to type of failure yields:
winding : 53%
others : 47%
A subdivision to failure characteristic yields:
automatic removal by protective device : 75 %
partial failure reducing capacity : 5%
manual removal : 20%
-8-
: 33.3%
: 3.1%
: 6.3%
: 26.0%
: 4.2%
: 9.4%
: 17.7%
power transformers
2.2.6. Cigr6 held an international survey on failures in large power transformers in
service in 1978 [CIGRE,1983]. Data was received from 13 countries (Australia, Austria,
Belgium, Canada, Czechoslovakia, France, Finland, Italy, japan, Switzerland, United
Kingdom, USA and USSR). The analysis took in more than 1000 failures that occurred
between 1968 and 1978, relating to a total population of more than 47,000 unit-years.
The results are summarized in table 4.
ETTF (veersl c.f.(veersl
Power station transformers
60 - 100 kV 65 [40 190]
100 - 300 kV 45 [35 60]
300 - 700 kV 32 [22 54]
without on-load tab-changers 44 [326n
with on-load tab-changers 42 [33 sn
Substation transformers
60 - 100 kV 48 [42 56]
100 - 300 kV 45 [40 50]
300 - 700 kV 46 [30 110]
Table 4: Results of 1978 Cigre survey.
2.2.7. One ofthe Dutch regional electricity authorities, N. V. PNEM, performed a survey
after the reliability of power system components in their 10 kV network, covering the
period 1980 - 1986 [van Amelsfoort et.al, 1986].
For 10 kV/380 V transformers the survey resulted in:
ETTF = 650 yr; c.i. = [500,900].
2.2.8. Waumans [1986] presents the result of failure registration in a number of medium
and low-voltage networks in the Netherlands from 1979 to 1984. For transformers
(apparently MV/LVI he gives:
ETTF = 500 years
2.2.9. A survey performed by the Dutch utilities [VDEN,1987] results in the following
values for 10kV/400 Volt transformers:
ETTF = 950 yr; c.i. = [700,1400]
2.2.10. Pijls [1988] analysed failures in the electricity system supplying a number of
chemical plants, during 8 years. The rated voltage of the system under study was 10
kV. For transformers 10 kV /6 kV and 10 kV /2 kV he found:
ETTF = 280 years; c.i. = < 90 years, 00 >
2.2.11. Franke [1990] observed during 12 years, failures in a chemical plant in Eastern
Germany. Early failures in the wear-in phase (1.5 ... 2 years) have been removed from the
data. For MV/LV transformers up to 1.6 MVA he reports:
ETTF = 85 years; c.i. = < 55,200>
-9-
power transformers
Included in this value are:
intervention by the protection due to internal fault;
alarm signal by Buchholz relay;
over temperature.
2.2.12. Bovy et al. [1991] analysed the outage occurances in some 4000 km of
underground 10 kV network during 6 years. For transformer stations they found:
ETTF: 625 years; repair time: 12 hours;
ETTF for phase-to-ground-faults: 7000 years;
probability of multiple fault due to phase-to-earth fault: 19%
2.2.13. Bruggeman et al. [1991] present some statistics about various types of
transformer faults from British utilities. They divide the faults by the place of occurence.
Their results for ground-mounted, cable-connected distribution transformers are given
Table 5.
cOO1)Onent where the faut t occured failure rate per 1000 unit.vr ETTF (vr)
cable tenminations 0.69 1 450
other external connections 0.13 7700
windings and comections 0.52 1 900
tanks radiators etc. 0.15 6,700
(mostly corrOSion)
taa-changers: mechanical 0.06 17 000
tap-changers: eLectrical 0.008 12 500
other sites mostly accessories 0.08 12 500
'Ilknown 0.14 7 000
all c ts 1.778 550
Table 5: Results of survey by Bruggeman et al.[1991J.
2.2.14. Verplanke [1991] studied failures that occured in the power system of a
chemical industry in The Netherlands between 1970 and 1991. For 50 kV/11 kV
transformers he reports
ETTF = 20 years; sample too small for c.i.
-10-
power transformers
2.3. Data used in reliability studies
2.3.1. Dickinson [1957] gives for transformers in an industrial power system a typical
value of
ETTF = 300 years.
He states:" whether available, actual service experience should preferably be used.
Actual failure rates wil vary widely depending on quality of equipment, installation,
maintenance and on service conditions."
The typical repair time (electrical downtime) after failure of a power transformer is
repair time = 360 - 720 hours.
2.3.2. Dickinson [1960] uses the following values for power distribution systems for
refinery process units. These are based on several published papers and on service
reports obtained from refineries. For transformers he uses:
ETTF = 350 years; repair time = 70 - 500 hours.
2.3.3. Gangel and Ringlee [1968] calculate service interruption rates and restoration
times for a distribution circuit. They use the following data for distribution transformers:
ETTF = 200 - 1000 years; repair time = 1 - 5 hours.
2.3,4. Guertin and Lamarre [1975] performed a reilability analysis of a high-voltage
substation. For HV/MV transformers they use:
ETTF = 10 years; repair time = 1000 hours;
maintenance is performed every year during 6 hours .

2.3.5. Bocker and Kaufman [1977] use the following values for a public distribution grid:
110/10 kV transformer: ETTF = 50 years; repair time = 90 days.
10 kV/400 V transformer: ETTF = 500 years; repair time = 44 hours.
2.3.6. Lonsdale and Hitchen [1977) use the following value for transformers 33/11 kV
and 33/6.6 kV in a public distribution network in the Northwest of England. Data were
based on an examination of system performance over recent years.
ETTF = 170 years.
2.3.7. McNab [1977] uses the following values for 33/11 kV transformers in a public
distribution network in the South of Scotland:
ETTF = 40 years; repair time = 14 hours.
2.3.8. Snaith [1977] uses the following values for 3.3 kV/415 V transformers in the
electricity supply for a nuclear power station:
-11-
power transformers
maintenance interval = 1 year;
2.3.9. Allan, Dialynas and Homer (1979) use the following values for 33/11 kV
transformers in a reliability study of a distribution system:
time to maintenance = 4 years; maintenance time = 8 hours.
2.3.10. Allan et al. [1980] use, in a reliability study of electrical auxiliary systems of a
power station, the transformer reliability data shown in Table 6.
ETTF (ye.rs) repair ti. ti. to dur.tfon of
active passive maintenance 118 f ntenance
400/23.5 kY 17 vr 18 vr 144 hr. 1 vr 168 hr.
132/11 kY 37 vr 42 vr 196 hr. 1 vr 168 hro
23.5/11 kY 70 vr 170 vr 196 hra 1 vr 168 hrs
11/3.3 kY 70 vr 170 vr 120 hrs 1 vr T2 hrs
3.3 kY/415 Y 70 yr 170 yr 120 hrs 1 yr T2 hrs
Table 6: values used by Allan et al. [1980J.
2.3.11. Nelson and Johnson [1982) compare three distribution voltages for the power
supply to a gas centrifuge uranium enrichment plant. They use the following reliability
data (based on 1974 IEEE survey and other available data) for power transformers:
345/13.8 kV : ETTF = 280 years;
345/24.9 kV : ETTF = 250 years;
345/34.5 kV : ETTF = 230 years.
2.3.12. Koval [1983) uses for a 50 MVA transformer in an industrial power system:
ETTF = 85 years;
replacement time = 168 hours;
restoration time = 12.0 hours.
2.3.13. Williams and Mudge [1983) use values drawn from distribution system
statistics. For failures of a 33/11 kV transformer they use:
2.3.14. Sillinton and Goel [1986) use the following data for 25 kVA, 14.4 kV/240 V
pad-mounted transformers in the reliability analysis of an existing 14.4 kV distribution
network in Saskatchewan, Canada:
ETTF = 200 years; replacement time = 48 hours.
2.3.15. Dialynas and Allan [1986] use for 33/11 kV transformers, when including local
generating facilities in the reliability evaluation of power distribution systems.
time to maintenance = 4 years; duration of maintenance = 8 hours.
-12- power transformers
2.3.16. Anderson et al. [1987) use the following values for HV/MV transformers in their
study after the reliability of HV substations:
mechanical failures
ETTF = 130 years; repair time = 12 hours
electrical isolation failures
maintenance is performed every 330 days during 12 hours
2.3.17. Billinton (1987) uses for 14.4 kV/LV transformers in the Anaheim distribution
system of the Saskatchewan Power Corporation the following data:
ETTF = 200 years; replacement time = 48 hours;
2.3.18. Dialynas and Allan [1987] describe a reliability model for a power distribution
network with local generation.
For transformers 33 kV/11 kV they use:
2.3.19. Goldberg et al. (1987) use the following values for distribution transformers:
ETTF = 500 years; repair time = 5.5 hours.
2.3.20. Kevers (1987) uses for 150 kV/ 10 kV transformers values of
The values are based on failure statistics in high-voltage networks.
2.3.21. Dialynas (1988) presents a reliability analyses of a Greek 150/20 kV
transmission and distribution system. He uses the following values for failures in power
transformers:
permanent failures
temporary failures
: ETTF = 7.6 years.
: ETIF = 21 years.
The transformers present in the system under study are 150/20 kV transformers.
2.3.22. Dialynas and Papadopoulos (1989) have determined the reliability of a. "typical"
20 kV rural distribution system in Greece. For a 150kv/20kV transformer they use:
ETTF = 25 years.
2.3.23. Fransen [1989] uses, in a reliability study of an industrial supply, for short-
circuits in transformers a value of
ETTF = 88 years; repair time = 11.5 hours
-13-
power transformers
This value is based on failure statistics during 7.5 years in the power system of a large
chemical plant in The Netherlands.
The transformers under study were mainly 30 kV/10 kVand 10 kV/ 2 kV.
2.3.24. Roos [1989] compares the supply reliability for different configurations of
medium-voltage power systems. For independent failure of HV /L V transformers he uses
the following values:
2.3.25. Whiting [1989] determines the reliability of power-supplies to broadcast
transmitting stations. Component data are taken from various sources. For 11 kV/415
V transformers he uses:
ETTF = 50 years; expected repair time = 120 hours.
2.3.26. Horton et al. [1989a] determine the reliability of service of an underground 21
kV feeder of Pacific Gas and Electric. They have used the following data for distribution
transformers:
2.3.27. Horton et al. [1989b] use their Distribution Reliability Assessment Model
[DREAM] to compute reliability indices for distribution feeders. They use the following
data for power transformers:
ETTF = 500 years; repair time = 330 minutes
2.3.28. Duke et al. [1989] assess the reliability of an industrial distribution system. For
66/11 kV transformers they use:
2.3.29. Allan and Inga-Rojas [1990] describe a method for distribution system reliability.
In an example they use for transformers:
ETTF = 9 years
2.3.30. Sallam et al. [1990] calculate, as an example, reliability indices for the MV-
network of Port-Fouad, Egypt. They have used the following data for power
transformers:
HV/MV transformers: ETTF = 330 years; repair time = 130 hours
MV/LV transformers: ETTF = 250 years; repair time = 1.85 hours.
2.3.31. Mohan Rao and Sekhar [1990] compare the reliability of a number of distribution
systems. They use the following values for 13.8 kV/480 Volt transformers:
-14-
power transformers
2.3.32. Allan at al. [1991] present a reliability test system for distribution networks. The
ETTF values used for power transformers are summarized in Table 7.
voltaaes cermanent fai lures active failures temoorarv failures
138/33 kV 100 vr 100 vr 20 vr
33/11 kV 70 vr 70 vr 20 vr
11 kV/415 V 70 vr 70 vr
Table 7: ETTF values used by Allan et al.[1991J
2.3.33. Dialynas and Koskolos [1991] determine the reliability performance of industrial
power systems. They study an existing industrial system in Greece. For transformer
failures they use the following values:
permanent failures
temporary failures
: ETTF = 7.6 years ; repair time = 89.0 hours.
: ETTF = 21.2 years ; repair time = 1.5 hours.
Maintenance is performed once every 2 years during 19 hours.
The transformers present in their study are 20 kVI 6.6 kV 10 MVA, 20 kV/ 380 V 1.6
MVA, and 6.6 kV/380 V 1.25 MVA.
2.3.34. Prescott et al. [1991] use the following values for the failure of a distribution
transformer:
2.3.35. Volkmann et al. [1991] use a value of ETTF = 500 years for distribution
transformers. The value has been determined from failure reports by Pacific Gas and
Electric.
2.3.36. Kj"lIe and Sand [1991] use the following values for transformers in distribution
systems in Norway:
22 kV rural network: ETTF = 100 years; repair time = 3.0 hours;
11 kV urban network: ETTF = 600 years; repair time = 7.8 hours.
2.3.37. Warren [1992] examines different UPS system configurations to formulate the
most reliable system. The failure rates used are based on field statistics and MIL-HDBK-
217-E. For MV-LV transformers he uses:
ETTF = 300 years.
-15- power transformers
2,4. Ageing data
2.4.1. Parascos and Arceri (1976) present bathtub curves for underground network
transformers. Only "internal winding failures" have been taken into account. The resultys
are shown in Figure 1. The numbers with the curves show the production years. Newer
types show less wear-out but
(relatively) more wear-in.
Figure 2 shows the results for "heavy
tank corrosion requiring removal". As
expected. only a wear-out phase is
visible.
w
...
"
'" w
'"
~
~
.,
~
~
"
z
z
0.3
0.2
"
ae 0.1
\
\
,
\
\
\
\
\
,
I
/
1939-1952
19531962
2
,
6 B 10 12 14 16
AGE OF EQUIPMENT (YEARS)
Figure 1: Failure rate of underground network
transformers. internal winding failure.
According to Parascos and Arceri [1976J.
w
...
0.40
~ 0.30
w
'"
~
~
;;
~ 0.20
'"
~
z
z
" ..
0.10
5 10 1S 20 25 30 35 40
AGE OF EQUIPMENT (YEARS)
Figure 2: Failure rate for underground network
transformers; heavy corrosion of tank material.
According to Parascos and Arceri {1976J.
-16-
power transformers
2.4.2. The IEEE performed a survey after failures of power transformers in industrial
plants and commercial buildings in 1979 [Aquilino, 1980, 1983). Table 8 gives some
results on ageing of liquid filled power transformers with rated power between 300 and
10,000 kVA. Age is defined in this survey as the age at the end of the reporting period.
Age ETTF (years) conf. interval
1 ~ 10 years 138 [90 3251
11 - 25 years 188 [135 3001
above 25 years 165 [140,2801
Table 8: Influence of age.
2.4.3. Cigre held an international survey on failures in large power transformers in
service in 1978 [CIGRE, 1983). Data was received from 13 countries (Australia, Austria,
Belgium, Canada, Czechoslovakia, France, Finland, Italy, japan, Switzerland, United
Kingdom, USA and USSR). The analysis took in more than 1000 failures that occurred
between 1968 and 1978, relating to a total population of more than 47,000 unit-years.
Some results for ageing are summarized in Table 9.
60 - 100 kV 100 - 700 kV
age (yrl ETTF (yrl c. i. ETTF (yrl c. i .
o - 5 36 [28 481 57 [48 701
5 - 10 49 [38 681 49 [41 601
10 - 20 58 [48 741 43 [37511
Table 9: influence of age.
2.4.4. Bar et al. [1990) use a Delphi-method to determine the position of the knee in the
bath-tube curve, i.e. the place where ageing starts to become important.
For "oil-filled" MVILV transformers they find in case of
good circumstances : 42 years;
average circumstances : 35 years;
bad circumstances : 25 years.
For cast-resin isolated transformers they find in case of
2.4.5. Ducket and McDonough [1990) have performed a study after the ageing of
transformers. During 5 years they have observed the failure behaviour of a population
with a known age structure. The data has been provided by California, Power and Light.
Over 8600 events of failure, which resulted in pole type transformers being removed
from the system and scrapped, occured during the years 1984 through 1988. These
units were purchased from a single supplier between the years 1947 and 1988.
The information available is: the number of transformers purchased for each year
between 1947 and 1988; the number of failures for each year between 1984 and 1988.
From this information an estimation is made of the failure rate as a function of time. The
result is shown in Figure 3.
-17-
power transformers
" 1000r---------------------------------,
100
10
0 . 1 L - - - ~ - - - - - L - - - - ~ - - ~ ~ - - ~ - - - - ~
o 5 10 15 20 25 30
Yee,.
Figure 3: Failure rate of transformers as a function of their age, according to Bucket
and McDonough [1990}.
For the first 5 years the failure rates are fairly accurate and can be assumed constant.
This leads to a value of:
ETTF = 93 years; c.L = <85,103>
I slightly changed the data of Ducket and McDonough [1990]. The number of failures
for the first year of operation has been doubled. For the first year they give a failure rate
of 0.423 % whereas the failure rate for the years 2 through 6 is around 1 %. As
transformers are put into operation the whole year through, but the counting of failures
stops at the end of the year, the first year of operation is on the average just half a year.
2.4.6. Horton et al. [1990] provide estimates of failure rates of underground distribution
system components. The estimates are based on information collected from a number
of utility sources throughout the United States of America. For each component and
each year they used the number of units installed or removed and the number of failures
which occured. This data has been fitted to a Weibull distribution.
From over 88,000 single-phase pad-mounted transformers installed by the NELPA
(Northwest Electric Light & Power Association) utilities between 1968 and 1988 they
find a constant failure rate:
ETTF = 350 years; c.L = [340,370].
-18-
power transformers
2.5. Conclusions
Table 10 and Table 11 summarize the data on transformer life time given in this chapter.
Table 10 gives recommended data from surveys, where Table 11 gives data used in
reliability studies. A distinction has been made between MV/LV transformers (e.g.
10kV/400V). MV/MV transformers (e.g. 33/11 kV) and HV/MV transformers (e.g.
150/10kV).
Section Reference MV/LV MY/MY HV/MV unknown
2.1.1. Green and Bourne 1972 190 2957 16
2.1.2. IEEE 1983 340-875
2.1.3. Kloeppel et al. 1990 100 50 20-30
2.2.1- Dickinson 1962 110-130 50
2.2.2. Todd 1964 180 70
2.2.3. Connor and Parkins 1966 200-400 70
2.2.4. IEEE 1974 270-400 77
2.2.5. AQUilino 1980. 1983 65-170
2.2.6. CIGRE 1983 32-65
2.2_7. van Amelsfoort et at. 1986 650
2.2.8_ Walln8ns 1986 500
2_2.9. VOEN 1987 950
2.2_10_ Pi j Is 1988 280
2.2.11. Franke 1990 85
2_2.12. BoVY et al. 1991 625
2.2.13. Bruggeman et at. 1991 550
2.2.14_ Verplanke 1991 20
Tab/e 10: Summary of transformer life times: recommended values and data from
surveys_
Most surveys result in an ETTF value in the order of 500 years for MV/LV transformers.
Deviations from this value are given by: Green and 80urbe [1972] (data from the 60's
7); Kloeppel et al. [1990] (data from Eastern Germany 7); Dickinson [1960] (data from
the 50's 7); and by Franke [1990] (data from Eastern Germany 7). Values around 500
years are also used in most of the reliability studies. Values below 100 years can clearly
be described as non-realistic. From the surveys a clear increase in reliability is visible in
the course of time: from about 100 years in the 50's to over 500 years for the latest
surveys. Improved manufacturing and maintenance clearly are fruitfull.
MV/MV transformers are less reliable than MV/LV transformers. Lifetimes of 50 to 75
years seem to follow from the surveys. Higher values are given by Todd [1964] and Pijls
[1988]. A lower value is given by Verplanke [1991]. The latter two (both data from
chemical plants in The Netherlands) use a small population, which could explain their
deviating values. It might also be due to the specific operating condition. Some reliability
studies use a value of 500 years. Apparantly the value for MV/LV transformers has been
used. Most other studies use values consistent with the results from surveys.
HV/MV transformers are even less reliable than MV/MV transformers, with lifetimes
between 25 and 30 years. Some reliability studies however use lifetimes up to 330
years. These values are certainly not realistic.
MV/MV as well as HV/MV transformers become somewhat better during the years, due
-19-
power transformers
to improved manufacturing and maintenance. This improvement in time can also be
concluded from Figure 1.
From the data presented in this chapter I would suggest the values given in Table 12.
Section Reference MY/LV MY/MY HV/MY u:>known
2.3.1. Dickinson 1957 300
2.3.2. Dickinson 1960 350
2.3.3. Gangel and Rinal .. 1968 2001000
2.3.4. Guertin and Lamarre 1975 10
2.3.5. Bocker and Kaufman 19n 500 50
2.3.6. Lonsdale and Hitchen 19n 170
2.3.7. McNab 19n 40
2.3.8. Snaith 19n 170
2.3.9. Allan Dialvnas Homer 1979 500
2.3.1D. Allan et at. 1980 70 70 1737
2.3.11. Nelson and Johnson 1982 230280
2.3.12. Koval 1983 85
2.3.13. WiLLians and Mudge 1983 80
2.3.14. Billinton and Goel 1986 200
2.3. IS. Oistynes and Allan 1986 500
2.3.16. Anderson et at. 1987 lID
2.3.17. Billinton 1987 2DD
2.3.18. Dlstynes and Allan 1987 500
2.3.19. Goldbero et al. 1987 500
2.3.20. Kevers 1987 IS
2.3.21. Dialynas 1988 7.6
2.3.22. Dialynas and Papadopoulos 1989 25
2.3.23. Fransen 1989 88
2.3.24. RODS 1989 650
2.3.25. \/hitin . 1989 50
2.3.26. Horton et at .. 1989a 500
2.3.27. Horton et at. 1989b 500
2.3.28. Duke et at .. 1989 40
2.3.29. Allan and Inga-Ro"ss 1990 9
2.3.30. Sallam et at .. 199D 250 33D
2.3.31. Mohan Roa and Sekher 199D 300
2.3.32. Allan et at. 1991 70 70 100
2.3.33. Dialynas and !Coskalos 1991 7.6 7.6
2.3.34. Prescott et at. 1991 42
2.3.35. volkmam et at. 1991 500
2.3.36. Kj.lle and Sand 1991 100600
2.3.37. Warren 1992 300
Table 11: Summary of transformer life times: data used in reliability studies.
existing transformers new transformers
MV/LV transformers 400 . 60D years 500 .. 1000 vears
MY/MY transformers 50 .. 75 years 75 . 100 years
HY/MV transformers 30 . 50 years 4D 70 vears
Table 12: Suggested values for power transformers.
20-
power transformers
3. CIRCUIT BREAKERS AND SWITCHES
3.1.1. Green and Bourne [1972] give average component failure rates for electrical
components. For circuit-breakers they give:
general (less than 33 kV):
415 V to 11 kV:
33 kV:
132 kV:
275 kV:
400 kV:
ETTF = 57 years
ETTF = 76 years
ETTF = 38 years
ETTF = 29 years
ETTF = 16 years
ETTF = 11 years
3.1.2. The IEEE gold book [IEEE. 1990] recommends the ETTF-values given in Table 13.
all values are in years:
o - 600 V
All
All 'vee.
0-600A 600 A +
600 V +
Fixed circuit breaker
192 240 285 105 57
Metal-clad circuit breaker 330 370 430 330 280
motor starter 72 65
enclosed disconnect switch 165
Table 13: recommended values (ETTF in years) according to the IEEE gold book.
3.1.3. Kloeppel et al. [1990] present recommended values of component data for
reliability studies. They were based on data from several industries in Eastern Germany
as well as from the public supply. For circuit-breakers the following values are
recommended:
below 1 kV: ETTF = 650 years; range
=
<200. 1000>
repair time = 4 hours; range
=
<3.8>
6. 10 kV: ETTF = 135 years; range
=
<85.150>
20 kV: ETTF = 75 years; range
=
< ... 115>
=
<12.27>
30 kV: ETTF = 26 years; range
=
<12.27>
110 kV open: ETTF = 48 years; range
=
<37.63>
=
< 12.24>
110kV closed: ETTF = 1000 years
=
<36. 72>
For disconnect switches the following values are recommended:
below 1 kV:
10-30 kV:
100 kV:
ETTF = 300 years;
repair time = 3 hours.
ETTF = 400 years; range = <200.600>
ETTF = 110 years; range = <85. 140>
-21- circuit breakers and switches
3.2. pata from surveys
3.2.1. Dickinson (1962) reports on an AlEE Survey undertaken in 1959. Responses were
received from 33 compagnies, covering 85 plants. For circuit breakers the results are:
Fixed Circuit breakers
above 15 kV:
2.4 - 15 kV:
ETTF = 70 years, sample too small
ETTF = 85 years, c.L < 60, 150 yr>
Draw-out metal-clad circuit breakers
above 600 V: ETTF = 96 years, c.L = <70, 140>
600 V and below: ETTF = 61 years, c.L <53,73>
Circuit breakers used as motor starters.
above 600 V: ETTF = 23 years; c.L = < 19, 30>
Contactor-type motor starters
above 600 V: ETTF = 50 years; c.i. = <41,66>
600 V and below: ETTF = 57 years; c.L = <54,60>
The reported repair time is:
Circuit breakers used as motor starters:
above 600 V: average 14-27 h: min. 2 h; max: 240 h
Contactor-type motor starters
above 600 V: average 15-17 h; min. 0 h; max: 32 h
600 V and below: average 5-8 h; min. 1 h; max: 125 h
For disconnect switches the results are:
Open switches:
Enclosed switches:
ETTF = 102 yr, c.i. = <80,140 yr>
ETTF = 230 yr, c.L = < 170, 365 yr>
3.2.2. Connor and Parkins (1966) report about a 14-year survey of failures in networks
with nominal voltages between 2 and 33 kV. The results for switchgear are given below.
The number of switchgear units in service has increased from 12,000 in 1951 to 23,000
in 1965. The authors note that the fault incidence of switchgear of all voltage has
remained practically constant over the period under consideration. The manufacture of
smaller units, to reduce costs, appears to be cancelled out by improving materials and
manufacturing techniques.
Circuit breaker failures ETTF = 6500 years;
Tripping or closing mechanism ETTF = 4000 years;
A.R. tripping or closing mechanism ETTF = 12,000 years;
Current transformers ETTF = 2700 years;
Voltage transformers ETTF = 17,000 years;
Other failures of outdoor switchgear ETTF = 3300 years;
Other failures of indoor switchgear ETTF = 3400 years;
Small wiring and auxiliary switches ETTF = 12,500 years;
Failure of metalclad fuse switch
Failure of metalclad
oil-immersed isolator or switch
Failure of air-break isolator
Miscellaneous
Total
ETTF = 1400 years;
ETTF = 7100 years;
ETTF = 10,000 years;
ETTF = 1100 years;
ETTF = 290 years.
3.2.3. Selseth [19721 presents results from a survey of outage records at the Norwegian
State Power System, over the period 1967-1970. The survey covered a population of
2,479 circuit-breaker years. The following results are derived for circuit breakers:
ETTF = 25 years; c.i. = <20,30>
A subdivision of failures yields:
mechanical parts : 30%;
electrical parts : 24%;
unknown parts : 12%;
others : 34%.
3.2.4. An IEEE survey of electrical equipment reliability in industrial plants was
completed in 1972 IIEEE,19741. This survey included a total of 1982 equipment failures
that were reported by 30 compagnies covering 68 plants in nine industries in the United
States and Canada. The results for circuit breakers are given in Table 14.
average actual estimated restore
down time (hours) time (hours)
ETTF (vears) c.L industry Dlant reoair restore
Fixed tYDe
All voltages 194 [140 310] 5.8 4.0 31.7 4.5
Below 600 Volt 225 [160 375] 4.7 4.0 6.0 2.0
Above 600 Volt 57 [30 500] 10.6 3.8 44.5 12.0
Metalclad drawout
All voltages 330 [260 425] 129 7.6 54.2 3.9
Below 600 Volts 370 [280 560] 147 4.0 47.2 2.9
Above 600 Vo l ts 280 [210 430] 109 168 62.4 5.2
Disconnect switches
Enclosed 160 [130 220] 3.6 2.8 50.1 13.7
Open 350 [170 ...
Table 14: Results from 1972 IEEE survey {IEEE, 19741.
The last four columns give the industry average and the plant average of the actual
down time per failure and the average estimated time to fix the failure during 24 hour
work day in case of repair of the failed component and in case of replace with a spare.
A subdivision to failure characteristic yields for circuit breakers:
fail-to-close : 5%
failed while opening : 9%
mal-trip : 44%
damaged while successfully opening: 7%
damaged while closing : 2%
failed while operating : 32%
A subdivision to damaged part yields:
insulation : 21 %
mechanical : 18%
auxiliary or protective device : 34%
other : 27%
3.2.5. Lonsdale and Hitchen [1977] give the following data based on an examination of
system performence over recent years:
Switchgear 33kV, 11 kV, 6.6 kV: ETIF = 500 years.
3.2.6. Mazza and Michacca [1981] as well as Heising [1986] presents the results of a
worldwide survey after failure of circuit breakers in the voltage range above 63 kV. The
survey was performed by working group 13.06 of CIGRE between 1974 and 1977. Their
results are summarized in Table 15 and Table 16. A distinction is made between major
failures and minor failures. A major failure occurs when the breaker can no longer
perform all of its fundamental functions, or when intervention within 30 minutes is
necessary. All other failures are referred to as minor failures. The results for the major
failures are shown in Table 15; those for the minor failures in Table 16.
reoair time
vol tOQe ETTF c.i. average modi ...
63100kY 250 yr <200 300. 29 hrs 5 hr.
100200kY 60 yr <55 70> 95 hrs 12 hrs
200-300kY 40 yr <35 45. 60 hr. 11 hr.
300-500kY 22 yr <19 25,. B4 hr. l' hra
>500kY 10 yr
<8 12' 142 hr. 27 hr.
Table 15: Results for major failures, according to Cigre survey [Heising, 1986J_
The data below holds for the voltage range from 63 to 100 kV. The data for other
voltage classes is too far away from our area of interest.
Does not open on command
ENTF = 60,000; c.L <35,000, 130,000>
Does not break the current
ENTF = 400,000
Does not close on command
ENTF = 18,000; c.L < 13,000, 26,000 >
Failures without a command to open or close
ETIF = 450 yr; c.i. = <350,600 yr>
(ENTF = Expected Number of operations To Failure)
The repair time has been defined as the time required to get to site, analyse the failure,
obtain spare parts, repair and return the circuit-breaker to service, excluding deliberate
delays.

repair time in hours
voltaae ETTF (vears) c.; . average medh.rn
63-100kV 60 [55 6n 16_5 5_0
100-200kV 24 [22 26] 19_9 5_0
200-300kV 16 [14 In 27_7 6_0
300-500kV 6 [5 n 73_1 8_0
>500kV 20 [15 40] 58_2 9_0
Table 16_- Results for minor failures, according to Cigre survey [Heising, 19861_
A subdivision to origin of the failure yields for major failures:
mechanical : 70 %
electrical (main circuit) : 10 %
electrical (auxiliary and control circuit) : 20 %
A subdivision to cause of the failure yields:
design or manufacture : 45 %
incorrect erection : 10 %
incorrect maintenance : 8 %
Condition of the circuit-breaker when failure was discovered:
major failures
in service
during maintenance
minor failures
in service
during maintenance
: 90 %
: 5 %
: 70 %
: 25 %
Characteristics of the major failure:
Does not close on command : 33.7 %
Does not open on command : 14.1 %
Closes without command : 1.7 %
Opens without command : 5.2 %
Does not make the current : 1.6 %
Does not break the current : 1.9 %
Fails to carry current : 2.5 %
Internal breakdown : 5.3 %
External breakdown : 4.0 %
For the major failure "internal or external breakdown" the following values are derived
from the survey results:
63-100 kV : ETTF = 4000 years ; c.i.
=
<2000,11000>
100-200 kV : ETTF = 550 years ; c.i.
=
<400,750>
200-300 kV : ETTF = 600 years : c.i.
=
<400,1100>
300-500 kV : ETTF = 400 years ; c.i.
=
<250,800>
>500 kV : ETTF = 60 years ; c.i.
=
<40,110>
3.2.7. According to Heising (1983) the expected life time of vacuum power circuit
breakers is:
ETTF = 1.400 years.
This value has been based on field experience between November 1971 and January
1982.
3.2.8. Waumans (1986) presents the result of failure registration in a number of medium
and low-voltage networks in the Netherlands, from 1979 to 1984. For MV circuit-
breakers he gives:
ETTF = 500 years
For disconnect switches he finds:
ETTF = 1000 years.
3.2.9. One of the Dutch regional electricity authorities, N.V. PNEM, performed a survey
after the reliability of power system components in their 10 kV network, covering the
period 1980 - 1986 [van Amelsfoort et aI., 1986). For circuit breakers the survey
resulted in:
ETTF = 350 yr c.i. = [280,440)
The survey resulted for disconnect switches in:
ETTF = 6000 yr c.i. = [4000,11000).
3.2.10. Stanek and Venkata (1988) compare failure rates of equipment in four different
coal mines. For "molded-case circuit breakers 400-A frame size" they report:
mine 1: ETTF = 4.3 years;
mine 2: ETTF = 65 days;
repair time = 1.25 h;
mine 4: ETTF = 0.5 years;
repair time = 2.0 h.
The authors conclude that these failure rates are one to three orders of manitude higher
than those of similar devices in other industries.
3.2.11. Norris (1989) reports on an IEEE survey conducted in 1985 on the reliability of
circuit breakers in industrial and commercial installation. Due to the low number of
responses, 13 plant location, detailled analysis was useless. The results are summarized
below:
0-600 Volt air-magnetic:
ETTF = 130 year, c.L = <90,200>
ETTFO = 13 year
ETTNFO = 44 days
repair time (average failure duration) = 2.8 hours
601 - 15,000 Volt air-magnetic
ETIF = 170 year ;sample too small
ETIFO = 740 days
ETINFO = 10 days
repair time = 2.25 hours
34.5 - 138 kV bulk oil
ETIF = 21 year, c.L = < 13,64>
ETIFO = 680 days
ETINFO = 16 days
345 kV air blast and SFs
ETIF = 11 year, c.i. = <8,19>
ETIFO = 215 year
ETINFO = 11 days
ETIFO = Expected Time To Fault Operation
ETINFO = Expected Time To Non-Fault Operation
For increasing rated voltages the time-to-failure of the circuit breakers decreased
considerably where as the repair time increases. The non-availability increases from
0.006 % for low-voltage breakers to 4 % for 345 kV breakers.
For air-magnetic breakers the ETIF is about equal for low-voltages and medium-voltages
despite the far greater amount of operations for medium-voltage breakers.
3.2.12. Radwan and EI-Marsafawy [1990] present failure data for MV circuit-breakers
collected from 1980 to 1984 in the Upper Egypt Zone of the Egyptian Unified Power
System network. The total number of circuit breakers considered was 881 (107 indoor
and 774 outdoor); leading to a population of 3054 unit-years.
66 kV circuit breakers; all types
ETIF = 12 years; c.i. = <9,20>
66 kV oil-filled circuit breakers
ETIF = 35 years; c.L = < 18, ... )
66 kV air-blast circuit breakers
ETIF = 5 years; c.L = <3,8>
ETIF = 23 years; c.L = < 18,35 >
33 kV oil-filled circuit breakers
ETIF = 21 years;c.L = <15,35>
33 kV air-blast circuit breakers
ETIF = 25 years; c.L = < 15, 70>
ETTF = 49 years; c.L = <37.70>
6.6 kV circuit breakers; all types
ETTF = 100 years (only two failures reported)
A subdivision to origin of failure yields:
66 kV circuit breakers
electrical insulation to ground: 63 %
electrical controls: 13 %
seals gaskets: 10 %
seals gas: 50 %
operating mechanism: 35 %
air system: 28 %
operating mechanism: 28 %
isolating contacts: 14 %
A subdivision to cause of failures yields:
pollution: 60 % (repair time = 1 hour)
wear/ageing: 20 % (repair time = 13 hours)
wear/ageing: 60 % (repair time = 2 hours)
maintenance: 5 % (repair time = 14 hours)
others: 35 % (repair time = 20 minutes)
wear/ageing: 50 % (repair time = 50 minutes)
installation: 20 % (repair time = 1 hours)
3.2.13. Franke [1990] observed. during 12 years. failures in a chemical plant in Eastern
Germany. Early failures in the wear-in phase (1.5 ... 2 years) have been removed form
the data.
6 kV circuit-breakers
ETTF = 170 years; c.L = < 100. 250>
Included in this value are
- short-circuit in the breaker
- failure in the control part
- incorrect intervention
3.2.14. Ishibashi et al. [1990] looked for vacuum failures of vacuum interrupters
between 1965 and 1987. All known failures occurred in the first 8 years of production
1965-1973: 23 failures; 1.5 x 10
6
unit-years
ETTF = 50.000 years
1974-1987: no failures; 8 x 10
6
unit-years
ETTF > 1.000.000 years
3.2.15. Filter and Jones [1990] performed laboratory tests on loadbreak elbow
connector opening. Their test program was designed to address the issue of switching
speeds and their influence on switching reliability.
Average speed = 0.38 m/s; stationary piston
sample reliability = 86.7 %; c.i. < 71,97 >
Average speed = 0.38 m/s; floating piston
sample reliability = 100 %; c.i. < 92,1 00 >
Average speed = 0.61 m/s
sample reliability = 100 %; c.i. <98,100>
Average speed = 1.17 m/s;
Average speed = 1.70 m/s;
3.2.16. Working Group 13-06 of CIGRE performed a follow up of their 1977 survey
(discussed in Section 3.2.6) during 1988-1989 [Cigre, 1989]. Some of their results are
given in Table 17.
regair time
vol taae ETTF c. i. 8veraae mediLrn
63100kV 400 vr <600 1500> 45 hr. 24 hr.
100200kV 140 vr <100 150> 49 hr. 8 hrs
200300kV 90 vr <75 125,. 70 hI's 10 hr.
300500kV 60 vr <50 15> 199 hr. 10 hr.
>500kV 55 vr <40 100> 464 hr. 120 hr.
Table 17: Results of 1988-89 Cigre survey {Cigre, 19911.
3.2.17. Bovy et al. [1991] analysed outage occurances in the 10 kV cable networks of
a Dutch utility, during 6 years. For circuit breakers they find:
ETTF = 5,000 years;
ETTF for phase-to-earth faults = 10,000 years;
probability of multiple fault following phase-to-earth fault: 41 %
3.2.18. Volkmann et al. [1991] give the following values for failures in underground
distribution systems for the source of their data they refer to two internal reports by
Pacific gas and Electric.
Non-load break elbow
Molded splice
Switch
: ETTF = 5,200 year
: ETTF = 5,200 year
: ETTF = 250 year
For overhead distributions they give,
Switch (urban feeder) : ETTF = 1,300 year
Switch (rural feeder) : ETTF = 800 year
These two values were computed from historical outage records for a group of 85 rural
and 95 urban feeders. The feeders were selected from areas known to provide accurate
and compehensive reports of distribution circuit failures. The data span a period of
approximately 5 years.
The average repair times. from historical outage data. are:
Molded splice (rural)
(urbani
Elbow (rural and urbani
overhead switch (rural)
(urban)
underground switch (rural)
recloser (rural)
(urban)
(urban)
: 303 minutes
: 205 minutes
: 210 minutes
: 95 minutes
: 116 minutes
: 79 minutes
: 226 minutes
: 200 minutes
: 74 minutes
3.3.1. Dickinson [1957) gives a typical value of
ETTF = 100 years
for circuit breakers. Typical values of the repair time are between 4 and 8 hours.
3.3.2. Dickinson [1960) uses the following values for power distribution systems for
refinery process units. These are based on several published papers and on service
records obtained from refineries.
Circuit breakers, fixed : ETTF = 200 years ; repair time = 40 hours.
Circuit breakers, drawout : ETTF = 250 years ; repair time = 4 hours.
Switches, integral with equipment : ETTF = 1000 years; repair time = 40 hours.
Switches, single throw : ETTF = 500 years ; repair time = 40 hours.
3.3.3. Capra et al. [1969) use the following data for a reliability analysis of 12 kV
underground distribution systems:
Switches, 600A
Switches, 200A
: ETTF = 500 - 2000 years;
: ETTF = 500 - 2000 years;
3-phase interrupters, 600A : ETTF = 100 - 200 years;
3-phase interrupters, 200A : ETTF = 100 - 200 years;
1-phase interrupters, 100A : ETTF = 200 - 500 years.
3.3.4. Grover and Billinton [1974) use the following data for a study after substation
reliability:
High-voltage breakers
short-circuits
incorrect trip
stuck-breaker probability
Low-voltage breaker
short-circuits
incorrect trip
stuck-breaker probability
disconnect switches
short-circuit
mal-trips
: ETTF = 30 year
: ETTF = 5 year
: 0.5 %
: ETTF = 100 year
: ETTF = 100 year
: 6 %
: ETTF = 50 years
: ETTF = 5 years
The average outage duration is 2.09 hours. the switching time 3 hours. Maintenance is
performed every 4 years during 4 hours.
3.3.5. Guertin and Lamarre (1975) performed a reliability analysis of a high-voltage
substation. For high-voltage circuit-breakers they use
ETTF (fault) = 30 years;
repair time = 70 hours;
maintenance every 4 years during 20 hours;
ETTF (mal-trip) = 5 years:
stuck-breaker probability = 1 %
For low-voltage circuit-breakers they use
ETTF (fault) = 100 years;
ETTF (mal-trip) = 30 years
For load-break switches they use:
ETTF (fault) = 50 years
repair time = 10 hours
ETTF (mal-trip) = 50 years;
3.3.6. Bocker and Kaufmann [1977) use the following values for a public distribution
network:
Switchgear bay 10 kV with circuit breaker:
Switchgear bay 10 kV with load-break switch:
3.3.7. Allan et al. [1977) use the following data for 11 kV circuit breakers in the
electrical auxiliary systems of power stations:
Active failures : ETTF = 67 years;
Passive failures : ETTF = 200 years;
repair time
stuck probability
time to maintenance
duration of maintenance
= 20 hours;
= 0.1 %;
= 1 year;
= 50 hours.
3.3.8. McNab [1977) uses the following values for a public distribution network in the
South of Scotland:
33 kV circuit breakers: ETTF = 110 years; repair time = 3.6 hours.
11 kV circuit breakers: ETTF = 72 years; repair time = 2.8 hours.
3.3.9. Snaith [1977) uses the following values for the electricity supply of a nuclear
power station:
3.3 kV or 415 Volt circuit breaker
ETTF = 200 years
time to maintenance = 0.5 years
Contactor starter
ETTF = 500 years
time to maintenance = 0.2 years
3.3.10. Chang [1977) uses the following data for load break connectors to evaluate
distribution system design:
ETTF = 300 years;
replace time = 4 1/2 hours.
3.3.11. Allan et al. [1979) use the following values in a reliability study of a distribution
system:
33 kV circuit breaker
ETTF = 200 years
time to maintenance = 4 years
maintenance time = 8 hours
11 kV circuit breaker
ETTF = 500 years
11 kV and 33 kV isolators
ETTF = 1000 years
3.3.12. Allan et al. (1980) use, in a reliability study of electrical auxiliary systems of a
power station, the circuit breaker reliability data shown in Table 18. They further use:
Stuck probability of normally closed breakers : 0.1 %
Stuck probability of normally open breakers : 1.3 %
~
ETTF repair
active passive
ti.
(years) (years) (hours)
400 kV 6 20 54
132 kV 40 125 45
11 kV 200 140 48
3.3 kV 200 140 36
415 V 200 140 36
time to I duration of
maintenance
(years) (hours)
1 168
1 168
1 24
1 8
1 24
Table 18: Circuit breaker data used by AI/an et al. [1980}.
3.3.13. Adams and Jasmon (1981) use the following data for circuit breakers in a
distribution system:
Active failures : ETTF
Passive failures : ETTF
Repair time
Stuck probability
Time to maintenance
Duration of maintenance
= 10 years;
= 7 years;
= 20 hours;
= 0.5 %;
= 2 Yz years;
= 12 hours.
3.3.14. Nelson and Johnson (1982) compare three distribution voltages for the power
data (based on 1974 IEEE survey and other available data) for circuit breakers:
13.8 kV
24.9 kV
34.5 kV
: ETTF = 278 years;
: ETTF = 265 years;
: ETTF = 253 years.
For switches they use:
13.8 kV
24.9 kV
34.5 kV
: ETTF = 670 years;
: ETTF = 450 years;
: ETTF = 330 years.
3.3.15. Ruoff and van Meteren (1983) use a value of
ETTF = 380 yr
for a circuit breaker being normally closed and
ETTF = 750 yr
for a circuit breaker being normally open. This applies to short circuits in circuit breakers
in distribution systems.
3.3.16. Williams and Mudge (1983) use a value of
ETTF = 500 year
with a repair time of 75 hours for 33 kV circuit breakers and a value of
ETTF = 800 year
with a repair time of 5 hours, for 11 kV circuit breakers. The values have been drawn
from distribution system statistics.
3.3.17. Koval (1983) uses for circuit breakers in an industrial power system:
Opens without command:
ETTF = 36 years;
restoration time = 20 hours.
Short circuit:
ETTF = 140 years;
For switches he uses:
ETTF = 300 years;
3.3.18. Declerq et al. (1985) compare two methods for HV/MV substation reliability.
They use the following data for MV circuit breakers:
3.3.19. Dialynas and Allan (1986) use the same values as Allan et al. (1979) when
including local generation facilities in the reliability evaluation of power distribution
systems.
3.3.20. Anderson et al. [1987) use the following data for circuit-breakers in their study
after the reliability of HV substations.
mechanical faults : ETTF = 250 years;
: repair time = 12 hours;
electrical isolation failures : ETTF = 230 years;
: repair time = 12 hours;
maintenance is performed every 240 days during 8 hours;
They use the following data for disconnectors:
mechanical faults
electrical isolation failure
: do not occur
: ETTF = 500 years
: repair time = 6 hours
maintenance is performed every year during 2 hours.
network with local generation.
For 33 kV circuit-breakers they use
ETTF = 200 years;
For 11 kV circuit-breakers they use
ETTF = 500 years;
For isolators they use:
ETTF = 1000 years
repair time 75 hours
3.3.22. Goldberg et al. [1987] use the following values:
Switches
ETTF: 250 years
Splices
ETTF = 1700 years (old splices)
ETTF = 17000 years (new splices)
Elbow
ETTF = 1700 years (old elbows)
ETTF = 17000 years (new elbows)
3.3.23. Allan (1988) uses in a reliability study a value of ETTF = 100 years for a short
circuit in a circuit breaker.
3.3.24. Dialynas [1988] uses a value of ETTF = 13 years for 150 kV circuit breakers
and a value of ETTF = 9.6 years for 20 kV circuit breakers.
He uses a value of ETTF = 75 years for 150 kV isolators.
3.3.25. Dialynas and Papadopoulos [1989] use the following values:
ETTF = 300 years for a 20 kV circuit breaker.
ETTF = 300 year for a 20 kV recloser.
ETTF = 250 years for a 20 kV sectionaliser.
ETTF = 250 years for a 20 kV isolator.
3.3.26. Fransen [1989] uses a value of
ETTF = 90 years
for short circuits in circuit breakers. The value is based on failure statistics during 7%
years in the power system of a large chemical plant. The repair time used is 8 hours.
The voltage levels under study are mainly 30 kV. 10 kV and 2 kV.
3.3.27. Whiting [1989] determines the reliability of power supplies to broadcast
transmitting stations. Component data are taken from various sources.
11 kV circuit breakers:
ETTF = 85 years
415 V circuit breakers:
ETTF = 200 years
kV feeder of Pacific Gas and Electric. They use the following component data:
Switches
ETTF: 250 years
Splices
ETTF = 1700 years (old splices)
ETTF = 17000 years (new splices)
Elbow
ETTF = 1700 years (old elbows)
ETTF = 17000 years (new elbows)
3.3.29. Horton et al. [1989b] use their Distribution Reliability Assessment model
(DREAM) to compute reliability indices for distribution feeders. They use the following
data:
Splice
ETTF = 5300 years
repair time = 90 minutes
Elbow
ETTF = 5300 years
Switch
ETTF = 250 years
circuit breakers they use the following data:
ETTF = 300 years;
ETTF = 350 years;
In an example they use for circuit breakers:
Open breakers
probability failing to close: 1 %
ETTF = 50 years
Closed breakers
ETTF = 25 years
3.3.32. Sallam et al. [1990] calculated, as an example, reliability indices for the MV-
network of Port-Fouad, Egypt. They use the following data:
13.8 kV circuit breakers:
ETTF = 280 years
Manual switches
ETTF = 10,000 years
Disconnect switches
ETTF = 170 years
systems. They use the following data:
13.8 kV disconnect switch (enclosed)
ETTF = 160 years
13.8 kV metalclad circuit breaker
ETTF = 280 years
480 Volt metalclad circuit breaker
ETTF = 370 years
3.3.34. Allan at al (1991) present a reliability test system for distribution networks. For
this network they use for circuit breakers the ETTF data in Table 19.
permanent active terJ1lOrary
failures failures fai lures
138 kV 175 vrs 285 vrs 20 vrs
33 kV 500 vrs 670 vrs 50 vrs
11 kV 170 vrs 250 vrs 17 vr.
Table 19. Circuit breaker data used by Allan et al. [1991 J.
3.3.35. Dialynas and Koskolos (1991) use a value of
ETTF = 9.61 yr
for permanent failures of circuit breakers. The repair time is 280 hours. Maintenance is
performed every two years during three hours.
the stuck breaker probability used is 0.1 %.
The system under study contains circuit breakers of 20 kV, 6.6 kV and 380 Volt.
For a 150 kV breaker they use ETTF = 12.9 yr, with a repair time of 21.0 hours.
They use a value of ETTF = 75 yr for isolators (6.6 kVand 150 kV).
The repair time used is 50.0 hours.
Maintenance is performed once every two years during 3.0 hours.
3.3.36. Dortolina et al. (1991) use in a substation reliability evaluation study for circuit
breaker failures:
ETTF = 66 years for active failures;
ETTF = 33 years for passive failures.
The repair time used is, in both cases, equal 24 hours. Maintenance is performed once
every three years during 24 hours.
They use a value of ETTF = 500 years for switches, with a repair time of 12 hours.
3.3.37. Kj"lIe and Sand (1991) give the following values for circuit breakers in
distribution networks in Norway:
22 kV rural network
ETTF = 59 years;
repair time = 6.0 hours.
11 kV urban network
ETTF = 77 years;
For disconnectors they give the following values:
22 kV rural network
ETTF = 71 years;
11 kV urban network
ETTF = 1400 years;
217E. For low-voltage circuit breakers the following value is used:
ETTF = 1650 years.
3.4. Ageing data
3.4.1. Wiseman [1976] presents "the
actual bath tub curve for power circuit
breakers". It is reproduced in Figure 4.
The curve is based on more than
600,000 breaker service years.
3.4.2. Mazza and Michacca [1981]
present the results of an enquiry on
circuit-breaker failures and defects in
service, observed during the period
1974-1977. The enquiry was
performed by Cigr6 Working Group
13.06. The survey concerns 77,892
breaker-years in service, obtained from
102 utilities in 22 countries.
0
...
~
hilUf. R8'8 r."..SHd in
pIorcent per ,..e

0
~ /
~ /
, 3
7 I II 10
AGE IN YEARS
Figure 4: Circuit breaker failure rate as a
function ofage, according to Wiseman [19761.
A distinction is made between major failures and minor failures. A major failure occurs
when the breaker can no longer perform all of its fundamental functions, or when
intervention within 30 minutes is necessary. All other failures are referred to as minor
failures. This survey is discussed in detail in Section 3.2.6.
Figure 5 shows the failure rate as a function of age. For minor failures, it shows a steady
decrease. For major failurs no trent is present .
c
,
-
,
u
10
g 08
- ~ .
.
-~

u _
~ .
, ~
-
-i E.o 6
u c
- .
= -
---
u
~ ~
~
0
~ : .04
_ u
-_ z
<
~ -
- -
~ !
< ~
~ 0
~ ~ .02
~ 0
~
_ x
- ~
< <
~ -
l-
-
-
-
Q

\

\
,
,
,
,

)$
\
\
I
2
P-
,
,
,
, ,
,
,
,
,
,
,

,

,

, Minor failures (.r)

Q
,
"15 Oefaillances .ineures

,
, ,
,
,
, ,

, ,
, ,
I
' ~
,
,
,
,
Major failures (MF)
p../ OHaillanees ujeures
,
,
"
--,
- "
.
" ,
r-- ;"',
// "',
,
/ '0..
" .... ...0...

,
' ..... --<l"' .... -'
'...,
l/ . ~
'<\
,
I
~
'1
4 6 8 10 12 YEARS
Figure 5: Relation between the failure rate and the age of circuit breakers according to
Mazza and Michaca [19811.
3.4.3. Kawamura et al. (1990) observed the fail behaviour of 77 kV air-blast type circuit
breakers used by The Kansas Electric Power Co., Kansai, Japan. The majority of the
breakers was manufactured between 1965 and 1970.
The causes of fault were divided in three categories.
initial failure (poor manufacturing, poor construction, etc.)
- random failure (natural disaster, contact of birds or animals, mistake
- wear-out failure (natural deterioration, poor maintenance, etc.)
Figure 6 shows the failure rate as a function of age. The upper figure for all failures, the
three lower ones after a subdivision for failure cause. The failure rate for all failures
shown no clear trent, whereas the initial failures show a decreasing failure rate and the
wear-out failures an increasing one.
Rat, of fults du 10 .11
TOlal numHf of tinll;1 bu-.ken: 9S(J
(17kV Air blllli typt')
Uud ,II'
(I) [nitial r.il.,1 ,-. Poor IUIIII.ehrilil.
Balhillb ear"
.: W,ar-ollt
_ I )1I;liol (.ilure \
: .......
(D) Random f.ilnt ... Nalull diusler.
(I) W,.r-od r.ihlrr ... Nahul dtteriontioA.
0.015
0.01
Contaet of birds.
.. im,b, .tt.
M iSlah. ete
10 IS 20 25
Poor mllintmann!
Figure 6: Bathtub for circuit breakers according to Kawamura et at. {1990}.
3.4.4. Bar et at (1990) use a Delphi-method to determine the position of the knee in the
bath-tube curve, i.e. the place where the wear-out phase starts to become important.
For low-voltage disconnect switches they find in case of
good circumstances 32 years;
average circumstances 26 years;
bad circumstances 17 years.
For medium-voltage minimum-oil circuit breakers in an MVtLV substation they find in
case of
good circumstances
average circumstances
bad circumstances
26 years;
20 years;
13 years.
For oil filled circuit breakers the values are in case of
For synthetic isolated circuit breakers they find in case of
For open circuit breakers they find
For medium-voltage minimum oil circuit breakers in an MV substation they find in case
of
good circumstances
and for air-blast breakers,
good circumstances
14 years
13 years
11 years
11 years
3.4.5. Horton et al. [1990] provide estimates of the failure rates of underground
distribution system components. The estimates are based on information collected from
a number of utility sources throughout the United States of America. For each
component and each year they used the number of units installed or removed and the
number of failures which have occured. This data is fitted to a Weibull distribution.
From 364,000 load break elbows installed by NELPA (Northwest Electric Light and
Power Association) utilities from 1968 to 1988 they find the following expression for
the failure rate:
A(t) = 0.09x10
3
t.
In terms of Weibull distributions this would imply:
shape factor = 2;
characteristic time-to-failure = 150 years.
As only 20 years of observation are available, the latter time has no relation to an ETTF
value found under the assumption of an exponential distribution (2400 years in this
case).
3.4.6. Kanai et al. [1991] determine parameters for the lifetime distribution of vacuum
interrupters from field data. They find a Weibull distribution with a characteristic number
of operations of 271,600 and a shape factor of 2.041. With the time as parameter they
find a characteristic lifetime of 3201 months (267 years) with a shape factor of 1.235.
3.5. Conclusions
The values for circuit breakers presented in this chapter are difficult to compare. Circuit
breaker failures include, among others, incorrect opening (either due to a failure of the
protection or due to a failure of the breaker itself) and short circuits in or near the circuit
breaker.
Table 20 and Table 21 summarize the values for "short circuit in or near a circuit
breaker". Table 20 gives recommended values and data from surveys. Table 21 gives
data used in reliability studies.
In Table 20 it has been assumed that "failure in electrical part" and "failure of electrical
insulation to ground" are synonym for "short circuit". A comparison with the other
values in the table shows that this assumption is probably not correct. From Table 20
the conclusion could be that the ETTF is in the range from 1000 to 5000 years, for LV
and MV circuit breakers. For higher voltages circuit breakers become less reliable. From
follow-up surveys by CIGRE and IEEE it follows that circuit breakers are more reliable
nowadays than in the past.
In Table 21 it has been assumed that "active failures" and "electrical insulation failures"
are synonym for "short circuit". It is remarkable that the data used in reliability studies
are systematically too low. No clear explanation for this has been found.
Table 22 and Table 23 summarize the ETTF values for disconnect switches. The results
from surveys differ over a wide range: from 100 to 6000 years. This is partly due to the
increase in reliability due to improved manufacturing and maintenance methods. But,
apparently, different types of disconnect switches have fairly different life times. This
is also made clear by the values presented by Volkmann et al. (1991).
The same spread in values is visible in Table 23, showing the data used in reliability
studies.
From the values presented here, an ETTF value in the range 250 - 1000 years seems
reasonable.
Section Reference ETTF Remarks
3.2.3 Selseth 19n 100 failure in electrical part
3.2.4 IEEE 1974 925 fixed type
1570 draw-out
3.2.6 Mazza and M;chacca 1981 4000 63-100 kV
550 100200 kV
600 200-300 kV
400 300-500 kV
60 >500 kV
3.2.12 Radwam and El-Marsafawy, 1990 19 66 k.V electrical insulation to grCM.l"ld
3.2.17 Bovy et al. 1991 5000 10 kV
Table 20. Summary of circuit breaker life times; short circuits in circuit
breakers; recommended values and data from surveys.
3.3.4 Grover and Billinton 1974 30 hinh-voLtafte
100 low-voltage
3.3.5 Guertin and Lamarre 1975 30 hiahvoltsae
100 low-vol taae
3.3.7 Allan et al. 19n 67 11 kv' active failures
3.3.12 Allan et at. 1980 6 400 kV' active failures
40 132 kV' active failures
200 11 kV' 3.3 kV 415 V active failures
3.3.13 Adams and Jasmon 1981 10 active fai lures
3.3.15 Ruoff and van Meteren 1983 380 normallYooen
750 normaLlv closed
3.3.17 Koval 1983 140
3.3.20 Anderson et al. 1987 230 electrical insulation failures
3.3.23 ALLan 1988 100
3.3.26 Fransen 1989 90
3.3.34 Allan et at. 1991 285 139 kV' active failures
3.3.36 Dortolina et al. 1991 66 active fai lures
Table 21. Summary of circuit breaker life times; short circuits in circuit
breakers; data used in reliability studies.
3.1.2 IEEE 1990 165 enclosed
3.1.3 Klo...,.1 et al. 1990 300 < lkY
200-600 10-30 kV
85-140 100 kV
3.2.1 Dickinson 1962 102 aDen
230 enclosed
3.2.2 Connor and Parkins 1966 7100 oB - ilrmersed
10 000 air-break
3.2.4 IEEE 1974 350 looen
160 enclosed
3.2.8 WallnBns 1986 1000
3.2.9 van Amelsfoort et at. 1986 6000
3.2.18 Volkmam et at. 1991 250 switch (underaround)
5200 elbow (underaround)
5200 sol ice (underaround)
1300 overhead switch' urban feeder
800 overhead switch' rural feeder
Table 22. Summary of disconnect switch life times; recommended
values and data from surveys.
3.3.2 Dickinson 1960 1000 integral with equi-"t
500 single throw
3.3.3 CaDre et al. 1969 5002000
3.3.4 Grover and Billinton 1974 50 short c i rcui t
5 maltriD
3.3.11 Allan et at. 1979 1000
3.3.14 Nelson and Johnson 1982 670 13.8 kV
450 24.9 kV
330 34.5 kV
3.3.17 Koval 1983 300
3.3.20 Anderson et al. 1987 500 electrical insulatfon faflure
3.3.21 Dialvnas and Allan 1987 1000
3.3.22 Goldberg et at. 1987 250 switches
1700 old splices/ elbows
17000 new sDl;ces/elbows
3.3.24 Dialynas 1988 75 150 kV
3.3.25 Dialynas and Paoadoooulos 1989 250 20 kV
3.3.28 Horton et at. 1989a 250
3.3.29 Horton et al. 1989b 250
3.3.32 Sallam et at. 1990 170
3.3.33 Mohan Roa and Sekhar 1990 160 enclosed
3.3.35 Dialvnas and Koskolos 1991 75
3.3.36 Dortolina et al. 1991 500
3.3.37 Kjolle and Sand 1991 71 22 kV rural network
1400 11 kV urban network
Table 23. Summary of disconnect switch life times; data used in
reliability studies.
4. PROTECTIVE EQUIPMENT; GENERAL AND RELAYS
4.1.1. IEEE standard 500 [IEEE, 1983] gives reliability data for components of nuclear-
from several surveys and data bases. For protective relays the following values are
recommended (catastrophic failures only):
All types:
Spurious operation
Fails to open
Fails to close
: ETTF = 2850 years;
: ENTF = 2 x 10
6
;
: ENTF = 300 x 10
3
;
Overcurrent alarm switches:
ETTF = 200 years
Thermal overload relay
ETTF = 250 years; range = < 65, 1000>
Overload, time delay relay
ETTF = 75 years
Under-voltage, time delay
ETTF = 90 years
Under-voltage, instant
ETTF = 50 year
Over-voltage
ETTF = 200 years
range =
range =
range =
(ENTF = Expected Number of operation To Failure)
-47-
<1200,11,000>
< 1 x 1 0
6
, 7 x 10
6
>
< 1 50 x 10
3
, 1 x 10
6
>
Protective equipment.
4.2.1. Dickinson [1962] presents the results of an AlEE survey in 1959. For "Protective
relays, including switchgear, starter overhead etc." the results are:
ETTF = 170 yr; c.i. = <160,180 yr>
The repair time reported is 5 hours.
4.2.2. Connor and Parkins [1966] report about a 14-year survey on faults in networks
with nominal voltages between 2 and 33 kV. The results for protective gear are given
below. There are approximately 50,000 protective gear in service. This figure includes
approximately 15,000 installations of current-release-operated equipment with time-limit
fuses.
The following subdivision according to cause is given for faults in protective gear:
relays and components: ETTF = 970 years; 9.1 %
incorrect settings: ETTF = 930 years; 9.5%
failure of trip supply: ETTF = 4800 years; 1.9%
AC trip circuit and t.1. fuses: ETTF = 350 years; 25.7%
wiring defects: ETTF = 2300 years; 3.8%
pilot cables: ETTF = 1700 years; 5.2%
incorrect connections: ETTF = 4800 years; 1.9%
incorrect circuit diagram: ETTF = 20,000 years; 0.5%
interference with secondary wiring: ETTF = 3700 years; 2.4%
testing errors: ETTF = 2700 years; 3.3%
vibration or mechanical shock: ETTF = 450 years; 20.0%
incorrect characteristic: ETTF = 1600 years; 5.7%
unkown at time of original report: 11 .0%
Based on 50,000 protective gear-units and a 14-year covering period, the following
values are found for "all causes":
ETTF = 89 years; c.i. = <87,91 >
The number of system faults caused initially by relays and protection is extremely small,
and on the average only about two cases occur per year (on a population of 50,000 that
implies: ETTF = 25,000 years). Cases of protection failing to operate correctly when
faults on the system occur are about 200 per year (this would imply ETTF = 250 years,
although this value is of no meaning as the number of faults occuring in the system is
not known). The average percentage of correct operations of relays for the past 14 years
is approximately 94%, and has tended to improve slightly in recent years.
The percentage of correct operations of H.V. fuses is now approaching 90%.
4.2.3. The IEEE sponsored 1972-survey [lEEE,1974] (discussed in Section 3.2.4)
resulted for protective relays in:
ETTF = 5000 years, c.i. = [2500,30000]
-48- Protective equipment.
The survey gives for the failure characteristic "mal-trip of circuit breaker":
fixed type : ETTF = 440 years;
draw-out : ETTF = 750 years.
For "failure of circuit breaker due to auxiliary or protective device" the survey results in:
fixed type : ETTF = 570 years;
draw-out : ETTF = 970 years.
4.2.4. Patterson and Teague [1974] present field experience with solid-state protective
relays over the period 1965-1973. Their results, as shown in Figure 7, are based on a
total of more than 10,000 equipment-years of service. The failure rate decreases from
0.25 per year to 0.03 per year, according to the authors, due to manufacturer's design
and quality control improvements and due to the user's increased familiarity with the
installation, setting testing and maintenance of the solid-state equipment.
,.
- -- - - - ., - -
CAlf NO""" Yf ......
IUCUUCAl COI""'fCTlONS
1165 ,_ 1M' ,... It" 1110 1"1 "" ""
C ... LtNO ..... vr ..... !
Figure 7. Field experience (left) and types of field problems (right) with
solid-state protective relays, according to Patterson and Taegue [19741.
4.2.5.The CIGRE survey [Mazza and Michaca, 1981] described in Section 3.2.6 gives
the following values for protection related failures.
Does not open on command:
63-100 kV : ETTF = 2400 years
100-200 kV : ETTF = 525 years
200-300 kV : ETTF = 220 years
300-500 kV : ETTF = 125 years
> 500 kV : ETTF = 40 years
Opens without command:
63-100 kV : ETTF = 5600 years
100-200 kV : ETTF = 1400 years
200-300 kV : ETTF = 550 years
300-500 kV : ETTF = 520 years
>500 kV : ETTF = 60 years
; c.i.
, c.i.
, c.i.
=
=
=
<1500,5000>
<400,700>
< 170,300>
; c. i. = < 100,175 >
; c.i. = <30,60>
; c.i.
=
< 3000,35000 >
; c.i.
=
<1000,2700>
; c.L
=
<400,1000>
; c.i.
=
<350,1250>
; c.L
=
<40,110>
4.2.6. Yoguchi et al. [1984] investigated 1037 failures, occured during April 1977 and
March 1980, on typical relay equipment for transmission line, busbar, transformer and
distribution line protection in Japan. A total of 14,755 pieces of equipment have been
observed.
They give the following data for protective relay equipment:
phase comparison relay equipment
ETTF = 7.5 years c.L = < 6.5, 8.5 >
directional comparison relay equipment
ETTF = 10.4 years c.i. = <9, 13>
differential relay equipment for transformers
ETTF = 70 years c.L = <60,90>
differential relay equipment for busses
ETTF = 20 years c.L = < 15, 25 >
The transmission line equipment (phase and directional comparison) has duplicated main
protection and complicated backup protection. This explains, according to the authors,
the low value of ETTF.
Data obtained for the elements of the equipment is given below.
Main relay (static type)
phase-comparison relay
undervoltage relay
overcurrent relay
distance relay
voltage balance relay
bus protective relay
logic element
Time
Auxiliary relay
Control power source unit
: ETTF = 265 years
: ETTF = 890 years
: ETTF = 1 500 years
: ETTF = 480 years
: ETTF = 470 years
: ETTF = 750 years
: ETTF = 75,000
: ETTF = 3,400 years
: ETTF = 230 years
For the main relay the percentages of unnecesarry and missing operating modes are 42%
and 58%, respectively.
The failures described previously were detected by the manual periodic test, automatic
monitoring, routine inspection and power system faults.
4.2.7. Franke [1987) gives data for 6 types of secundary equipment, obtained form 30
years of service experience in the power system of a chemical plant in the German
Democratic RepubliC. There seem to be however some contradictions in his data. I will
give here only his value for the probability of failure per interruption (mittere
Fehlverhaltens quote):
RD 7/ RD 110 "Complicated relay system"
ROU/ROS differential protection
RSZ 3f/RSf independent overcurrent time
RSZ 3t dependent overcurrent time
high-voltage automatic transfer system
low-voltage automatic transfer system
: 0.02 0.01 %
: 0.4 0.2 %
: 0.0061 0.0003 %
: 0.01 0.005 %
:0.3 %
: 0.03 %
4.2.8. Chan [1988) compares the performance of different protection arrangements. for
which he uses two relays.
Relay A
probability of fail-to-operate: 10 %
probability of mal-operation: 5%
Relay B
probability of fail-to-operate: 6 %
probability of mal-operation: 7 %
He uses two indices to quantify relay performance:
DSFPI (Discriminative System Fault Performance Index)
NSFPI (Non-system Fault Performance Index)
DSFPI = 1 _ Number of faults incorrectly cleared
total number of system faults
NSFPI = 1 _ non-system fault circuit-breaker operations
number of circuit-breakers installed
He gives values for these indices for the EHV. transmission and sub-transmission
systems of Hong Kong. (based of observations from 1981 to 1987)
DSFPI = 95.1 % ; c.L = < 94.96 >
From the NSFPI an ETTF for mal-trip can be determined:
ETTF = .--;;;1=",
1-NSFPI
This leads to the following ETTF values:
1981 ETTF = 16 years
He further uses two indices for relay performance:
the index of correct performance:
k = correct operations
C correct operations + mal operations
the index of fail-to-operate:
k
f
= fail-to-operate
correct operations + fail to-operate
The results for three types of distance relays are given in Table 24.
Relay 1 (43 units) Relay 2 (56 units) Relay 3 (115 units)
k, Ko k, Ko k, Ko
1980/82 0.2 X 99.3 X 1.3 X 91.2 X 0 90.5 X
1983/85 1.2 X 98.7 X 1.6 X 90.9 X 3.9 X 97.5 X
1985/87 1.4 X 99.0 X 2.5 X 98.3 X
2.6 "
90.5 X
Table 24: Relay data according to Chan [1988J.
4.2.9. Heising and Patterson (1989) give typical failures rates for protective relays.
Electromagnetic, time overcurrent
Electronic single function ( 1970's design)
Electronic relay system
: ETTF = 1000 yr.
: ETTF = 300 yr
: ETTF = 20 yr
The values are obtained from
published and unpublished industry
and general Electric sources.
Figure 8 gives the failure rate of
electronic relays and electromagnetic
relays as a function of complexity.
For very complex systems an
electronic relay will become more
reliable than an electromechanic one.
The authors also present some data
on the effectiveness of continuous
monitoring; over the period from
1977 to 1984 a total of 48 out of
55 failures which could result in
incorrect operation were detected by
continuous monitoring.
4.2.10. The Canadian Electric
Association [19901 gives detailled
failure data for high-voltage
100 r---------------
Number 01 Functions in a Protection System
Figure 8. Relay failure rate versus functional
complexity, according to Heising and Patterson
[1989J.
equipment. For each component they give a sub-component "control and protection
equipment". The results for this SUb-component are given in Table 25. Rated voltages
are from about 100 kV to 700 kV.
ETTF c.i.
Cables 8 . 5 ~ r [6 12]
Lines 19.8 yr [18 22]
Transformer banks 34.8 yr [31 39]
Circuit breakers 63 yr [57 70]
Synchronous compensator 7.4 yr [5 12]
Static compensator
88 da""
[75 1 05 days]
Shunt reactor bank 27 yr [20 40]
Shunt caDScitor bank 62 yr [40 110]
Series capacitor bank 600 days [450 900 days]
Table 25: Failure of control and protection equipment, according to the Canadian
Electrical Association 119901.
4.2.11. Vlutters [1991 J gives failure rates for earth leakage relays, derived from testing
of over 44,000 installation in Bavaria, Germany. A subdivision to the place of the
installation yields:
Workshops : ENTF = 80 ; c.L
=
<55,160>
Offices, meeting rooms : ENTF = 75 ; c.i.
=
<50,130>
Assembly halls : ENTF = 60 ; c.i.
=
<40,115>
Houses : ENTF = 35 ; c.i.
=
<30,45>
appartment buildings : ENTF = 28 ; c.i.
=
<24,32>
Cellars, heating rooms, etc. : ENTF = 20 ; c.i.
=
<14,32>
Outdoor equipment : ENTF = 13 ; c.i.
=
<8,50>
4.3. Data used in reliabilitv studies
4.3.1. Snaith (1977) uses the following values for the electricity supply of a nuclear
power station.
Time delay relay; probability of failure on demand: 1 %
under-voltage relay; unrevealed fault: ETTF = 100 years;
maintenance interval = 0.5 years.
4.3.2. Allan and Adraktas (1982) use the following failure rates for a reliability study.
The values are based on information given by [Green and Bourne, 1972) and [IEEE,
1974)
Fault detector
Relay
Trip signal device
Breaker
: 0.04 failures per year
(ETTF = 250 yr);
(ETTF = 2000 yr);
(ETTF = 300 yr);
(ETTF = 1250 yr).
All these refer to "dormant fail-to-trip" situations.
The "fault detector" includes appropriate current transformers, voltage transformers and
comparators. The "relay" contains operating and restraint coils. The "trip signal device"
contains the trip signal device and associated power supply. The "breaker" is the actual
fault breaking component.
4.3.3. Ruoff and van Meeteren (1983) use a value of ETTF = 10 years for incorrect
tripping of a circuit breaker. The probability of fail-to-open they use is 1 %.
4.3.4. Bunch, Stalder and Tengdin (1983) use past performance from a number of
electric utilities to establish reliability targets for automated distribution system
equipment which are deemed acceptable. They arrive at the following targets for the
protection:
false trip detected by self test
false trip not detected by self test
failure to trip detected by self test
failure to trip not detected by self test
: ETTF = 20 years;
: ETTF = 40 years;
: ETTF = 10 years;
: ETTF = 40 years.
4.3.5. Anderson et al. (1987) describe a study after the reliability of HV substations in
which they include the secondary equipment. For failures of the secundary equipment
they distinguish between (dormant) fail-to-trip and mal-trip. The second one is
incorporated in their study by increasing the failure rate of the circuit breaker, the first
one by introducing a failure probability at operation for the circuit breaker.
They have used the following data:
failure of circuit-breaker due to mal-trip
failure probability at operation
: ETTF = 800 years
: 0.5 - 2.0 %.
4.3.6. Allan [1988) uses in a reliability study for an incorrect trip of a circuit breaker
(thus for an incorrect trip of a relay) the value ETTF = 100 years. The value for the
propability of fail-to-open used is 6 %.
4.3.7. Fransen [1989) uses a value of ETTF = 140 years of incorrect interventions by
a relay or by an operator. The value is based on 7% years of failure data in the supply
to a large chemical plant. The repair time used is 4 hours.
4.3.8. Mohan Rao and sekhar [1990) compare the reliability of a number of distribution
systems.
Incorrect trip due to failure of a protective relay:
ETTF = 1700 years;
4.3.9. Dortolina et al. (1991) determined the reliability of three different designs for a
400 kV switching substation in Venezuela. They explicity include the behaviour of the
protective relaying system by extending the circuit-breaker representation.
Their circuit-breaker model is a logical series connection of a processing and decison
subsystem (PDs), a switching subsystem (55) and a auxiliary subsystem (AS). The latter
is considered totally reliable.
The PDs has an ETTF of 20,000 years for active failures; an ETTF of 39 years for
passive failures and a repair time of 12.0 hours.
The 55 has an ETTF of 65 years for active failures; an ETTF of 200 years for passive
failures and a repair time of 24.0 hours.
The probability of a breaker being stuck is 0.5 %.
4.3.10. Koglin et al. [1991) describe a reliability study of an EHV system. For the
protection they use the following data:
missing operation of protection for EHV line:
expected value of conditional probability for one end of the line: 0.46 %.
unnecessary operation of protection for switching bay:
expected value of conditional probability; 0.03 %.
4.3.11. Kula et al. [1991] include the reliability of the protection in the design of high-
voltage substations. They use the following data:
probability of failure of relay : 1 %
probability of failure of communication between relay and circuit breaker: 2%
probability of fail-to-operate of circuit breaker : 2%
probability of failure of the circuit-breaker failure protection : 2%
4.3.12. Lau et al. [1991) give failure data for components of modern computer-based
protective equipment. These data are based on a review of experience with similar
components in other applications and engineering judgment of a few experts in the field.
They give the following ETTF values:
logic processor : 6.7 years;
communication processor : 9 years;
input interface
input card
input protection
output interface
output card monitor Ckt
outputs (32)
overall card
watch-dog-timer
power supply
: 14 years;
: 7 years;
: 50 years;
: 14 years;
: 12 years;
: 20 years;
: 8 years;
: 12 years;
: 10 years.
All types of failure take 8 hours to repair. Periodic maintenance is carried out once in two
years to check out the failures that cannot be detected otherwise.
4.3.13. Kialle and Sand [1991] give the following values for control equipment in
distribution networks in Norway:
22 kV rural network
ETTF = 20 years; repair time = 0.8 hours;
11 kV urban network
4.4. Ageing data
4.4.1 . Gusciora [1988) treats bathtub curves, Weibull distibutions and relay reliability.
for a number of relays and related apparatrus he determined the parameters of the
weibull distribution for the lifetime.
For snap-action switches before process fixes he finds the superposition of two
Weibull distributions:
char. life = 7 million operations; shape factor = 0.9
For snap-action switches after process fixes he finds the superposition of two other
Weibull distributions:
Infant mortality of reed relays that had been declared marginally defective":
char. life = 18 billion operations; shape factor = 0.2
General-purpose cia per style heavy-duty relay type 1
char life = 410,000 operations; shape factor = 3.8
General-purpose cia per style heavy-duty relay type 3
char. life = 265,000; shape factor = 7.8
4.4.2. Bar et al. [1990) use a Delphi method to determine the position of the knee in the
bath-tub curve, i.e. the age where the wear-out phase starts to become important.
For an overcurrent-time relay in a MV-substation they find, in case of
average circumstances : 13 years.
The same values are given for polygon-shaped relays and for differential relays.
For signalling equipment they find, in case of
For the protective relay in an MV/LV substation they find:
Buchholz relay
good circumstances
bad circumstances
overcurrent-time relay
good circumstances
bad circumstances
: 21 years;
: 20 years;
: 20 years;
: 14 years;
: 9 years;
: 6 years;
differential relay
good circumstances
bad circumstances
: 15 years;
: 13 years;
: 10 years.
For the on-off coil in an MV-substation they find, in case of
For the auxiliary contacts they find, in case of
For the remote control they find, in case of
4.4.3. Ugokwe [1992] presents the results of a reliability case study on the offline
1774 programmable logic controller (PLC)". From a total of 20 failures of this type of
PLC they conclude that the failure rate consists of three Weibull-distributions, with the
following shape factors:
p = 0.457, 0 - 20 days;
p = 1.842, 7 - 25 days;
p = 2.185,60 - 150 days.
4.5. Conclusions
Failures of the protection can be divided into "fail-to-trip" and "incorrect trip. The latter
can be further divided into spontaneous fail-to-trip" and "fail-to-trip due to fault in
another zone". Unfortunately, many studies do not make this distinction. This makes it
difficult to compare the values.
Table 26 and Table 27 summarize the values for relay life time given in this chapter. The
values show a large spread. Several reasons for this can be given: difference in definition
of failure of the relay; different life times for different relay types.
Due to the limited number of values given, it is of no use to suggest some new values.
Some of the referred publications give recommended values that seem to be reasonable
[e.g. Heising and Patterson, 19891. I would suggest 250 to 1000 years for
electromagnetic relays; 100 to 200 years for electronic relays (single function) and 10
to 30 years for electronic relay systems. The life time of modern computerized relays is
not clear.
4.1.1 IEEE 1983 2850 spurious operation
200 overcurrent
250 thermal overload
75 overload- time delay
90 undervoLtage" time delay
50 instant
200 overvoltage
4.2.2 Connor and Parkins 1966 89
4.2.3 IEEE 1974 5000
4.2.4 Patterson and Teague 1974 430 new type of solid-state relay
4.2.6 Yoguchi et al. 1984 7.5 phase COll1l8"; son
10.4 directional comparison
70 transformer differential
20 bus differentiaL
4.2.8 Chan 1988 25 mal-trip
4.2.9 Helsing and Patterson 1989 1000 electromagnetic' time overcurrent
300 electronic' single function
20 electronic relay system
Table 26. Summary of relay life times; recommended values and data
from surveys.
4.3.1 Snaith 1977 100 undervoltage" unrevealed fault
4.3.2 Allan and Andraktas 1982 250 fault detector
2000 relay
300 triD signal device
4.3.5 Anderson et al. 1987 800 mal-triD of circuit breaker
4.3.6 Allan 1988 100
4.3.7 Fransen 1989 140 mal-trip due to relay or operator
4.3.8 Mohan Roa and Sekhar 1990 1700 mal-trip due to relay
4.3.13. Kjolle and Sand 1991 20 control eQUipment" 22 kV rural network
60 l' tv urban network
Table 27. Summary of relay life times; data used in reliability studies.
5. FUSES
5.1.1. IEEE standard 500 [IEEE, 1977] gives reliability data for components of nuclear-
form several surveys and data bases. For fuses the following values are recommended
(catastrophic failures only)
All types
Fuses (open) below rating
Fails to interrupt
Up to 1000 Volts
Fails to interrupt
: ETTF = 8800 years ; range = < 80, 30000 >
: ENTF > 100,000 cycles
: ETTF = 5500 years ; range = <550, 19000>
: ENTF > '00,000 cycles
Above 1000 Volts to 36.6 kV, indoor
Fuses (open) below rating : ETTF = 10,000 years; range = <700,30000)
Fails to interrupt : ENTF > 100,000 cycles
From 2.3 to 138 kV outdoor
Fails to interrupt
: ETTF = 12,500 years ; range = < 600, 40000)
: ENTF > 100,000 cycles
-60- Fuses.
was completed during 1972 [lEEE,1974). This survey included a total of 1982
industries in the United States and Canada. The survey resulted for fuses in the following
values:
ETTF = 500 years, c.i. = [250,3000).
5.2.2. The PNEM-survey [van Amelsfoort et aI., 1986) covering the period 1980 - 1986,
resulted for fuses in the following data:
ETTF = 4000 years, c.i. = [2500, 13000).
-61- Fuses.
5.3.1. Dialynas and Papadopoulos (1989) use a value of ETTF = 250 years for a 20 kV
fuse-cutout.
5.3.2. Whiting (1989) determines the reliability of power supplies to broadcast
transmitting stations. Component data are taken from various sources. For fused
switches they use:
ETTF = 500 years
5.3.3. Saliam et al. (1990) calculate, as an example, reliability indices for the MV-
network of Port. Fouad, Egypt. For fuses they use:
ETTF = 430 years
5.3.4. Volkmann et al. (1991) give a value of ETTF = 250 years for a fuse in an
underground distribution system. For the source of their data they refer to two internal
reports of Pacific Gas and Electric. For a fuse in an overhead distribution system they
give
ETTF = 220 years
ETTF = 270 years
(rural feeder)
(urban feeder)
These values were computed form historical outage records for a group of 85 rural and
95 urban feeders, over a period of approximately 5 years.
The repair times are.
rural overhead fuse
urban overhead fuse
rural underground fuse
urban underground fuse
: 155 hours
: 130 hours
: 200 hours
: 74 hours
-62- Fuses.
5.4. Ageing data
No ageing data for fuses has been found.
-63- Fuses.
5.5. Conclusions
Table 28 summarizes the values for fuse life time given in this chapter. From [IEEE, 1977)
it becomes clear that the probability that a fuse fails to interrupt is very small. The values
in the table are therefore, probably, only for "fuse mal-trips".
The values used in the reliability studies appear too low. I would suggest values between
1000 and 5000 years. One should keep in mind however that the probability of an
incorrect trip (or fail-to-trip) because of choosing the wrong fuse size, can be much higer.
5.1.1 IEEE 19n 5500 <100OV
10000 >1000V Indoor
12000 >1000V outdoor
5.2.1 IEEE 1974 500
5.2.2 van Amelsfoort et al. 1986 4000
5.3.1 Oletyne. end Papadopoulos 1989 250 fuse cutout
5.3.2 Whiting. 1989 500 fused aw; tch
5.3.3 Saltern et al. 1990 430
5.3.4 Volkmann et at. 1991 220 rural feeder
270 urban feeder
Table 28. Summary of fuse life times.
-64- Fuses.
6. VOLTAGE AND CURRENT TRANSFORMERS
6.1.1. IEEE standard 500 [IEEE,1984] gives reliability data for components of nuclear-
from several surveys and databases.
For instrument transformers the following values are recommended (catastrophic failures
only):
voltage transformers:
0-10 kV : ETTF = 325 years ; range = <145,425>
open circuit : ETTF = 1600 years ; range = < 1300, 2000>
<200,550> short circuit : ETTF = 400 years ; range =
over 10 kV : ETTF = 170 years ; range =
open circuit : ETTF = 1000 years ; range =
< 110, 300>
<600, 1700>
<140,400> short circuit : ETTF = 200 years
Current transformers
0-10 kV
open circuit
short circuit
over 10 kV
open circuit
short circuit
: ETTF = 500 years
: ETTF = 2300 years
: ETTF = 650 years
: ETTF = 350 years
: ETTF = 1500 years
: ETTF = 420 years
; range =
; range =
; range =
; range =
; range =
; range =
; range =
<250, 1000>
<1100,4500>
<350,1400>
<280, 1000>
<300,4200>
<350,1200>
6.1.2. Kloeppel et al. [1990] give recommended values of component data for reliability
studies. They were based on data from several industries in Eastern Germany as well as
from the public supply. The following values are recommended for voltage and current
transformers.
: repair time = 4 hours ; range = <2,6>
10,30 kV : ETTF = 500 year ; range = <250,1000>
110 kV : ETTF = 230 years ; range = <150,350>
-65- Voltage and current transformers.
6.2.1. Connor and Parkins [1966] report about a 14-year survey of faults in networks
with nominal voltages between 2 and 33 kV. Some results are given below:
current transformers : ETTF = 2700 years;
voltage transformers : ETTF = 17,000 years.
6.2.2. Cigre Working Group 23.07 performed a study after the reliability of instrument
transformers rated at 72.5 kV and above that consist of a paper-oil system [Cigre, 1989].
The survey covered a total of 2.31 million transformer-years (136.033 transformers
during 17 years).
The results of the survey are given in Table 29. All ETTF-values and confidence intervals
are in years:
violent failures non-violent failures
ETTF c. i. ETTF c. i .
current transformers 7500 [6000 10 000) 4000 [3800 4800)
magnetic voltage transformers 5300 [4000 7500) 3500 [3000 4300)
capacitive voltage transformers 20 000 [15000 33000) 3000 [2600 3500)
Table 29: failure data according to C i g r ~ survey [Cigre, 19891.
6.3.1. De Clerq et al. (1985) compare two methods for HV/MV substation reliability.
MV voltage and current transformers
ETTF = 500 years
6.4. Ageing data
6.4.1. Biir et al. (1990) use a Delphi-method to determine the position of the knee in the
bath-tub curve, i.e. the age where the wear-out phase starts to become important.
For current transformers in an MV/LV substation they find, in case of
For measurement transformers in an MV substation they find, in case of
6.5. Conclusions
Table 30 summarizes the results of this chapter. From both surveys the conclusion is
that the lifetime of voltage and current transformers is several thousands of years. The
data from other sources is not convincing enough to overrule this conclusion. I therefore
suggest an ETTF value between 2000 and 3000 years.
Section Reference vol tage current Remarks
6.1.1 IEEE 1984 325 500 < 10 kV
170 350 > 10 leV
6.1.2 Kloeooel et at. 1990 1000 1000 6 kV
500 500 10 30 kV
230 230 110 kV
6.2.1 Connor and Parkins 1966 17000 2700 2 - 33 kV
6_2.2 CIGRE 1989 2300 7500 > n.5 kV
6.3_1 Oe Clerq et al. 1985 500 500 Hedilll1 Voltage
Table 30. Summary of voltage and current transformer lifetimes.
7. GENERATORS
L.1...1 Green and Bourne [1972] give average component failure-rates for electrical
components. For generators they give:
AC, general : ETTF = 16 years
DC, general : ETTF = 13 years
Tachometers : ETTF = 23 years
Synchros : i:TTF = 14 years
7.1.2. IEEE standard 500 [lEEE,1984] gives reliability data for components of nuclear-
from several surveys and databases. For generators the following recommended values
are given (catastrophic failures only):
Steam turbine driven
fails once started
fails to start
Gas turbine driven
fails once started
fails to start
: ETTF = 140 years
: ETTF = 230 years
: ETTF = 2.2 years
: ETTF = 8 years
; range = <75,3000>
; range = < 120, 6000 >
; range = <0.5, 250>
; range = <1.5,1000>
7.1.3. The IEEE gold book [IEEE, 1991] gives recommended values for the components
of industrial and commercial power systems. For generators it recommends:
steam turbine driven generators in continous service
ETTF = 5.9 year
emergency and standby units (reciprocating engine driven)
ETTF = 185 hours
failure rate per start attempt: 0.0135
-70- Generators.
7.2.1. Dickinson (1962] presents the results of an AlEE survey after the reliability of
electrical equipment in industrial plants. The survey was held in 1959. For steam turbine
generators the results are:
ETTF = 700 days, c.L <600,900 days>
The failure data appears to be different for different locations, therefore some detail is
presented below:
USA and Canada : ETTF = 1800 days
Latin American : ETTF = 365 days
Middle East pipeline : ETTF = 270 days
The repair time reported is 111 hours.
;c.i. = <1300,2900days>
; c.L = <260,600 days>
; c.L = <210,360 days>
was completed in 1972 (lEEE,1974]. This survey included a total of 1982 equipment
failures that were reported by 30 compagnies covering 68 plants in nine industries in the
Unites States and Canada. For generators the following data were found:
Steam turbine driven
Gas turbine driven
driven by motor, diesel or gas engine
: ETTF = 32 year
: ETTF = 1.5 year
: ETTF = 15 year
; c.L = (16,45]
; c.L = (1.1,2.3]
; c.L = (6,00 >
7.2.3. Dopazo et al. (1976] have collected data to find the relation between unavailabilty
and time to maintenance. The data was derived from supercritical and cyclone-fired
generating units, in the period 1970-1975. Figure 9 shows the duration of maintenance
as a function of the time to maintenance. Figure 10 shows the forced unavailability as
a function of the time to maintenance.
7.2.4. O'Donnell [1980] describes the results of a survey after the reliability of
generators in industrial plants and commercial buildings. He gives the following data:
Continuous units (steam turbines)
Continuous units; age :s 10 years
Continuous units; age > 10 years
: ETTF = 6 years; c.L = <4.5, 8.5 >
: repair time = 32.7 hours
: ETTF = 1.7 years; c.i. = <1,2.5>
: ETTF = 8.3 years; c.L = <6,13>
Emergency standby units (reciprocating engines)
: ETTF = 185 hours; c.L = < 170,200 >
: probability of fail to start: 1.3 %, c.i. = <0.8, 1.9>
The survey was too small to draw any additional conclusions.
-71- Generators.
iii
?i
e
w
60
'"
~
~
0
"-
60
0
....
X
'"
Z
W
...J
40
30
20
10
O ~ ____ -L ______ L-____ -L ____ ~ ______ ~ ____ ~ ______ ~ ____ __
o 100 200 300 400 600 600 700
LENGTH OF PRECEDING RUN (DAYS)
Figure 9. Duration of maintenance as a function of the time time maintenance, according
to Dopazo et al. [1976J.
-72- Generators.
>
20%
>-
:::;
iii 18%
"
~
!
16%
z
::>
0
14%
w
~
0 12%
0
,.
10%
8%
6%
~ .
...--- 15% CONFIDENCE INTERVAL
4%
L>.
2%
O % ~ ~ - - ~ - - ~ - - ~ - - ~ - - ~ ~ - - ~ - - ~ - - - - - -
o 13 4-6 7-9 10-1213-'616-1819-2122-2425-27
MOUTHS SINCE PlANNES OUTAGE
Figure 10. Forced unavailability as a function of the time to maintenance, according to
Dopoza et al. [1976J.
7.2.5. Schilling et al. [1988] give failure data for 10 power plants in Brazil: four thermal
oil units and six thermal coal units. The data for all units cover the period since their
initial commercial commitment, thus representing their real commercial age. Some results
for these 10 units are given in Table 31. Note that the ETTF is given in days I.
Age X util. ETTF c. i.
(Years) (days)
sel 15.92 30.4 X 110 <85 150>
se2 15.25 32.2 X 80 <65 110>
se3 10.58 23.3 X 50 <40 65>
se4 10.00 24.3 X 50 <40 65>
jll 8.67 80.6 X 32 <27 40>
jl2 8.67 81.6 X 29 <24 36>
H3 8.67 86.3 X 50 <40 65>
il4 8.33 81.2 X 50 <40 65>
~ 1 5 3.25 67.9 X 25 <20 40>
j 16 2.5 71.6 X 31 <23 50>
Table 31: Failure data of 10 power plants according to [Schilling et al., 1988J.
The first column gives a code for the unit, sc1 - sc4 are oil units (2x81 MW, 2x218MW),
jl1 - jl6 are oil units (2x50MW, 2x66MW, 2x125MW). The third column gives the per-
centage of time that the unit was in operation.
-73- Generators.
7.2.6. Smith [1989) gives equivalent unavailability factors for cogeneration on power
plants. For utility cogeneration plants they find EUF = 13.4 %, c.i. = [8,24). The
planned unavailability is 9.0 %, the unplanned unavailability 4.4 %.
For industrial cogeneration plants they find EUF = 4.4 %, c.i. = [2,6). The planned
unavailability is 2.8 %, the unplanned unavailability is 1.6 %. All plants under study
were put into operation in the 1980's, had relatively high service factor, and an electrical
power between 75 and 100 MW.
(EUF = Expected Unavailability Factor: the percentage of time that a component is not
available due to forced as well as scheduled outages.)
7.2.7. Morzelle et al. [1989) have analysed the reliability of 76 UD-45 emergency sets
installed before 1980 and 175 sets installed between 1980 and 1988 and compared the
results with data obtained from sets installed before 1980.
Before 1980
failure to start
failure while running
1980-1988
failure to start
failure while running
: ENTF = 600 ; sample too small.
: ETTF = 425 hours ; c.i. = <300, 650>
: ENTF = 28,000 ; sample too small
: ETTF = 1000 hours ; c.i. = <800, 1400>
7.2.8. Ficek and Grolich [1989) followed about 90 steam turbines from one
manufacturer during 12 years (1974-1985). The authors conclude that these turbines
have a high availability. They give the following data:
Availability (ratio beteen the time of operation and the desired time of operation):
98.3 0.3%
Technical availability (percentage of time spend on normal operation, stand-by, and
non-operation due to external causes): 87.8 2.0%
Usefullnessfactor (percentage of time spend on normal operation and stand-by): 86.1
2.5%
Technical use factor (percentage of time spend on normal operation): 63.2 2.0%
Failure halt factor (percentage of time spend on non-operation due to failures): 1.0
0.2%
Stand-by halt factor (percentage of time spend on stand-by): 22.3 1.1 %
Maintenance halt factor (percentage of time spend on maintenance): 9.6 1.8%
7.2.9. Farmer [1989) presents some results of a study of diesel generator ageing in
nuclear power plants. The following subdivision of ageing failures, with resoect to the
cause, has been made:
vibration and shock
poor manufacturing
:28%
: 18%
-74- Generators.
adverse environment
maintenance errors
: 17%
: 9%
The following subdivision with respect to the location of the failure has been made:
governor control system : 30%
fuel oil supply system : 13%
the diesel engine itself : 10%
One of the major potential contributors to ageing turned out to be the fast start
requirements imposed by technical specifications for nuclear plants.
7.2.10. Smith et al. (1990) present the results of a reliability survey of 600 to 1800 kW
diesel and gas-turbine generating units. The survey was performed in 1988 and
sponsored by the U.S. Army Engineering and Housing Support Center. The data was
needed to support the analysis of power systems at command, control, communications,
and intelligence installations worldwide.
Plants were selected from a wide variety of applications (e.q. electric utilities,
cogenerators, hospitals, airfields, military installations, and computer and control
facilities). Twenty-two plants participated in the study, providing data on 708 unit-years
of operating experience.
A distinction is made between failure rates based on period hours (calendar time) and
operating hours (power production time).
Based on period hours the results are:
Diesel auxiliary
continous : ETTF = 93 days ; c.i. = <80, 105 days>
standby : ETTF = 290 days ; c.i. = <260,330 days>
Diesel package
continous: ETTF = 83 days ; c.i. = <75,90 days>
standby : ETTF = 380 days ; c.i. = <320,460 days>
Gas turbine
continous: ETTF = 80 days ; c.i. = <70,95 days>
standby : ETTF = 1160 days ; c.i. = < 900, 1500 days>
Based on operating hours the following results are given:
Diesel auxiliary - continous
- standby
Diesel package - continous
- standby
: ETTF = 11 days
: ETTF = 68 days
: ETTF = 31 days
: ETTF = 23 days
-75- Generators.
Gas turbine - continous
- standby
: ETIF = 49 days
: ETIF = 8 days
The high failure rates based on operating hours can be attributed to the relatively low
utilization of these units.
The mean time between planned outages (MTIPO) and the duration of planned outages
are given below, based on period hours
Diesel auxiliary
continuous
standby
Diesel package
continous
standby
Gas turbine
continous
standby
: MTIPO
=
28 days
: MTIPO
=
286 days
:MTIPO
=
62 days
: MTIPO
=
180 days
: MTIPO = 66 days
: MTIPO = 390 days
; duration = 1.3 hours
In the discussion to their paper, the authors give data on the probability of fail-to-start.
Here they refer to a number of other publications. I will reproduce their conclusions
below; where ENTF = Expected Number of starts To Failure.
The results of a number of gas-turbine starting reliability studies are given in Table 32.
Some results of diesel starting reliability studies are given in Table 33. The references
are given after Table 33.
S t ~ ENTF c.; .
1 210 <140 400>
2 155 <125 200>
3 125 <105 150>
4 128 <105 160>
Table 32: results of gas-turbine starting reliability studies.
Study ENTF c. ; .
1 33 not available
5 270 <220 350>
60 55 <35 130>
6b 200 ..... 1. too small
6c 300 sampl. too small
6d 55 <40 90>
6e 100 sample too small
7 11 not available
8 58 not eva; labLe
Table 33: results of diesel starting reliability studies.
1 ARINC Research Corporation. Final Report - RAM Study of Diesel and Gas-Turbine
Generator Sets. Publication 4219-03-01-4803, October 1988.
2 Booz, Allen Applied Research. Small Gas Turbine Start Investigation, April 1970.
-76- Generators.
3 Kongsberg Dresser Power. Internal Study Comparing Diesels with Gas-Turbine Engines
(unpublished). 1984.
4 AT&T. Internal Study for Gas-Turbine Reliability (unpublished). 1980.
5 Electric Power Research Institute. Reliability of Emergency Diesel Generators at U.S.
Nuclear Power Plants. NSAC 108. September 1986
6 U.S. Nuclear Regulatory Commission. Nuclear Computerized Library for Assessing
Reactor Reliability (NUCLARR). NUREG/CR-4639 EGG-2458. Volume 5. RX. June 1988.
6a Consumer Power Compagny - Big Rock point
6b Northeast Utilities - Milstone
6e Northeast Utilities - Connecticut Yankee
6d Commenwealth Edison Compagny - Zion
6e Consolidated Edison Compagny of New York Inc. - Indian Point
7 Institute of Nuclear Power Operations. Nuclear Plant Reliability Data Systems. 1982
Annual Report. 1983.
8 Electric Power Research Institute. Diesel Power Reliability at Nuclear Power Plants:
Data Prelimanary Analysis. NP-2433. June 1982
7.2.11. Yotsumoto et al. (1990) give data for origin of failure in an engine-generator set.
Actual statistical data from
telecommunications offices are used. The
functional block diagram of an engine-
generator set. as they use it. is shown in
Figure 11.
For this configuration they give the
following probabilities of these blocks
contribution to a failure:
Main block
failed during operation
failed to start
Common block
failed to start
Detecting block
failed to start
Controlling block
failed to start
: 13 %
: 22 %
: 5 %
: 16 %
:4%
: 15 %
: 25 %
Detecting Block
Main Block
Temperature
Fuel system
Lubricating oil
Lubrication sYstem
pressure
Cooling system
Water flow
Generating system
Output voltage
Speed
I
l
Common Block
Controlling Block
~ t a r t - u p system
Distribution board
Fuel system
f- Cooling systefTl
Control board
Figure 11: Functional block diagram of
Engine-Generator set used by [Yutsomoto et
al., 19901.
7.2.12. Verplanke (1991) studied failures that occured in the power system of a
chemical industry in The Netherlands between 1970 and 1991.
Generators 11 kV
ETTF = 2 years; c.i. = <1.5.3.5>
-77- Generators.
7.3.1. Snaith [1977] uses the following values for the electricity supply of a nuclear
power station.
Diesel generator
probability of failure to start : 3%
ETTF, while running : 300 hours
Diesel generator circuit breaker
probability of failure to close on command : 1.6%
7.3.2. Magnon et al. [1977] use a fault tree to determine the reliability of stand-by diesel
generator sets in nuclear plants. From data for components of the diesel generator (3000
kW, 1500 rpm) they find the following reliability data:
probability of set failure in idle phase revealing at start up: 1.6%
(the set is tested every 15 days)
probability of failure to start owing to a fault at attempted start-up: 0.36%
failure in the steady running phase: ETTF = 15 days.
7.3.3. Allan et al. [1980] use, in a reliability study of electrical auxiliary systems of a
power station, the following data for standby generations (23.5 kV):
ETTF = 30 days (while running)
average repair time = 78 h.
probability of failing to start = 10 %
7.3.4. Dialynas and Allan [1986] use the following generating units data for 1 MVA, 2
MVA and 10 MVA units in 33 kV and 11 kV distribution networks:
ETTF = 10 years
starting failure probability = 10% (20% for 1 MVA unit)
time to maintenance = 1 year
durating of maintenance = 168 hours
stuck breaker probability = 0.2%
network with local generation. They use the same values as in [Dialynas and Allan 1986]
transmiting stations. Component data are taken from various sources. He uses for 415
V, diesel standby generators:
ETTF = 320 days
-78- Generators.
7.3.7. In a study on energy planning for Curacao [Government, 1989] the following non-
availability values are used for generating units.
7.5 MW steam unit:
planned maintenance
forced outages
25 MW steam unit:
planned maintenance
forced outages
18 MW gas turbine; 3.1
planned maintenance
forced outages
: 8.2%
: 13.8%
: 12.3%
: 9.7%
and 3.75 MW steam units:
: 7.0%
: 3.0%
For diesel units the data are given in Table 34 (based on observation):
capacity in operation since planned maintenance forced outages
4.5 MW 1971 13.4% 8.4%
1.1 MW 1967 4.5% 13.8%
2.6 MW 1967 12.6% 13.8%
4.5 MW 1968 13.1% 22.4%
4.5 MW 1970 14.4% 23.4%
2.0 MW 19n 0.0% 9.4%
2.0 MW 19n 35.5% 6.2%
2.0 MW 1979 2.1% 17.3%
2.0 MW 1979 O.OX 3.3%
Table 34. Failure date for diesel units on Cura9ao {Government, 1989J
7.3.8. Billinton and Bebnath [1990] present an alternative four-state model for peak lead
units. They give the following non-availability data, based on 1989-1984 ERIS data. The
non-availability is defined as the chance that the unit is forced out when it needs to be
available.
(ERIS = Canadian Electrical Association, Equipment Reliability Information System.)
Combustion Turbine Units
1-9MW :8.41%
10-24 MW : 9.24%
25-49 MW : 16.73%
> 50 MW : 31.90%
Diesel generating units
1-4 MW : 7.72%
> 5 MW : 30.60%
In an example they use for standby generators:
probability of failure to start: 3 %, 4%, 5%
-79- Generators.
failure while operating: ETTF = 14 days.
7.3.10. Dialynas and Koskolos (1991) study an industrial power system with 5 local
generators: a parallel steam unit (6.88 MVA), two parallel gas units (6.47 and 6.22
MVA), and two stand-by diesel units (0.50 and 0.70 MVA). For all units they use a
forced outage rate (FOR) of 10 %. Maintenance is performed once a year during 168.0
h. For the stand-by units a starting failure probability of 20 % and a stuck breaker
probability of 0.2 % are used.
The FOR is defined as the percentage of time that the component is out of operation due
to failures (i.e. due to forced outages). The time that the component is out of operation
for maintenance (scheduled outages) is not included in the FOR.
7.3.11. Prescott et al. (1991) use a value of ETTF = 10 years for the failure of standby
generation to a large computer and communications installation. The repair time is 8
hours.
7.3.12. Patton [1992] investigates the effect of the duty cycle experienced by a
generating unit on the unit's failure rate and availability. He uses the following failure
data:
800 MW fossil-fired
600 MW fossil-fired
400 MW fossil-fired
200 MW fossil-fired
50 MW combustion turbine
: ETTF = 290 hours;
: ETTF = 290 hours;
: ETTF = 280 hours;
: ETTF = 540 hours;
: ETTF = 680 hours.
The following relation is used between the failure rate (per year) and the number of
starts per year (5):
A = Ao + 1.315 S ,
where "0 is the failure rate for zero number of starts (according to the above-given
values). The number of starts appears to have a considerable influence on the failure
rate. One start per month already more than doubles the failure rate.
The above relationship is based on an EPRI-study after the reliability of combined-cycle
units [Brown, Gardner, 1982)
-80- Generators.
7.4. Ageing data
7.4.1. Schilling et al. (1987) obtained failure data for large thermal generators. They fit
a number of expressions for the failure rate to the times-to-failure assuming a repair
process that is "as-bad-as-old".
Assuming a Weibull distribution, they obtain.
char. life-time = 2.353 months
shape factor = 1.709
7.4.2. Simpson and Stoll (1989) give a lot of data on availability and outage factors for
50 MW to 400 MW oil/gas fired subcritical units. Figures 12 through 15 give some of
the results obtained from the utility industry (North American Reliability Council). Figure
16 gives the forced outage rate as a function of the unit's age. A significant increase
occurs for units older than 25 years. This relation is obtained by using utility industry
data for units smaller than 200 MW.
PER CENT FORCED OUTAGE FACTOR
10r----------------------------------------------,
-- 10 HRS/ST
8
~ 300 HRS/5T
6
4
2
t)
25 HRS/ST
800 HRS/ST
120 HRS/ST
1500 HRS/ST
O ~ - - ~ - - - - ~ - - ~ ~ - - - L - - __ ~ ____ ~ __ _L ____ ~
o 2 3 4 5 6 7 a
THOUSANDS OF SERVICE HOURS
Figure 12. Forced non-availability versus annual service hours, for different values of the
service hours per start, according to Simpson and Stol/ [1989J.
7.4.3. Vesely [1990) determined ageing parameters for components of nuclear plants.
For fail-to-start of a diesel generator he finds.
ETTF = 0.96 years c.i. = <0.07,16>
shape factor = 1.7 c.i. = <1.2,2.4>
-81- Generators.
30
25
20
15
10
5
____________________
5 STARTS
STARTS
10 STARTS
80 STARTS
-:-
20 STARTS
150 STARTS

o 2 3 4 5 6 7
Figure 13. Failure rate as a function of annual service hours, for different values of the
number of starts per year, according to Simpson and Stoll {1989J.
20
15
10
5
__________________
5 STARTS
.0 STARTS
-
10 STARTS
80 STARTS
20 STARTS
.. -:- 150 STARTS
____ __ _L __ ____ __ ____ __
o 23456
7 8
Figure 14. planned unavailability as a function of annual service hours, for different
values of the number of starts per year, according to Simpson and Stoll {1989J.
-82-
Generators.
PER CENT AVAILABILITY

90
85
80
5 STARTS
40 STARTS
10 STARTS
80 STARTS
20 STARTS
150 STARTS
__ ____ __ ____ __ ____
a 1 2 3 4 5 6
7 8
Figure 15. Availability (forced and planned) versus annual service hours, for different
values of the number of starts per year, according to Simpson and Stoll [1989J.
PER CENT AVAILABILITY PER CENT FORCED OUTAGE RATE
90 14
-- AVAILABILITy FORCED OUTAGE RATE
12
85
10
/ B
80
.' 6
.'
".
'-.
-. -----_. 4
75
2

o 5 10 15 20 25 30 35 40
YEAR
Figure 16. Availability and failure rate as a function of age, according to Simpson and
Stoll {1989J.
-83- Generators.
7.5. Conclusions
Table 35 and table 36 give the ETTF values for generators. as presented in this chapter.
The values fall into two groups: life times of several days and life times of several years.
From the description of the different studies I concluded that the low value holds for
stand-by units and the high one for continous units. A serious descripancy in this are the
values for conituous units presented by Smith et al. [1990].
From the values presented here I would suggest 5 to 20 days for stand-by units and 1
to 3 years for continuous units.
Table 37 and Table 38 give the expected number of starts to failure (ENTFI. according
to the different sources. From the surveys the fail-to-start probability appears to vary
between 0.5 and 2%. The probability values used in reliability studies are in general
higher: 2 - 20 %.
7. 1. I Green and Bourne 19n 13 - 23 vears
7.1.2 IEEE 1984 140 years steam-turbine driven
2.2. years gas-turbine driven
7.1.3 IEEE 1991 5.9 years continous service
185 hours stand by
7.2. I Dickinson 1962 2 years steam-turbine driven
7.2.2 IEEE 1974 32 years steam-turbine driven
1.5 Years RBs-turbine driven
15 years engine driven
7.2.4 O/Donnell 1980 1. 7 vears continuous- age! 10 years
8.3 years continuous" age> 10 years
185 hours stand'by
7.2.7 Morzelle at al. 1989 425 hours stand by: before 1980
1000 hours standb)r; 1980-1988
7.2.10 Smith et at. 1990 II 49 days continuous
8 - 68 days stand'by
7.2.12 Ver"lanke 1991 2 vears
Table 35: Summary of generator life times: recommended values and data from surveys.
7.3. I Snaith 19n 300 hours
7.3.2 Magnon et a1. 19n 15 dayS stand'by
7.3.3 Allan et a1. 1980 30 dayS stand'by
7.3.4 Dialynas and ALlan 1986 10 years
7.3.6 WhHing, 1989 320 daYS stand by
7.3.9 Allan and IrI9!I'Rojas 1990 14 days standby_
7.3.11 Prescott et a1. 1991 10 years stand by
Table 36. Summary of generator life times: data used in reliability studies.
-84- Generators.
Section Reference ENTf Remarks
7.2.4 O'Donnell 1980 75
7.2.7 Morzelle et al. 1989 600 before 1980
28 000 19801988
7.2.10 Smith et at. 1990 210 gas turbine driven
155 gas turbine driven
125 gas turbine driven
128 Aasturbine driven
33 diesel driven
270 diesel driven
55 diesel driven
200 diesel driven
300 di esel driven
55 diesel driven
100 diesel driven
11 diesel driven
58 diesel driven
Table 37. Summary of fail-to-start data: data from surveys.
Section Reference ENTF Remarks
7.3.1 snaith 1977 33 diesel generator
63 c i rcui t breaker
7.3.2 Magnon et a l. 1977 51
7.3.3 Allan et al. 1980 10
7.3.4 Dialynas and Allan 1986 10 2 10 MVA
5 1 MVA
7.3.9 Allan and (noa-Roias 1990 20 - 30
7.3.10 Dialynas and Koskolos 1991 10
Table 38: Summary of fail-to-start data: data used in reliability studies.
-85- Generators.
8. "UNINTERRUPTABLE" POWER SUPPLIES
8.1. Recommended yalues
8.1.1. IEEE standard 500 [IEEE,1984] gives reliability data for components of nutlear-
from several surveys and databases. For uninterruptable power supplies the following
data are recommended.
All types
No output
Fail to transfer
Degraded
Incipient
: ETTF = 30 years ; range = <4, 600>
: ETTF = 20 years ; range = <3,450>
: ETTF = 57 years ; range = < 8, 1300 >
: ETTF = 57 years ; range = <8,1100>
Single-phase static inverter
No output : ETTF = 110 years ; range = < 10, 400 >
Fail to transfer : ETTF = 85 years ; range = < 7, 300 >
Degraded : ETTF = 350 years ; range = < 30, 1100 >
Incipient : ETTF = 165 years ; range = < 14, 600 >
Three phase static inverter
No output : ETTF = 40 years ; range = < 4, 600 >
Fail to transfer : ETTF = 30 years ; range = < 3, 450 >
Degraded : ETTF = 55 years ; range = < 6, 900 >
Incipient : ETTF = 120 years ; range = < 10, 2000 >
8.1.2. The IEEE gold book [IEEE,1991] recommends:
rectifiers : ETTF = 26 years.
inverters : ETTF = 291 days.
-86- 'Uninterruptable power supplies'.
8.2.1. The IEEE sponsored survey completed in 1972 [lEEE,1974) results in:
rectifiers : ETTF = 26 years ; c.L = [15.65).
inverters : ETTF = 300 days ; c.i. = [200,600 days).
-87- Uninterruptable power supplies.
8.3.1. The IEEE orange book [IEEE, 1987) states that large non-redundant UPS systems
can have an ETTF of 20,000 hours (2 years) using handbook reliability data; but field
experience indicates ETTF's in the order of 40,000 hours. A redundant system is 2 - 4
times more reliable than a nonredundant system.
8.3.2. Suntio and White [1988) use different types of UPS to optimize the power supply.
They use the following data:
Traditional Battery
ETTF = 100 years
maintenance checking one a year
SLA battery
ETTF = 50 years
maintenance checking once every 3 years
Rectifier
ETTF = 1 5 years
repair readiness 24 hours
maintenance checking once every year
8.3.3. Van der Vaart en Bouwkneght [1989) give data for the failure of UPS-systems.
They refer to this data as generally accepted figures.
rectifier : ETTF = 4.5 years;
battery : ETTF = 11 years;
inverter : ETTF = 3 years;
defective set isolator : ENTF = 20, N = number of inverter failures;
bypass system : ENTF = 20, N = number of inverter failures;
8.3.4. Schneider [1989] presents reliability parameters for components of UPS-systems.
Load-bus : ETTF = 8000 years ; repair time = 4 hours
DC-bus : ETTF = 8000 years ; repair time = 4 hours
Static switch : ETTF = 12 years ; repair time = 4 hours
Inverter : ETTF = 4 years ; repair time = 4 hours
1 battery : ETTF = 4 years ; repair time = 39 hours
2 batteries in parallel : ETTF = 1800 years ; repair time = 19 hours
8.3.5. Suntio et al. (1989) give the following data for a standby UPS:
Inverter : ETTF = 10.3 years
Battery charger : ETTF = 17.1 years
Transfer switch : ETTF = 22.8 years
For a true UPS they give
Inverter : ETTF = 3.5 years
Rectifier : 6.2 years
Bypass switch: 22.8 years
Figure 17 shows the difference between a standby UPS and a true UPS.
Public
p"""r
L.i ...
-
StatJ.c Dypua Switch
R.cti- Inver-
tier
1-,-
tor
aattery
S .. _
Transrer Swi tc:h
CRITICAL
AC
LOAD
Batt. In-
Char- _ ~
pI" ter
CRITICAL
AC
LOAD
Figure 17: True UPS (left) and standby UPS (right); from Suntio et al. [19891.
8.3.6. Bakker [1989] mentiones the following data, as given by compagny A for their
UPS.
Rectifier/charger : ETTF
=
9 years.
Battery : ETTF
=
14 years
Inventer : ETTF
=
6 years
Static interrupter (ultra rapid stoppage) : ETTF
=
60 years
static by-pass switch : ETTF
=
60 years
For Compagny B the following data were given:
Rectifier/charger
Battery
Inverter
: ETTF = 7.6 years
: ETTF = 9.5 years
: ETTF = 3.6 years
Static by-pass switch
: ETTF = 16.5 years
DC bus
AC bus
: ETTF = 11.4 years
: ETTF = 43.1 years
From this data he determines ETTF-values for emmergency supplies from 3
manufacturers. The results are given in Table 39.
Manuf. A. Manuf. B Manuf. C
Static Static Dvnamic
Standard UPS 24E" 20 yr 23 yr
Extra inverter 41 yr 29 yr
2 UPS chains 60 yr 33 yr 132 yr
3 UPS chains 60 yr
2 9 ~ 59 -"ir
Table 39: ETTF for different UPS types from
three manufacturers [Bakker, 19891.
8.3.7. Fiorina and Chevalier [1990] compare two methods for the reliability analysis of
UPS-systems. They use the following data (derived from Merlin Gerin studies on
elements):
Bypass switch : ETTF = 50 years
Inverter : ETTF = 10 years
Battery : ETTF = 14 years
Rectifier : ETTF = 8 years
;repair time = 6 hours.
; repair time = 6 hours.
8.3.8. Dialynas and Koskolos [1991] use data for failure of UPS for a reliability study of
an industrial power system. Redundant UPS's are used. They use the following data.
rectifier : ETTF = 27.4 yr ; repair time = 1 hour;
inverter : ETTF = 5.5 yr ; repair time = 1 hour;
main time to discharge battery: 8 hours;
probability of static interrupter not successfully isolating faulted inverter: 1 %.
8.3.9. Prescott et al. [1991] use a value of ETTF = 35 years for the failure of a UPS
plant.
8.3.10. Lee [1991] compares a dual-conversion UPS with a single-conversion on-line
UPS. Figure 18 shows both types. He states that for a dual-conversion UPS:
ETTF - 400 days
This low value is due to the constant heating and stress produced in the power
semiconductors.
The new single-conversion on-line designs are said to have a higher reliability:
ETTF - 2000 days.
-'"
SWkswiU'h
,-------lC
'" - -
'"
~ = - - - - a -
I!
- '--_...J--
- SappiJ1'DaCewimucWtyanUable
- Silpplrnme wIda taiUtJ fIiJun!
.- ~
-. l
T
- Supply rowt wttb oUlky .,aiiIbIt
- Supply I"Mlte wid! ~ . faihlre
Figure 18: Single conversion UPS (left) and dual conversion UPS (right);
from Lee [1991J.
217-E. He uses the following data:
Rectifier
Battery
Inverter
Static Switch
Electromagnetic transfer switch
: ETIF = 20 years;
: ETIF = 11 years;
: ETIF = 6 years;
: ETIF = 45 years;
: ETIF = 22 years;
-91-
Uninterruptable power supplies.
8.4. Ageing data
No ageing data for UPS's has been found.
8.5. Conclusions
Table 40 summarizes the life time values for components of Uninterruptable power
supplies. Table 41 summarizes these values for complete UPS systems. For batteries,
switches and busses no survey data were available. This makes it hard to judge the
worth of the different values used in reliability studies.
The only survey available gives a lifetime of about 1 year for the inverter, making it by
far the most sensitive part of the UPS. Remarkably, all reliability studies use a higher
reliablity for the inverter: ETIF = 3 - 10 years. The lower values is however confirmed
by [IEEE, 1987) and [Lee, 1991) who give UPS lifetimes of 2 and 1 year, respectively.
All other studies use UPS lifetimes that are probably much too optimistic.
From the available information I would suggest 0.5 to 2 years for the inverter and 10 to
30 years for the rectifier.
Section Reference inverter rectifier battery bypass load-bus DC bus Remarks
switch
8.1.2 IEEE 1991 291 d 26 yr
8.2.1 IEEE 1974 300 d 26 yr
8.3.2 suntio and White 1988 15 yr 50100 yr
8.3.3 van der Veart 1989 3 yr 4.5 yr 11 yr 10
8.3.4 Scheider 1989
4 ~
4 yr 12 yr
8.3.5 suntio et ale 1989 3.5 yr 6.2 yr 22.8 yr true UPS
10.3 yr 17.1 yr 22.8 yr stand-by UPS
8.3.6 Bakker 1989 6 yr
9E
14 yr 30 yr manuf. A
3.6 yr 7_6 yr 9.5 yr 16.5 yr 43.1 yr 11.4 yr manuf. B
8.3.7 Fiorine et al. 1990 10 yr 8 yr 14 yr 50 yr
8.3.8 Dialvnas et at. 1991 5.5 yr 27.4 yr
100 *
8.3.11 Warren 1992 6 yr 20 yr 11 yr 22-45 yr
Table 40: Summary of UPS component lifetimes.
8.1_ 1 IEEE 1984 30 years no output
20 years fail to transfer
57 years de.raded
57 years inci2!ent
8.3.1 IEEE 1987 2 - 4 years nonredundant
4 - 16 years redundant
8.3.6 Bakker 1989 20 - 23 years standard UPS
29 - 41 years extra inverter
33 - 132 years 2 UPS chains
29 - 60 years 3 UPS chains
8.3.9 Prescott et ale 1991 35 years
8.3.10 Lee 1991 400 days dual conversion
2000 days single conversion
Table 41: Summary of UPS lifetimes.
9. CABLES AND ACCESSORIES
All values in this chapter are for 1000 meter of cable, unless otherwise noted I
9.1.1, Green and Bourne [19721 give average component failure-rates for electrical
components. For cables they give:
less than 1 kV ETTF
1 to 33 kV: ETTF
33 kV to 275 kV
: 190 years
: 26 years
: ETTF = 1 5 years
9.1.2. Kloeppel et al. [19901 give recommended values of component data for reliability
studies. They were based on data from several industries in Eastern Germany, as well
as from the public supply. They recommend the values below:
1 KV : ETTF = 17 years, range = <7,60>
: repair time = 12 hours, range < 1 0, 30 >
6, 10 kV : ETTF = 14 years, range = <7,50>
: repair time = 12 hours, range = < 10, 30>
20, 30 kV : ETTF = 11 years, range = <8, 50>
: repair time = 30 hours, range = < 15, 50 >
110 kV : ETTF = 11 years, range = < 10, 100>
: repair time = 40 hours, range = <30, 100>
9.1.3. The IEEE gold book [IEEE, 1991 1 recommends the ETTF values given in Table 42.
Type of cable ETTF (ye.rs) average median
repair time repair time
(hours) (hours)
Above ground and aerial
o 600 Volt 216 457.0 10.5
600 - 15 000 Volt
All 21.6 40.4 6.9
In trays above ground 33.0 8.9 8.0
In conduit above ground 6.2 140.0 47.5
Aerial cable 2.1 31.6 5.3
Below ground and direct burial
o - 600 Volt 78.6 15.0 24.0
600 - 15 000 Volt
All 49.4 95.5 35.0
In duct or conduit 49.7 96.8 35.0
Above 15 000 Volt 90.7 16.0 16.0
thenmoDlastic 600 15 000 Yolt 78.8 44.5 10.0
thermosetting 600 - 15 000 Volt 34.3 168.0 10.0
lpaper insulated lead covered 600 V - 15 kV 33.4 48.9 26.8
other type of insulation 600 - 15 000 Volt 16.6 16.1 28.5
Table 42: Recommended values according to IEEE Gold Book [IEEE, 1991 J.
-94- Cables and acessoiries
The last two colums give the industry average and the median plant average of the
actual downtime per failure.
The IEEE gold book recommends the following ETTF values for for cable joints with rated
voltages between 600 and 15,000 Volts ..
all types of insulation; in duct or conduit below ground
ETTF = 1150 year;
average repair time = 36.1 hours;
median repair time = 31.2 hours
thermoplastic
ETTF = 1325 year;
median repair time = 8.0 hours.
paper insulated lead covered
ETTF = 950 year;
The IEEE gold book recommends the following ETTF-values for cable terminations for
rated voltages between 600 and 15,000 kV.
thermoplastic
ETTF = 239 year;
thermosetting
ETTF = 3260 year;
paper insulated lead covered
ETTF = 1280 year;
9.2.1. Dickinson (1962) gives the results of an AlEE survey held in 1959 after the
reliability of electrical equipment in industrial plants. the survey covered 33 compagnies
with 58 plants. The results for power cables are reproduced below.
Lead-covered power cable
aerial : ETTF = 25 yr
above-ground conduit : ETTF = 410 days
underground conduit : ETTF = 12 yr
direct burial : ETTF = 22 yr
Non-leaded power cable
aerial
above-ground conduit
underground conduit
direct burial
: ETTF '" 5% yr
: ETTF '" 12 yr
: ETTF = 8 yr
: ETTF = 6 yr
The following repair times are reported:
Lead-covered power cable.
aerial : 5 hours
above-ground conduit : 5 hours
underground conduit : 111 hours
direct burial : 21 hours
Non leaded power cable
aerial
above-ground conduit
underground conduit
direct burial
: 93 hours
: 36 hours
: 53 hours
: 190 hours
For cable joints and terminations results are:
; sample too small for confidence interval
; c.i. = <300,750 days>
; c.i. = <9,15 yr>
;c.i. = <17, 34yr>
; c.i. = <4, 7% yr>
; c.i. = < 10, 16 yr>
; c.i. = <7,9 yr>
; c.L = <3, 12 yr>
leaded : ETTF = 550 year ; c.L = < 450, 700 yr>
non-leaded: ETTF = 310 year; c.i. = <260,380 yr>
The repair times reported are:
leaded joints! terminations : 15 hours
non-leaded joints/terminations : 17 hours
9.2.2. Connor and Parkins (1966) report about a 14-year survey of failures in networks
with nominal voltages between 2 and 33 kV. The records of some 130,000 km.years
were available. For underground cables the following values are found:
2-5 kV solid or resistance earthing: ETTF = 55 years;
6.6 kV solid or resistance earthing : ETTF = 19 years;
11 kV arc-suppression coil : ETTF = 19 years;
11 kV solid or resistance earthing : ETTF = 23 years;
33 kV solid or resistance earthing : ETTF = 58 years.
Table 43 gives the relative contributions of the different faulure causes to the failure
rate:
S.YS.tem vol tege 2-5 kV 6.6 kV 11 kV 11 kV 22 kV 33 kV
method of earthing sol id/res sol id/res arc sup co; 1 sot id/res sol id/res sol id/res total
hlll16n agency 49% 42X 27% 29% 43X 21X 30X
mechanical damage to sheath 4X 7% 3X 4X 5X 4X 4%
corrosion 4X lX 2X
insulation failure l1X l1X 20X 17% 7% 29% 17%
failure 5X 17% lOX lax 7% 11%
termination failure 7% 16X 20X 15X 20X 4X 16X
I joint fai lure l1X 13X ax 13X 2X 25X l1X
I Around subsidence 7% 1X 3X 5X 4% 4X
miscellaneous 7% 5X 4% 8X 7% 5X
Table 43. Contribution of different causes to power cable failures, according to Connor
and Parkins [1966J
The most serious single fault cause is human agency. This is (according to the authors)
mainly due to the increasing use of excavators and other mechanical aids. Since 1951
the fault rate due to human agency has increased in both urban and rural areas. The
most likely times for this type of fault to occur are round 9.00 am and round 3.00 pm.
For failures in the underground cable itself, follows from the above data:
2-5 kV solid or resistance earthing : ETTF = 67 years;
6.6 kV solid or resistance earthing: ETTF = 27 years;
11 kV arc-suppression coil : ETTF = 26 years;
33 kV solid or resistance earthing : ETTF = 82 years.
The lower failure rate of 33 kV cables, in comparison with 11 kV cables, is said to be
due to the fact that 33 kV cables are generally larger, mechanically stronger, and usually
buried deeper than 11 kV cables. The main failure causes for 33 kV cables are: insulation
failure and joint failure.
For cable joints they found:
11 kV
33 kV
: ETTF = 1100 years.
For cable terminations the results are:
11 kV
33 kV
: ETTF = 2500 years.
was completed during 1972 IIEEE,1974J. This survey included a total of 1982
industries in the United States and Canada. The results for cables are given in Table 44.
The last four columns give, respectively: the industry average of the actual downtime
per failure; the median plant average of the actual downtime per failure; the average
estimated time to fix the failure during 24 hour work day in case of repair of the failed
components and in case of replace with a spare.
Type of cable ETTF c. i. Bvesrege actual down time estimated restore
(years> (hours) time (hours)
; n d u s t ~ plant recall' replace
Cable all type of insulation 50 [40 60]
Above ground and aerfal
o 600 Volt 200 [100 2000] 457.0 10.5 20.8 39.7
600 . 15 000 Vol t
All 20 [15 30] 40.4 6.9 26.8 60.4
In trays above ArOlSld 30 [20 65] 8.9 8.0 49.4 119.0
In conduit above ground 6 [319] 140.0 47.5 19.8
Aerial cabLe 20 [15 30] 31.6 5.3 10.6 28.0
Below growd and direct burial
o - 600 Volt 75 [40 700] 15.0 24.0 26.8
600 - 15 000 Volt
All 50 [40 60] 95.5 35.0 20.4 26.8
In duct or conduit 50 [40 60] 96.8 35.0 20.9 26.8
Above 15 000 Volt 90 [50 450] 16.0 16.0 16.0
Cable - thenmoplastic
600 - 15 000 Volt 75 [50 140] 44.5 10.0 22.5 29.3
Cable - thermosetting
600 - 15 000 Volt 35 [25 50] 168 26.0 27.2 55.2
Cable - paper insulated lead
covered
600 - 15 000 Volt 35 [25 50] 48.9 26.8 17.3 18.3
CabLe - other type of insulation
600 - 15 000 Volt 17 [10 35] 16.1 28.5 23.2 44.8
Table 44: Cable data according to 1972 IEEE survey [IEEE,19741.
The survey resulted in the ETTF-values and corresponding confidence intervals for cable
joints for rated voltages between 600 and 15,000 Volts, as shown in Table 45.
type of joint ETTF c. i. average actual down time estimated restore time
(years) (hours) (hours)
industry i plant repair restore
All tvoes of insulation
all 1000 [800 2000]
above ground and aerial 1200 [600 7500]
in duct or conduit below ground 1150 [850 1800] 36.1 31.2 14.7 5.5
thermoelastic 1300 [900 2300] 15.8 8.0 12.6 22.0
thermosetting 800 [400 5000]
paper insulated lead covered 950 [600 2250] 31.4 28.0 30.0
Table 45. Data for cable joints according to the 1972 IEEE survey [IEEE, 19741
From the confidence intervals one can conclude that it is not realistic to consider
different ETTF-values for the different types of cable joints.
The survey result for cable terminations are given in Table 46.
type of termination ETTF c. ;. average actuaL down time estimated restore time
(vears) (hours) (hours)
industry 'plant reoair restore
All types of insulation 1650 [1350 2100]
o 600 VoLts
Above ground and areal 7900 [4500 32000] 3.8 4.0 8.0 8.0
600 - 15 000 Volt 1130 [850 1700] 198 11.1 34.6 40.6
Above ground and areaL 1130 [850 1700]
in trays above ground 3000 [1700 12000] 8.0 9.0 48.8 58.3
in condui t above ground 780 [435 4000]
aedal cable 540 [380 1000] 48.5 11.3 15.3 18.0
duct or conduit below ground 3300 [1850 13000] 25.0 23.4 28.8 30.0
thennoDLastic 240 [150 600] 10.6 11.5 12.0 12.0
thermosetting 3300 [2100 7000] 451 11.3 30.2 42.8
paper insulated lead covered 1300 [850 2500] 68.8 29.2 39.0 30.0
Table 46. Data for cable terminations according to the 1972 IEEE survey [IEEE, 19741
9.2.4. Braun [1986] presents some results of the failure data kept by the Northwest
Underground Distribution Committee of the Northwest Electric Light and Power
Association. Their annual report on component reliability is only concerned with natural
failures. All failures caused by abnormal external means, such as through dig-ins or
damage prior to installation are not intended to be included in the data. The author
mentiones that there are still problems with field people not reporting the material
failures. All failure rates reported should therefore he considered on the low side. For 15
kV-cable he finds the ETTF-values in Table 47.
Year HMWPE XLPE
175mn
2
220 1J1112
175 nm
2
220 nm
2
1973 179 yr. 115 yr. 596 yr. -
1974 392 yr. 114 yr. 304 yr. 287 Yr.
1975 223 yr. 107 yr. 488 yr. -
1976 135 yr. 126 yr. 342 yr. 94 yr.
19n 129 yr. 155 yr. 413 yr. -
1978 78 yr. 233 vr. 152 vr. 2011 vr.
1979 60 yr. 179 yr. 237 yr. 1341 yr.
1980 47 vr. 248 yr. 5363 vr. -
1981 30 yr. 169 vr. 1609 yr.
1982 34 yr. 2299 yr. -
1983 37 yr. 93 yr. 304 yr.
-
Table 47: Data for 15 kV cables according to Braun 19861.
The results for low-voltage (oS 600 Volts) cables are given in Table 48.
Figure 19 shows the ETTF for 175 mm
2
and 220 mm
2
, 15 kV, HMWPE-cables plus the
confidence interval for the consecutive years. The author only gives the size of the
population for 1983. It is assumed that the size of the population did not change. There
is a clear decrease in ETTF for 175 mm
2
as a function of time. No explanation for this
decrease is given. There is no significant change in ETTF for 220 mm
2
.
year ETTF year ETTF year ETTF
1969 107 yr. 1974
322 Yr. 1979 220 yr.
1970 129 yr. 1975
413 yr.
1980 335 yr.
1971 217 yr. 1976
304 yr.
1981 201 yr.
1972 260 Yr. 19n 220 yr. 1982 230 Yr.
1973 460 yr. 1978 227 yr. 1983 206 yr.
Table 48: Data for low-voltage cables according to Braun [1986J.
' ~ r - - - - - - - - - - - - - - - - - - - - - ~
!
5
,.
1 ~ - - - - - - - - - - ~ - - - - - - - - - - ~
,L-__________ ~ ________ ~
1In 1171 1113 1171 117. 1113
ob .. rved y .. r ob .. fwd , r
Figure 19. ETTF for 175 mm
2
(left) and 220 mm
2
(right) 15 kV HMWPE cables,
according to Braun [1986J. The shaded areas represent the 95% confidence intervals.
From the 1983 data, for which the population size has been given, confidence intervals
can be determined:
HMWPE, 175 mm
2
15 kV
HMWPE, 220 mm
2
15 kV
XLPE, 175 mm
2
15 kV
HMWPE, 25 kV
XLPE, low voltage
: ETTF = 37 years, c.i. = <33,41 >
: ETTF = 93 years, c.i. = < 73, 130 >
: ETTF = 300 years, c.i. = < 200,500>
: ETTF = 18 years, c.i. = <15,23>
: ETTF = 180 years, c.i. = < 160, 200>
For primary splices, 15 kV, molded rubber, the 1983 data results in:
ETTF = 470 years; c.i. = <400,550 years>.
For primary splices, 25 kV, molded rubber:
ETTF = 3000 years,
For primary s[lices, 35 kV, molded rubber:
ETTF = 2200 years.
In the two latter cases, there is not enough data for a confidence interval.
9.2.5. Ferran et al. [1986) give failure data for underground cables in the EdF-network
during the first half of the eighties. Table 49 gives the ETTF for 1 km of underground
MV network. The network under study had a total length of about 75,000 km. From the
text one can conclude that these values include all failures in the network, i.e. not only
those in the underground cable. The increase of the ETTF is significant.
year EnF c.; .
1982 11.2 yr. <10.9 11.4>
1983 11.9 yr. <11.6 12.2>
1984 13.4 yr. <13.0 13.7.
'82-'84 12. I yr. <11.9 12.3>
Table 49: Data for underground MV networks according to Ferran et al. [1986J.
The results for about 70,000 km of underground LV network are given in Table 50. From
the text one can conclude that these values include all failures in the network, i.e. not
only those in the underground cable.
year EnF c. i .
1982 12. I yr. <11.8-12.4>
1983 12.1 yr. <11.812.4.
1984 13.2 yr. <12.813.5.
1985 12.5 yr. <12.3'12.6.
Table 50: Data for underground L V networks according to Ferran et al. [1986J.
Table 51 gives the ETTF-values for underground MV-cables.
lQ)rignated paper isolation
Year EnF c. i.
1980 27.3 yr. <26.228.6.
1981 32.9 yr. <31.4]4.5.
1982 32.3 yr. <30.833.8>
1983 44.3 yr. <42.0-46.8>
1984 53.2 vr. <50.3-65.4>
'80- '84 36.0 yr. <35.336.8.
XLPE insuLation
Year EnF c. i.
1980 30.3 yr. <28.432.3.
1981 35.1 yr. <33.2,37.2.
1982 42.0 yr. <39.8-44.4>
1983 50.8 yr. <48.3,53.6.
1984 54.0 vr. <51.456.9>
'80- '84 43.8 yr. <42.8-44.8>
Table 51: Data for underground MV
cables according to Ferran et al.
[1986J.
I.
-
- .-
5.
00 00 000 000 00 .. ~ ..
/
e
/

/
~

.! 40
00 000 . ~ . 000
-Imprlgnated pap.,
/
"
/
"-XLPE
~
_r
-
31
Y
-
. - .. .. _. _. 00 000
2.
1980 1 1 U82 1983 1984
obMrved y .. ,
Figure 20. Trent in cable life time, according
to Ferran et al. [1986J.
The trent in reliablity is shown in Figure 20. It can easily be seen that the reliability of
cables increases. As the values for both isolation type are about the same and remain
about the same, one can conclude that the failure mechanism may have nothing to do
with the type of isolation. Before 1980 the newly installed cables were mainly paper-
insulated lead-covered; After 1980 mainly XLPE insulated 3-core twisted.
The values found for cable joints are shown in Table 52.
XLPE/XLPE ~ o i n t s Paper/XlPE joints
observed year ETTF observed year ETTF
1980 1600 yrs 1980 830 yr.
1981 1800 yr. 1981 760 yr.
1982 2000 yrs 1982 910 yr.
1983 4800 yr. 1983 no yr.
1984 14 000 yr. 1984 760 yrs
Table 52: Cable joint data according to Ferran et al. (1986J.
For cable terminations they give:
1980:
1981 :
1982:
1983:
1984:
ETTF = 6,000 years
ETTF = 20,000 years
ETTF = 50,000 years
ETTF > 20,000 years
ETTF > 200,000 years
9.2.6. One of teh Dutch regional electricity authorities, NV PNEM. performed a survey
after the reliability of components in their 10 kV network, covering the period 1980 -
1986 [van Amelsfoort et al.,1986). It resulted for cables in:
ETTF = 100 years, c.i. = [90,110J.
9.2.7. Waumans [19861 presents the result of failure registration in a number of medium
and low-voltage networks in the Netherlands, during 1979 through 1984. The results
are summarized in Table 53.
polymer insulated PBl:Ier fnsulated all twe.
non armored armored lead covered includinLl9ints and terminations
mediun voltage 103_yeafs ,. 1000 years 70_ years 5S years
low voltage 13 years 60 years 32 years 21 years
house service connection 48 years 42 year. 18 years
Table 53. Data for underground cables, according to Waumans {1986J.
9.2.8. During 1987 VDEN registered all failures of power system equipment in Dutch
medium-voltage and low-voltage distribution networks [VDEN, 19881. For the 400 Volt
networks they registered 2006 failures in 69,692 km of underground cable, leading to:
ETTF = 34.7 yrs, c.i. = [33,371.
Detailed results are given in Table 54.
For 10-12.5 kV cables the survey results in:
ETTF = 79 years c.i. = [74,861 (not including failure of a joint)
ETTF = 59 years c.i. = [55,63) (including failure of a joint).
For 10-12.5 kV the survey results in a subdivision of failure causes for underground
cables, joints and terminations.
Manufacture and assembly
Excavation
: 9 %
: 43 %
: 7 % Soil warping
Moisture
Pollution and wear
Others external
Others internal
: 5 %
:4%
: 14 %
: 16 %
~ r insulated lead covered
All hi lures
Excavations
Others
oolvmer insulated armoured
All failures
Excavations
others
polymer insulated non-armoured
All failures
Excavations
Others
ETTF (yrl c.;. (yrl
38 [35 42]
IT [70 86]
IT [70 86]
71 [60 88]
110 [90 150]
200 [150 300]
20 [18 22]
32 [28 36]
52 [45 62]
Table 54. Data for 400 Volt cables according to VDEN survey (VDEN, 19881.
9.2.9. Pijls (1988) analysed failures in the electricity supply to a number of chemical
plants during 8 years. For a 10kV cable connection he finds
ETTF = 90 year, c.i. = <65,140 year>.
The typical cable length in the system under study is one kilometer.
70% of the failures occured in the cable itself;
20% in the cable connectors;
10% in the cable terminators.
This implies for failure in the underground cable itself:
ETTF = 125 years; c.i. = <80,275 years>.
From the failure data collected we find for cable terminations of 10 kV cables:
ETTF = 1850 years, c.L = < 900, ->
9.2.10. A group of experts on synthetically insulated MV-cables performed a survey
after the reliability of MV-cables with different types of insulation IBlechschmidt,1989).
Results were obtained from a population of almost 500,000 km.year.
The results of the survey over the period 1980-1986 are given in Table 55. The results
of the survey over the period 1980-1988 are given in Table 56 for LOPE and XLPE
insulated cables [Blechschmidt et al.,19911. Table 57 gives the results of the survey
over the years 1987 and 1988, i.e. the difference between the two tables above. The
differences between these three tables are considerable.
10 kV 20 kV 30 kV
ETTF c. i. ETTF c.i. ETTF c. i ..
XLPE 260 [240 285] 330 [305 360] 38 [33 45J
LOPE 112 [100 128] 34 [32 36) 26 [23 30)
PVC 25 [23 28) 18 [15 23) 7.3 [6 9)
EPR 32 [28 36) 80 [60 115) 39 [33 46)
Table 55: Cable data over 1980 - 1986, according to Blechschmidt [19891.
10 kV 20 kV 30 kV
ETTF c.L ETTF c. i .. ETTF c. i ..
XLPE 325 [30OJ50) 250 [240 270) 35
LOPE 150 [135 170) 27 [26 28) 23 [21 26)
Table 56: Cable data over 1980-1988, according to Blechschmidt et al. [1991].
10 kV 20 kV 30 kV
ETTF c. i .. ETTF c. i. ETTF c. f ..
XLPE 540 [470 650) 183 [170 200) 33 [27 41)
LOPE 270 [200 400) 16 [15 In 17 [14 20)
Table 57: Cable data over 1987 and 1988 (difference of Table 56 and 57).
9.2.11. The Rural electrical
Association (a credit agency of
the U.S. Department of
Agriculture) conducted a survey
after failures of cables in 1988
and 1989 [Dedman and
Bowles, 19901. By observing a
population of about 400,000
km.year and over 9,000 cable
failures they find:
ETTF = 44 years.
They also registered the year of
installation of the cable. This
made it possible to determine
25o,----------------------------------,
200 1---------------t'I:--I--'I-------l

b

year of installation
the average failure rate during Figure 21. ETTF for cables of different year of
the passed life time of cables installation, according to Dedman and Bowles [19901.
as a function of the year of
installation. The results are
given in table 58. Figure 21 gives the inverse failure rate (ETTF) of cables of different
year of installation, during 1988 and 1989. The value for 1985 (ETTF = 651 years) has
been omitted for clearity. In general, older cables show more failures than younger
cables; but for very young cables the trent appaers to reverse.
year of observed ETTF (1 kin) c.i. (years)
installation kin. year
1970 13 298 36 vr <33 40>
1971 25 544 17 vr <16 18>
1972 26724 20 vr <19 21>
1973 31 588 42 vr <39 45>
1974 36 347 22 vr <21 23>
1975 36560 40 Yf <37 43>
1976 31 522 38 yr <35 41>
1977 36 569 72 vr <65 80>
1978 45 203 161 vr <140 180>
1979 39 710 89 yr <80 100>
1980 26 370 56 yr <51 62>
1981 19 823 215 yr <180 270>
1982 14 867 162 yr <135 205>
1983 13 303 242 yr <190 330>
1984 11 166 115 yr <95 145>
1985 10 420 651 yr <430 1300>
1986 6 526 96 yr <75 130>
1987 5 738 25 yr <22 29>
1988 2079 40 yr <30 55>
All 433 357 44 yr
Table 58: "Ageing" data according to REA survey [Dedman, Bowles, 19901.
9.2.12. Franke [1990) observed, during 12 years, failures in a chemical plant in Eastern
Germany. Early failures in the wear-in phase (1.5 .. 2 years) have been removed from the
data. For 6 kV cables he finds:
ETTF = 45 years; c.i. = < 30.70>
Included in this value are
single and multi-phase shortcircuits
short-circuits in cable joints
short-circuits in cable terminations
9.2.13. Horton et al. [1990) provide estimates of the failure rates of underground
distribution system components. Data on molded rubber splices was available since
1982 from NELPA (Northwest Electric Light and Power Association) surveys:
15 kV splices : ETTF = 325 years
25 kV splices: ETTF = 550 years
35 kV splices : ETTF = 400 years
; c.i. = <300,350>
; c.i. = <450,700>
;c.i. = <300,650>
9.2.14. Heiss and Schweer [1991] present the analysis of failure events during 5 years
in 3000 km of underground cables in compensated 20 kV networks. From this they find
for underground cables:
ETTF = 47 years; c.i. = <42,53 years>
70 % of the faults were caused by foreign influences (dredges, trench work and stones),
10% by atmospheric overvoltages (the systems under study contained some 5000 km
of overhead lines),
20% by material faults
74% of all fault locations concerned the underground cables,
6% the cable joints,
20 % the cable terminations.
This implies for failures in the undergroudn cable itself:
ETTF = 63 years; c.i. = < 56,73 >
9.2.15. Morrison and Arhart [1991] give data for cable system performance in the period
1961-1990. Data has been collected from 36 plants of a major industrial manufacturer.
Their results are given in Table 59. A subdivision in failure location is given in Table 60.
Incorporating only failures in the cable itself leads to the lifetime expectations given in
Table 61.
ETTF c.; .
All cables 41 years <35 49>
EPR 29 years <Z3 41>
HMllPE 26 years <ZO 38>
other 66 ye.rs <51 91>
Table 59: Cable lifetimes, according to Morrison and Arhart [1991J.
cable joints term
EPR 98X ZX
.
HMllPE 73X 10X 17X
Other BOX 10X 1 0 ~
Table 60: Failure location, according to Morrison and Arnhart [1991J.
ETTF c.;'
All cables 49 vrs [41 59]
EPR 30 yr. [Z3 4Z]
HMWPE 35 yr. [Z6 56]
Other 8Z vr. [6Z 1Z0]
Table 61: Only failures in the cable itself, according to Morrison and Arnhart [1991J.
The authors made a comparison between cables in a wet environment and cables a dry
environment. The reliability of cables in a wet environment is significantly lower than
that of cables in a dry environment. The holds expecially for EPR insulated cables where
54 failures occured in 43 circuit-miles of wet environment but none in 23 circuit miles
of dry environment. Some results are given in Table 62.
wet env; rorrnent dry env;rorment
ETTF c. i. ETTF c. i .
EPR 1 4 ~ . [11 19]
. .
HMWPE 23 yr. [17 35] 26 yrs [14 156]
others 73 yrs [53 115] 55 yrs [39 m
all cables 34 yr. [29 42] 73 yrs [53 116]
Table 62: Influence of environment, according to Morrison and Arnhart [1991].
9.2.16. Bovy et al. (1991) analysed the outage occurances in some 4000 km of
underground 10kV cable of a Dutch utility, during 6 years. For 1 km of underground
cable they find:
ETTF = 77 years, c.L = < 68,87 years>
ETTF (phase-to-earth faults) = 1100 years, c.i. = < 1500, 2500 years>
probability of multiple fault following phase-to-earth fault: 26%
9.2.17. Verplanke (1991) studied failures that occured in the power system of a
chemical industry in The Netherlands between 1970 and 1991.
50 kV and 11 kV cable connections:
ETTF = 180 years; c.L <120,300.
25 % of the failures occured in the cable joints, 75 % in the cable itself.
This implies for failures in the cable itself:
ETTF = 240 years.
9.3.1. Dickinson [1957) gives a typical value of ETTF = 12 years for 1 km of cable
2,400 volt and up. and a typical value of ETTF = 22 years for 1 km of cable 600 volt
and below.
Typical repair times are 28 hours if a spare length is on hand and 41 hours if the run is
too long for the spare length.
For cable joints in industrial power systems he gives a typical value of:
ETTF = 200 years.
For the repair time after failures in cable terminations he gives:
2400 volt and below
6900 volt and below
26,400 volt
: 10 hours
: 14 hours
35 hours and more:
9.3.2. Dickinson [1960) uses the following values for power distribution systems for
refinery process units. The values are based on several published papers and on records
obtained from refineries.
straight jOint : ETTF = 1000 years ; repair time = 40 hours.
tee joint. leaded : ETTF = 500 years ; repair time = 40 hours.
tee joint. rubber or equal : ETTF = 1000 years ; repair time = 30 years.
9.3.3. Capra et al. [1969) use the following data for a reliability analysis of 12 kV
underground distribution systems of Pacific Gas and Electric.
three-phase cable : ETTF = 32-160 years;
single-phase cable : ETTF = 80 - 320 years
For 12 kV cable terminations they use the following data:
600 A terminations : ETTF = 670 - 1250 years;
200 A terminations : ETTF = 400 - 1250 years.
They use the following estimated time range for different operations in underground
distribution systems:
receive alarm. travel to open substation breaker. or open main-line interruptor
receive trouble call. travel to open line protective device
identify new sectionalizing location
read fault locators for single-phase system
read fault locators for three-phase system
remove or replace grate
operate switch or automatic interruptor
-108-
: 15 - 30 minutes;
: 40 - 90 minutes;
: 3 - 6 minutes;
: 3 - 6 minutes;
: 6 - 12 minutes;
: 6 - 12 minutes;
: 2 minutes;
Cables and acessoiries
remove 200 A terminations
replace 200 A terminations
drive between adjacent sectionalizing points:
single-phase system
three-phase laterals
three-phase main
operate nonload-break disconnects
call for test
: 9 - 18 minutes;
: 8 - 17 minutes;
: 2 - 3 minutes;
: 3 - 6 minutes;
: 6 - 9 minutes;
: 3 minutes;
: 6 - 12 minutes.
9.3.4. Bocker and Kaufmann [1977] use the following values for 10 kV cable in a public
distribution network:
9.3.5. Lonsdale and Hitchen [1977] use the following value for underground cable in a
public distribution network in the Nothwest of England. The data were based on an
examination of system performance over recent years:
ETTF = 42 years.
9.3.6. Chang [1977] uses the following data for single-phase lateral cable to evaluate
distribution system design:
ETTF = 40 years;
For cable terminations he uses:
ETTF = 1000 years;
average repair time required in restoration work:
receive call, travel to location : 45 minutes;
interrogate fault indicators : 36 minutes;
operate load break connector/switch : 15 minutes;
locate fault with fault locator : 1 hour;
repair cable : 5 hours.
replace cable terminations : 6 hours.
9.3.7. McNab [1977] uses the following values for 33 kV cable in a public distribution
network in the South of Scotland:
9.3.8. Allan, Dialynas and Hormer [1979] use the following data for cables in a reliability
study of a distribution system:
ETTF = 31 years;
repair time = 33.5 hours;
time to maintenance = 4 years;
maintenance time = 8 hours.
9.3.9. Allan et al. [1980] use, in a reliability study of electrical auxilary systems of a
power station, the cable reliability data shown in Table 63.
ETTF repair time
23.5 kV 200 yr 120 h
11 kV 200 yr 48h
3.2 kV 200 yr 36h
Table 63: Cable data used by Allan et al. [1980}.
9.3.10. Nelson and Johnson [1982] compare three distribution voltages for the power
data (based on 1974 IEEE survey and on other available data) for cables:
13.8 kV: ETTF = 50 years;
34.5 kV: ETTF = 42 years.
For cable joints they use:
For cable terminations they use:
9.3.11. Ruoff and van Meeteren [1983] use a value of ETTF = 53 years for short-
circuits in underground distribution cable.
9.3.12. Williams and Mudge [1983] use a value of ETTF = 25 years for an 11 kV
incoming cable and of ETTF = 20 years for a feeding cable. The repair time used is 30
hours. The values have been drawn from distribution system statistics.
9.3.13. Billinton and Goel [1986] use the following data for cable sections in a reliability
analysis of an existing 14.4 kV distribution network in Saskatchewan, Canada:
ETTF = 59 years;
9.3.14. Dialynas and Allan [1986) use the following values for cables, when including
local generation in the reliability analysis of power distribution systems:
ETTF = 31 years;
time to maintenance = 4 years;
maintenance time = 8 hours.
9.3.15. Dialynas and Allan [1987) describe a reliability model for a power distribution
network with local generation. For cables they use:
ETTF = 30 years;
9.3.16. Goldberg et al. [1987) use the following values for an underground 21 kV feeder
of Pacific Gas and Electric:
HMWPE cable after 10 years of service:
ETTF = 32 years
XLPE cable in the first year of operation:
ETTF = 530 years
9.3.17. Billinton [1987) uses for cable sections in the Annaheim distribution system of
the Saskatchewan Power Corporation:
ETTF = 60 years;
9.3.18. Dialynas [1988) uses a value of ETTF = 120 years for 150 kV cable.
9.3.19. Brzozowska-Theil and Theil [1988) incorparate reliability in the switching of
medium voltage networks. For 20 kV-cable they use the following data:
ETTF = 7 years;
switching time = 0.5 hours.
9.3.20. Dialynas and Papadopoulos [1989) use a value of ETTF = 20 years for 20 kV
cables.
9.3.21. Fransen [1989) uses a value of ETTF = 195 years for short circuits in cables.
The value is based on 7 Y. years of failure data in the power system of a large chemical
plant. The author does not relate the ETTF value to a length of cable. Typical cable
lengths in the system under study are between one and two kilometers. The repair time
used is 24 hours. The author mentions 19 failures in 7 Y. years in about 500 cables. This
leads to c.L = < 125. 375 years>
9.3.22. Roos [1989) compares the supply reliability for different configurations of
medium voltage power systems. For cable strings he uses the following values:
Independent failure of cable string due to failure in cable (for 1 km of cable):
ETTF = 31 years;
Independent failure of cable string due to failure in station (for 1 cable string):
ETTF = 170 years;
transmitting stations. Component data are taken from various sources. For single-phase
cable he uses:
11 kV lead-covered cable:
ETTF = 180 years
11 kV pvc - insulated cable
ETTF = 400 years
415 V lead-covered cable
ETTF = 180 years
415 V pvc-insulated cable
ETTF = 400 years
kV feeder of Pacific Gas and Electric. They use the following data:
HMWPE cable after 10 years of service
ETTF = 32 years
XLPE cable in the first year of operation
ETTF = 530 years
9.3.25. Horton et al. [1989b] use their Distribution Reliability Assessment Model
(DREAM) to compute reliability indices for distribution feeders. They use the following
cable data:
HMPWE cable
ETTF = 32 years
repair time: 90 minutes
XLPE cable
ETTF = 530 years
In an example they use for cable connections:
ETTF = 10 years.
9.3.27. Sallam et al. [1990] calculate. as an example. reliability indices for the MV-
network of Port-Fouad. Egypt. For 3-phase cable they use the followinf values:
ETTF = 7.8 years
For cable terminations at 13.8 kV they use:
ETTF = 550 years
systems. They use the following data:
13.8 kV cable, conduit below ground
ETTF = 50 years
480 V cable, conduit above ground
ETTF = 230 years
cable terminations, 13.8 kV
ETTF = 550 years
cable terminations, 480 V
ETTF = 5000 years
9.3.29. Dialynas and Koskolos [1991] use a value of ETTF = 120 years for 1 km of
cable. The repair time used is 96.0 hours. The cable in their system is operated at 20
kV.
9.3.30. Volkmann et al. [1991] use a value of
ETTF = 32 years
for underground HMWPE-cable, and a value of
ETTF = 535 years
for underground XLPE-cable.
The average repair time, from historical outage data, is 302 minutes for cables in a rural
system and 229 minutes for cables in an urban system.
They refer to two internal reports of Pacific Gas and Electric as a source for these
values.
9.3.31. Kjelle and Sand [1991] give the following values for cables in distribution
networks in Norway:
22 kV rural network
ETTF = 20 years; repair time = 13.7 hours;
11 kV urban network
9.4. Ageing data
9.4.1. Bar et al. [1990) use a Delphi-method to determine the position of the knee in the
bath-tub curve, i.e. the age where the wear-out phase starts to become-important.
For paper-insulated-Iead-covered low-voltage cable they find, in case of
good-circumstances : 54 years
average-circumstances : 44 years
bad-circumstances : 28 years
For polymer insulated low-voltage cable they find, in case of
For paper-insulated lead-covered medium-voltage cable they find, in case of
For polymer-insulated medium-voltage cable they find, in case of
average-circumstances: 30 years
For "mass" terminations of low-voltage cables they find, in case of
For low-voltage synthetic terminations they find, in case of
For medium-voltage oil-filled cable terminations they find, in case of
For medium-voltage "dry" terminations they find, in case of
For medium-voltage "mass" terminations they find, in case of
9.4.2. Horton et al. [1990) provide estimates of the failure rates of the failure rates of
underground distrbution system components. The estimates are based on information
collected from a number of utility sources throughout the United States of America. For
each component and each year they used the number of units installed or removed and
the number of failures which have occured.
A Weibull-distribution has been assumed for the time-to-failure with repair as-bas-as-old.
Consequently the form of the failure rate is taken to be:
J.( t) = :;:. t
m
-
1
,
where m is the shape factor and 8 the characteristic time-to-failure.
From 764 failures in 98,797 km. years of 15 kV HMWPE unjacketed cable in operation
by San Diego Gas and Electric they find:
m = 1.3; 8 = 85 years.
Assuming a constant failure rate would have led to ETTF = 129 years.
From 226 failures in 86,191 km.years of 15 kV XLPE 175 mm
2
unjacketed cable in
operation by Northwest Electric Light and Power Association they find a constant failure
rate:
m = 1; 8 = 380 years.
An analysis of 18 years and 4700 km of 220 mm
2
cable again gives a constant failure
rate
m = 1; 8 = 1200 years.
The higher value for 220 mm
2
cable might be attributed to lower average voltage stress
in the larger cable.
9.5 Conclusions
If the results of surveys on cable reliability show one thing, it is the difficulty in obtaining
usefull fail data. Braun [1986] shows a large yea r-to-yea r variation in failure data, for the
same system. The year to year variation is considerably larger than the confidence
intervals. The number of failures in a certain year is thus no good predictor for the
number of failures in later years. Also Ferran et al. [1986]. Blechschmidt et al. [1991]
and Dedman and Bowles [1990] show this large variation. The change in failure rate is
due to, among others, ageing of cable types, introduction of new cable types, changes
in the amount of excavations. Variations might even be due to improved registration of
failures.
As an example, Figure 64 gives the prediction from 1980-1986 data and the value for
1987-1988, according to Blechschmidt [1989] and Blechschmidt et al. [1991]. For the
prediction a 99.5% confidence interval has been chosen.
voltage type 19801986 1987-1988
10 kV XLPE 230-300 years 540 years
LOPE 95-135 years 270 years
20 kV XLPE 290-375 years 183_1ears
LOPE 3137 y.ars 16 years
30 kV XLPE 30-50y.ars 33 years
LOPE 2032 y.ars 17 years
Table 64: Predicted value (1980-1986) and
actual value (1978-1988) for cable life time,
according to Blechschmidt et al. [1989, 19911.
For 10 kV cables the prediction is too pessimistic, for 20 kV cables it is too optimistic;
for 30 kV cables the prediction turned out to be fairly reasonable.
Despite these problems I will try to obtain some acceptable values for cable lifetimes,
Tables 65 and 66 summarize therefore the available data. Table 65 gives data from
suveys; Table 66 data used in reliability studies. The values hold for failures in an
underground cable itself (i.e excluding failures in cable joints and in cable terminations).
Between the strong variations, values between 40 and 75 years appear to be acceptable
for reliability studies. These values are indeed often used in reliability studies.
Data for cable terminations are summarized in Table 67 and Table 68. Values between
1000 and 3000 years seem to be reasonable.
Data for cable jOints are summarized in Table 69 and Table 70. Values between 500 and
2000 years seem to be reasonable.
Section Reference LV M'I Remarks
9.1.1 Green and Bourne 1972 190 26
9.1.2 Klo...,.1 et al. 1990 17 11-14
9.2.1 Dickinson 1962 22 armored
6 non-armored
9.2.2 COMer and Parkins 1966 2682 depending on nom;nal voltage
9.2.3 IEEE 1974 75 50-90
9.2.4 Brall"l 1986 250 122 M'I strOM Iv time dependent
9.2.5 Ferran et at. 1986 36 DBDer insulation
44 polymer insulation
9.2.6 van Amelsfoort et al. 1986 100
9.2.7 lJaLlOans 1986 42-60 .1000 oolvmer armored
13-48 103 oolvmer non-armored
1832 70 paper insulated lead covered
9.2.8 WEN 1988 35 79
9.2.9 pij Is 1988 90
9.2.10 Blechschmidt, 1989 7.3-260 depending on nominal voltage and on
tYPe of insulation
Blechschmidt et at., 1991 16-540 depending on nominal voltage and on
tYPe of insulation
9.2.11 Dedman and Bowles 1990 44 strong year-to-Year variation
9.2.12 Franke 1990 45
9.2.14 Heiss and Schweer 1991 47
9.2.15 Morrison and Arhart 1991 30-35
9.2.16 BoyY et at. 1991 77
9.2.17 VerplBnke 1991 150
Table 65. Summary of cable life times: recommended values and data from surveys.
-, , 7-
Cables and acessoiries
section Reference LV MY Remarks
9.3.1 Dickinson 1957 200
9.3.3 Capra et al. 1969 32-320
9.3.4 Boeker and Kaufmann 1977 20
9.3.5 Lonsdale and Hitchen 1977 42
9.3.6 Cheng. 1977 40
9.3.7 McNab 1977 42
9.3.8 Allan Dialynas and Hormer 1979 31
9.3.9 Allan et a1. 1980 200
9.3.10 Nelson and Johnson 1982 42-50
9.3.11 Ruoff and van Meeteren 1983 53
9.3.12 Will; oms end MudJje 1983 20-25
9.3.13 Billinton and Goel 1986 59
9.3.14 Dialynas and Allan 1986 31
9.3.15 Dialvnas and Allan 1987 30
9.3.16 Goldberg et al. 1987 32-530
9.3.17 Bil linton 1987 60
9.3.19 BrzozowskaTheil and Theil 1988 7
9.3.20 Olelynas end P!l"ld<>l1QUlos 1989 20
9.3.21 Fransen 1989 195
9.3.22 Roos 1989 31
9.3.23 Whiting, 1989 180 180 lead covered
400 400 lool_r Insuleted
9.3.24 Horton et al. 198ge 32-530
9.3.25 Horton et al. 1989b 32-530
9.3.26 Allan and InQaRojas 1990 10
9.3.27 Sallern et at. 1990 7.8
9.3.28 Mohan Roo end Sekher 1990 230 50
9.3.29 Dialynas and Koskolos 1991 120
9.3.30 Volkmann et a1. 1991 32-535
9.3.31 Kjelle and Sand 1991 20-22
Table 66. Summary of cable lifetimes: data used in reliability studies.
9.2.1 Dickinson 1962 550 leaded
310 non leaded
9.2.2 Connor and Parkins 1966 1000 11 kV
2500 33 kV
9.2.3 IEEE 1974 3300
9.2.5 Ferran et at. 1986 > 6000
9.2.9 Pi jls 1988 1850
Table 67. Summary of cable termination life times: recommended
values and data from surveys.
9.3.3 Capra et al. 1969 670-1250 600 A
400-1250 200A
9.3.6 Chan . 19n 1000
2200 24.9 kV
1650 34.5 kV
9.3.27 Sallam et at. 1990 550
9.3.28 Mohan Rao and Sekhar 1990 550 13.8 kV
5000 480 Vol t
Table 68. Summary of cable termination life times: data used in
reliability studies.
9.2.1 Dickinson 1962 550 leaded
310 non-leaded
9.2.3 IEEE 1974 1000
9.2.4 Braun 1986 470 15 kV
3000 25 kV
2200 35 kV
9.2.5 Ferran et al. 1986 4800 XLPE/XLPE
800 XLPE/paper
9.2.13 Horton et a l. 1990 325 15 kV
550 25 kV
400 35 kV
9.2.17 Verplanke 1991 180
Table 69. Summary of cable joint lifetimes: recommended values and
data from surveys.
9.3.1 Dickinson 1957 200
9.3.2 Dickinson 1960 1000 straigt joint
500 leaded tee joint
1000 rubber tee joint
2200 24.9 kV
1650 34.5 kV
Table 70. Summary of cable joint life times: data used in reliability
studies.
10. BUSBARS
10.1.1. Kloeppel et al. [19901 give recommended values of component data for reliability
studies. The values were based on data from several industries in Eastern Germany, as
well as from the public supply. For busbars they give a relation between the failure rate
of the circuit-breakers, AC.b.' and the failure rate of the busbar, Abusbar:
open busbars:
Abusbar = 0.1 * A
c
.
b
. * # c.b.
closed, air isolated busbars:
Abusbar = 0.05 * A
c
.
b
. * # c.b.
closed, solid isolated busbars:
Abusbar = 0.01 * A
c
.
b
. * # c.b.
Where # c.b. is the number of circuit breakers connected to the bus.
10.1,2. The IEEE gold book IIEEE,19911 recommends for switchgear buses (indoor and
outdoor):
insulated switchgear with rated voltages between 600 and 15,000 volts
ETTF = 885 yrs; repair time = 28 - 261 hours;
bare switchgear with rated voltages below 600 volts
ETTF = 1250 yrs; repair time = 27 - 550 hours;
bare switchgear with rated voltages above 600 volts
ETTF = 521 yrs; repair time = 17.3 - 36 hours.
All these values are for one "unit". The number of units is the number of connected
circuit breakers and connected switches.
The gold book recommends for bus ducts:
ETTF = 2450 years; repair time = 9.5 - 128 hours.
The unit is 1 circuit meter.
-120- Busbars
10.2.1. The 1972 IEEE sponsored survey [lEEE,1974] resulted in the following values
for switchgear busses: (the unit was the number of compartements).
All switchgear busses
insulated 600 - 15,000 Volt
bare 0 - 600 Volt
bare 600 - 15,000 Volt
: ETTF = 1 500 year
: ETTF = 600 year
: ETTF = 3000 year
: ETTF = 1 600 year
; c.i. = [1000,2300]
; c.i. = [400,1300]
; c.i. = [1500,12000]
; c.i. = [900,5000)
The survey resulted in the following values for bus ducts (1 unit = 1 circuit meter):
ETTF = 2500 years; c.i. = [1500,5500)
10.2.2. O'Donnell [1979) presents the results of an IEEE survey after switchgear bus
reliability of industrial plants and commercial buildings, performed in 1977. This survey
was held because a major controversy emerged in the results of the 1972 survey [IEEE,
1974). Insulated bus showed a higher failure rate than bare bus, but data were heavily
influenced by chemical industry. The new survey shows the opposite of this, with less
chemical industry influence. The main results are given below:
All buses : ETTF = 950 years ; c.i. = <750, 1300>
Insulated buses
above 600 V : ETTF = 890 years ; c.i.
=
<650, 1450>
Bare buses
all : ETTF = 1 000 years ; c.L
=
<700, 1700>
below 600 V : ETTF = 1250 years ; c.L
=
<850,2350>
above 600 V : ETTF = 500 years ; c.L
=
<300,1800>
The number of units is the number of switchgear connected circuit breakers and
connected switches.
The following data have been found for the repair time:
Insulated busses
repair "round clock" average repair time
median repair time
repair "normal hour" average repair time
median repair time
Bare busses
repair "round clock" average repair time
median repair time
repair "normal hour" average repair time
median repair time
-121-
: 87 hours
: 24 hours
: 430 hours
: 240 hours
: 39 hours
: 32 hours
: 154 hours
: 24 hours
Busbars
10.2.3. The PNEM survey [van Amelsfoort et al.,1986) resulted for busbars (the authors
refer to substations") in:
ETTF = 1800 years, c.i. = [1200,3200).
10.2.4. Pijls [1988] finds from the failure data of an industrial power system collected
during 8 years, for 10 kV busbar systems.
ETTF = 90 years, c.i. <40, ..... >
Most busbars in the system being studied consist of three rails.
10.2.5. Wahlstrom et al. [1988] describe the experience with new 84 kV-420 kV GIS
in Sweden, between 1974 and 1986. By observing a population of 27 GIS with a total
number of 102 bays they find:
ETTF = 40 years
The repair time is between 2 and 4 days.
This low value is due to the problems in manufacturing, installation and maintenance of
this new type of switchgear. This value is therefore not representive for GIS switchgear.
10.2.6. Bovy et al. [1991] analysed the outage occurances in the 10 kV networks of a
Dutch utility for a timeperiod of 6 years. For one section of a 10kV busbar they find:
ETTF = 10.000 years
ETTF = (phase-to-earth faults) = 5000 years
probability of mUltiple fault following phase-to-earth fault: 41 %
10.2.7. Verplanke [1991] studied failures that occured in the power system of a
chemical industry in The Netherlands between 1970 and 1991. He gives the following
data:
Busbars 50 kV and 11 kV
ETTF = 125 years; c.i. <75,300>
-122- Busbars
10.3.1. Snaith [1977) uses the following values for the electricity supply of a nuclear
power station:
3.3 kV busbar, per outer section
ETIF = 200 years;
maintenance interval = 1 year;
415 Volt busbar, per outer section
ETIF = 200 years;
maintenance interval = 0.5 years;
10.3.2. Bocker and Kaufmann (1977) use the following values for the 10 kVequipment
in a system substation 10/0.4 kV in a public distribution network.
ETIF = 300 years; repair time = 26 hours.
10.3.3. Allan et al. [1977) use the following values for 11 kV busbars in the electrical
auxiliary systems of power stations:
ETIF = 200 years;
time to maintenance = 1 year;
duration of maintenance = 20 hours.
10.3.4. Allan et al. [1980) use, in a reliability study of electrical auxiliary systems of a
power station, the busbar reliability data shown in Table 71.
ETTF repair time time to I duration of
maintenance
(vr) (h) (vr) (h)
11 tv 220 120 2 8
3.3. tv 220 48 2 8
415 V 220 24 2 4
Table 71.' Busbar data used by Allan et al. [1980J.
10.3.5. Adams and Jasmob [1981) use the following values for busbars in a distribution
system:
ETIF = 42 years; repair time = 3 hours.
10.3.6. Nelson and Johnson [1982) compare three distribution voltages for the power
data (based on 1974 IEEE survey and other available datal for one section of a bus:
13.8 kV : ETIF = 1580 years;
24.9 kV : ETIF = 1580 years;
-123- Busbars
34.5 kV : ETTF = 1580 years.
For the bus duct between the power transformer and the switchgear they use:
34.5 kV : ETTF = 100 years.
10.3.7. Ruoff and van Meeteren [1983] use a value of ETTF = 1000 years for busbar
faults in distribution systems.
10.3.8. William and Mudge [1983] use a value of ETTF = 5000 year, with a repair time
of 60 hours, for an 11 kV busbar. These values have been drawn from distribution
system statistics.
10.3.9. Koval [1983] uses the following values for secondary busses in an industrial
power system:
ETTF = 500 years;
replacement ime = 10 hours;
network with local generation. For busbars they use:
ETTF = 1000 years;
10.3.11 . Dialynas [1988] uses a value of ETTF = 45 years for 1 km of busbar (i.e ETTF
= 45,000 year for 1 meter of busbar).
transmitting stations. Component data are taken from various sources. He uses the
following data:
11 kV bus bars
ETTF = 200 years
415 V bus bars
ETTF = 500 years
an 11 kV bus they use the following data:
10.3.14. Allan and Inga-Rojas [1990] describe a method for distribution system
reliability. In an example they use for busbars:
-124- Busbars
ETTF = 200 years.
10.3.15. Sallam et al. [1990] calculate, as an example, reliability indices for the MV-
network of Port Fouad, Egypt. They use the following data.
Switchgear bus, connected to 2 circuit breakers
ETTF = 150 years
repair time = 26,8 hours
Switchgear bus, connected to 4 circuit breakers
ETTF = 75 years
10.3.16. Mohan Rao and Sekhar [1990] compare the relaibility of a number of
distribution systems.
They use the following data:
Switchgear bus, 480 V, bare, connected to 7 breakers
ETTF = 420 years
Switchgear bus, 13.8 kV, insulated, connected to 1 breaker
ETTF = 300 years
10.3.17. Allan at al [1991] present a reliability test system for distribution networks. For
bus bars they use the data given in Table 72.
permanent active ten.,orary
fai lures fai lures fai lures
33 kV 1000_yrs 1000 vrs 100 vrs
11 kV 1000 yr. 1000 yrs 100 yrs
Table 72: Busbar data used by Allan et al. [T99T}.
10.3.18. Dialynas and Koskolos [1991] use a value of ETTF = 45 years for permanent
failures on busbars. The repair time is 19.0 hours. Maintenance is performed every two
years during 4.0 hours. The busbars in the system under study are operated at voltages
of 150 kV, 20 kV, 6.6 kV and 380 V.
10.3.19. Prescott et al. [1991] use a value of ETTF = 200 years for the failure of an
HV distribution switchboard, with and average repair time of 120 hours. For the LV
distribution switchboard they use ETTF = 330 years and an average repair time of 24
hours.
10.3.20. Dortolino et al. [1991] use in a substation reliability avaluation study a value
of ETTF = 100 years for busbars, with a repair time of 24 hours. Maintenance is
performed once every 3 years during 8 hours.
-125- Busbars
10.4. Ageing data
10.4.1. Bar et al. [1990] use a Delphi method is determine the position of the knee in
the bath-tub curve, i.e. the age where the wear-out phase-starts to become important.
For the busbar in a low-voltage substation they find, in case of
For a low-voltage busbar in an MV/LV-substation they find, in case of
bad circumstances : 25 years;
For a closed busbar in and MV substation they find, in case of
For an open bus bar in an MV-substation they find, in case of
-126- Busbars
10.5. Conclusions
The failure rate of a busbar is dependent on the number of sections connected to the
busbar, as well as on the actual length of the bus duct. Some authors therefore give a
failure rate per section. The 1972 IEEE survey [IEEE, 1974] also gives a failure rate per
meter. Unfortunately not authors refer their data to the dimensions of the bus, making
a comparison with others difficult.
The results for busbars are summarized in Table 73 and Table 74. From this survey a
value between 500 and 2000 years for the ETTF of one section might be concluded. The
influence of failures per meter of bus duct appears to be neglectable.
Most reliability studies use a value of some hundreds of years for the whole busbar. This
is consistent with the results from surveys.
Section Reference ETTF cer Remarks
section meter busbar
10.1.2 IEEE 1991 1250 < 600 Volt- bare
885 600-15 000 Volt" insulated
521 > 600 Volt- bare
2450 All types
10.2.1 IEEE 1974 3000 < 600 Volt- bare
600 600-15 000 Volt" insulated
1600 600-15 000 Volt bare
2500 All types
10.2.2 O'Donnell 1979 1250 < 600 Volt- bare
890 > 600 Volt- insuLated
500 > 600 Volt- bare
10.2.3 van Amelsfoort et at. 1986 1800
10.2.4 Pi jls 1988 90
10.2.6 Bovy et al. 1991 10 000
10.2.7 Verplanke 1991 125
Table 73. Summary of busbar lifetimes: recommended values and data from surveys.
-127- Busbars
Section Reference ETTF ""r Remarks
section meter busbar
10.3.1 snatth 1977 200
10.3.2 BOeker and Kaufmam 1977 300
10.3.3 A.lIan et at. 1977 200
10.3.4 Allan et at. 1980 220
10.3.5 Adams and Jasmb 1981 42
10.3.6 He 1 son ard Johnson 1982 1580
10.3.7 Ruoff and van Meeteren, 1000
1983
10.3.8 Will i ams and Muclge 1983 5000
10.3.9 Koval 1983 500
10.3.10 Dialynas and Allan 1987 1000
10.3.11 Diahnas 1988 45 000
10.3.12 WhithiM. 1989 200 11 tv
500 415 Volt
10.3.13 Duke et at. 1989 750
10.3.14 Allan and InRa-Ro"ss 1990 200
10.3.15 Sallam et al. 1990 300
10.3.16 Mohan Rao and Sekhar 1990 3000 480 Volt
300 13.8 tv
10.3.17 Allan et al. 1991 1000
10.3.1S Dialvnas and Koskolos 1991 15
10.3.19 Prescott et al. 1991 200 HV
330 LV
10.3.20 Dortolino et at. 1991 100
Table 74: Summary of busbar life times: data used in reliability studies.
-128- Busbars
11. LARGE MOTORS
11 .1. Recommended values
11.1.1. Green and Bourne [1972) give average component failure-rates for electrical
components. For motors they give;
Motors in general
induction motors above 200 kW
below, 200 kW
synchronous motors
small, general motors
stepper motors
: ETTF = 11 years
: ETTF = 11 years
: ETTF = 23 years
: ETTF = 16 years
: ETTF = 29 years
: ETTF = 23 years
11.1.2. IEEE standard 500 [lEEE,1983) gives reliability data for components of
nuclear-power stations. The data has been derived by a Delphi-method combined with
the results form several surveys and databases. For motors the following values are
recommended (catastrophic failures only)
Induction squirrel cage, 150 kW and smaller
Fails to run once started : ETTF = 105 years
Fails to start : ETTF = 98 years
Induction squirrel cage, above 150 kW
Fails to run once started : ETTF = 185 years
Fails to start : ETTF = 185 years
Wound rotor induction
Fails to run once started
Fails to start
Single phase induction
Fails to once started
Fails to start
000>
Synchronous single phase
Fails to run once started
Fails to start
: ETTF = 120 years
: ETTF = 60 years
: ETTF = 175 years
: ETTF = 165 years
: ETTF = 200 years
; range = < 150, 1000 >
; range = <50, 1000>
; range = < 30, 350 >
; range <45, 500>
; range = < 35,475 >
; range = < 30, 400 >
; range = < 50, 4000 >
; range = < 15 000,250 -
; range = < 135, 200>
; range = < 170, 260>
11.1.3. Kloeppel et al. [1990) give recommended values of component data for
reliability studies. The values were based on data from several industries in Eastern
Germany, as well as from the public supply. For motors they recommend the
following values:
below 1 kV
6 kV
: ETTF = 23 years
: ETTF = 12 years
; range = <14,50>
; range = <1,10>
; range = <3,65>
; range = < 10, 50 >
-129- Large motors.
11.2.1. Dickinson [19621 gives the results of an AlEE survey held in 1959 after the
reliability of electrical equipment in industrial plants. For electric motors the results
are presented below.
Synchronous motors.
180 kW and larger: ETTF = 15 year ; c.i. = <10,20>
Induction motors.
180 kW and larger : ETTF = 9.2 year ; c.i. = <8, 11>
150 kW and below: ETTF = 19.5 year; c.i. = <19,20>
The repair times reported are:
synchronous motors : 209 hours
large induction motors : 162 hours
small induction motors : 41 hours
11.2.2. A 1981 survey [lEEE,19821 in the electric utility industry reported, for motors
750 kW and up and not over 15 years of age, for insulation-related failures:
ETTF = 30 years; c.i. = < 25, 40>
11.2.3. O'Donell et al. [1983,19841 performed a survey after the reliability of motors
of rated power 200 hp (150 kW) and higher. The population under study consisted of
1141 motors in 75 plants of 33 compagnies. A total of 360 failures occured in 5085
unit years i.e.
ETTF = 14years,c.i. = <12,16yr>.
A subdivision to type of motor is given in Table 75. DC motors appear to be somew-
hat more reliable than AC motors, but the difference is not significant.
ETTF c. i.
" DOD
induction motors: 14 yr. <12, 16 yr.>
79"
synchronous motors: 13 yr. <9, 19 yr.>
10 "
"wollld rotor" 19 yr. <13, 34 yr.>
8"
D.C. Motors 2S yr. <14 140 yr.>
3 "
Table 75: Subdivision to type of motor.
Table 76 gives a subdivision to rated voltage. The rated voltage has no influence on
the reliability of a motor.
ETJF c. i.
" pop
0-1 kV 13 yr. <10, 16 yr.>
27 "
1-5 kV 15 yr. <13, 17 yr.>
70"
5-15 kV 16 yr. <9 50 yr.>
3"
Table 76: Subdivision to rated voltage_
Table 77 gives a subdivision to rated power. High-power motors appear to be less
reliable than low-power motors.
-130- Large motors.
ETTF c. i.
150-375 kW 15 yr <13. 17 yr.>
375-3750 kW 14 yr <11. 17 yr>
3750-7500 kW 5 yr <3 12 yr>
Table 77: Subdivision to rated power_
Table 78 gives a subdivision to motor speed_ The higher the speed of the motor, the
higher its reliability_
ETTF c. i .
0-720 RPM 10 yr <8. 13 yr.>
721-1800 RPM 14 yr <12. 16 yr>
1801-3600 RPM 19 yr <15 26 Yr>
Table 78: Subdivision to motor speed.
The repair time found from this survey is:
97.7 hours in case of repair during normal working hours,
81.4 hours in case of repair round the clock,
18.2 hours in case of replacement by a spare motor.
11.2.4. Albrecht et a!. [1987] describe a survey after the reliability of large motors in
power stations. The survey was limited to generating units above 150 MW, low-
voltage motors above 75 kW and all medium-voltage motors, generator units that
went into service between January 1969 and December 1979.
The report covers information on 5797 motors in service at 132 generating units
owned by 56 utilities. From the total population they find:
ETTF = 29.4 years; c.i. = < 28,31 >
There is a large spread in
reliability among different
units, utilities and manufac-
tures. Figure 22 shows the
distribution of estimated
failure rate by unit. A statis-
tical analysis by the authors
shows that there is little
spread among units of one
utility. This leads them to
the conclusion that "a ma-
jor factor in the reliability of
the motors is a function of
who is using them".
40
FREOUENCY 30
(UNITS)
20
10
2

6 8 10 >10
ESTIMATED fAILURE RATE, ./. YEAR
Figure 22. Distribution of estimated failure per unit,
according to Albrecht et a!. [1987J.
-131-
Large motors_
Also among different manufacturers there is a large spread in reliability:
Manufacturer 1 : ETTF = 39 years ; c.i.
=
<34,45>
Manufacturer 2 : ETTF = 33 years ; c.L
=
<28,38>
=
<19,25>
=
<33,57>
=
<32, 54>
=
<5,8>
Manufacturer 7 : ETTF = 120 years ; sample too small
Manufacturer 8 : ETTF = 19 years ; c.i. = <28,31 >
None of them shows a confidence interval overlapping the overall confidence interval
<28,31 >.
A subdivision to failure mode yields for this survey:
bearing related
stator related
rotor related
: 41 %
: 37 %
: 10 %
11.2.5. Franke [1990] observed, during 12 years, failures in a chemical plant in
Eastern Germany. Early failures in the wear-in phase (1.5 ... 2 years) have been
removed from the data. He present the following data:
6 kV motors
motors < 1000 V
: ETTF = 14 years; c.i. = <10,25>
: ETTF = 17 years ; c.i. = < 12,25>
-132- Large motors.
11.3.1. Snaith [1977) uses the following value for the electricity supply of a nuclear
power station:
probability of failure to start of 415 V motor: 0.1 %.
11.3.2. Allan et al. [1980) use the following data in a reliability study of electrical
auxiliary systems of a power station:
11 kV motors
3.3 kV motors
415 V motors
: ETIF = 75 years
: ETIF = 100 years
: ETIF = 250 years
-133- Large motors.
11,4. Ageing data
No ageing data on large motors has been found.
-134- Large motors.
11.5. Conclusions
Table 79 summarizes the available data on large motors. From the surveys follows an
ETTF value between 15 and 30 years.
Section Reference ETTF
11.1.1 Green and Bourne 19n 11-29
11.1.2 IEEE 1983 40-175
11.1.3 Kloeppel et ale 1990 12-23
11.2 _1 Dickinson 1962 9.2-19.5
11.2.2 IEEE 1982 30
11.2.3 O'DonneL l 1983 1984 14
11.2.4 Albrecht et ale 1987 30
11.2.5 frenke 1990 14-\7
11.3.2 Allan et ale 1980 75-200
Table 79. Summary of motor life times.
-135- large motors.
12. References
Adams, R.N. and G.B. Jasman
TOPOLOGICAL ASPECTS OF DISTRIBUTION SYSTEM RELIABILITY.
In: Proc. of the 7th Power Systems Computation Conf (PSCC), Lausanne, 12 - 17 July
1981. Ed. by M.A. Laughton.
Guildford, UK: Westbury, 1981. P.256-260.
Albrecht, P. and J.C. Appiarius, E.P. Cornell, DW. Houghtaling, R.M. McCoy, E.l,
Owen, D.K. Sharma
ASSESSMENT OF THE RELIABILITY OF MOTORS IN UTILITY APPLICATIONS.
IEEE Transactions on Energy Conversion, Vol. 2 (1987), p.396-402.
Allan, R.N.
EFFECTS OF PROTECTION SYSTEM OPERATION AND FAILURES IN COMPOSITE
SYSTEM RELIABILITY EVALUATION.
Electric Power and Energy Systems, Vol. 10 (1988), p.180-189.
Allan, R.N. and A.N. Adraktas
TERMINAL EFFECTS AND PROTECTION SYSTEM FAILURES IN COMPOSITE SYSTEM
RELIABILITY EVALUATION.
IEEE Transactions on Power Apparatus and Systems, Vol. 101 (1982), p.4557-4562.
Allan, R.N. and R. Billinton, I. Sjarief, l, Gael, K.S. So
A RELIABILITY TEST SYSTEM FOR EDUCATIONAL PURPOSES - BASIC DISTRIBUTION
SYSTEM DATA AND RESULTS.
IEEE Transactions on Power Systems, Vol. 6 (1991). p.813-820.
Allan, R.N. and M.F. de Oliveira, J.A. Chambers, R. Billinton
RELIABILITY EFFECTS OF THE ELECTRICAL AUXILIARY SYSTEMS IN POWER
STATIONS.
In: Reliability of Power Supply Systems, London, 21 - 23 February 1977. Ed. by M.N.
Eggleton et al"
London: Institution of Electrical Engineers, 1977. P.1-6.
lEE Conf. Publication no. 148.
Allan, R.N. and M.F. de Oliveira, A. Kozlowski, G.T. Williams
EVALUATION THE RELIABILITY OF ELECTRICAL AUXILIARY SYSTEMS IN MULTI-UNIT
GENERATION STATIONS.
Proceedings lEE, Part C, Vol.127 (1980). p. 65-71.
Allan, R.N. and E.N. Dialynas, I.R. Horner
MODELLING AND EVALUATING THE RELIABILITY OF DISTRIBUTION SYSTEMS.
Allan, R.N. and T.O. Inga-Rojas
TIME SEQUENTIAL RELIABILITY ASSESMENT OF ELECTRICAL AUXILIARY SYSTEMS.
In: Advances in Reliability Technology Symposium, Liverpool, U.K., 18 - 20 April 1990.
London: Elsevier, 1990. P. 183-198.
-136- References
Anderson, B.G. and J.C. Declerq, P. van Miegroet, H. Rohsler, H.G. Schutte
THE RELIABILITY OF HV SUBSTATIONS WITH SPECIAL REGARD TO THE SECONDARY
EQUIPMENT.
Electra (CIGRE), no. 111 (1987), p. 17-23.
Aquilino, J.W.
REPORT OF TRANSFORMER RELIABILITY SURVEY - INDUSTRIAL PLANTS AND
COMMERCIAL BUILDINGS.
In: Industrial and Commercial Power Systems Technical Conference, 12 - 15 May 1980,
Houston, Texas. Ed. by. B.Bridger et al..
New York: IEEE, 1980. P. 31-39.
Aquilino, J.W.
REPORT OF TRANSFORMER RELIABILITY SURVEY - INDUSTRIAL PLANTS AND
IEEE Transactions on Industry Applications, Vol. 19 (1983), p.858-866.
Bakker, G.W.
BETROUWBAARHEIDS-ANAL YSE NOODSTROOMINSTALLATIES (Reliability analYSis of
emergency power installations; in Dutch).
Faculty of Electrical Engineering, Delft University of Technology, 1989.
Student report of practical work.
Bar, J.W.H. and H.l. Doppen, H. Juckers, H.M.A. Konings, P. Wiersma, H. Zuijderduin,
M.J. Voeten
ONDERHOUDSANAL YSE (Maintenance analysis; in Dutch).
Arnhem, The Netherlands: VEEN, Commissie distributievraagstukken 1990.
Billinton, R.
DISPERSION ANALYSIS IN DISTRIBUTION SYSTEM RELIABILITY EVALUATION.
In: Proc. of the 9th Power Systems Computation Conf. (PSCC), Cascais, Portugal, 30
August - 4 September 1987. Ed. by H. Glavitsch.
London: Butterworths, 1987. P.147-153.
Billinton, R. and R.N. Allan
RELIABILITY EVALUATION OF POWER SYSTEMS.
Boston: Pitman, 1984.
Billinton, R. and K. Debnath
GENERATING UNIT MODELING IN POWER SYSTEM ADEQUACY ANALYSIS.
Reliability Engineering and System Safety, Vol. 27 (1990), p. 155-165.
Billinton, R. and R. Goel
AN ANALYTICAL APPROACH TO EVALUATE PROBABILITY DISTRIBUTIONS
ASSOCIATED WITH THE RELIABILITY INDICES OF ELECTRIC DISTRIBUTION SYSTEMS.
IEEE Transactions on Power Delivery, Vol. 1 (1986), no. 3, p.245-251.
-137- References
Blechschmidt, H. et al.
INTERNATIONAL CABLE FAULT STATISTICS ON SYTHETICALLY INSULATED MV-
CABLES.
In: Int. Conf. on Electricity Distribution (CIRED). Brighton, U.K . 8 - 12 May 1989. Ed.
by. H. Lhussier et al ..
London: Institution of Electrical Engineers. 1989. P. 200-206.
Blechschmidt, H.H. and H. Klockhaus, L. Litters, H.D. Reus
FAILURE OF LDPE-/XLPE-INSULATED MEDIUM VOLTAGE CABLES IN THE FEDERAL
REPUBLIC OF GERMANY.
In: Int. Conf. on Electricity Distribution (CIRED), Liege, Belgium, May 1991.
Liege, Belgium: A.I.M . 1991. Paper 3.8, p.1-8.
Boeker, H. and W. Kaufmann
OPTIMIZING THE EXPENDITURE ON RELIABILITY OF SUPPLY IN SYSTEM PLANNING
FOR THE BENEFIT OF THE CONSUMER.
In: Reliability of Power Supply Systems. London. 21 -23 February 1977. Ed. by M.N.
Eggleton et al ..
Bovy, A. and K.W. Edwin, W.R. Nick, A.H. Schweer
PLANNING DISTRIBUTION NETWORK REDUNDANCY USING RELIABILITY METHODS.
In: Electric Power Systems Reliability Symp., Montreal, Canada, 16 - 18 September
1991.
Paris: CIGRE, 1991. Paper 1 B-5, p.1-6.
Braun, W.F.
RELIABILITY STUDY OF CABLE TERMINATIONS AND SPLICES BY ELECTRIC UTILITIES
IN THE NORTHWEST.
In: Industrial and commercial power systems Technical Conf . Cleveland. Ohio. 5 - 8 May
1986. Ed. by C.E. Becker et al ..
New York: IEEE. 1986. P. 75-85.
Braun, W.F.
RELIABILITY STUDY OF CABLE TERMINATIONS AND SPLICES BY ELECTRIC UTILITIES
IN THE NORTHWEST.
Brown H.W. and N.J. Gardner
RELIABILITY AND AVAILABILITY ASSESSMENTS OF SELECTED DOMESTIC COMBINED-
CYCLE POWER GENERATING PLANTS.
Palo Alto, CA: Electric Power Research Institute (EPRI), August 1982.
EPRI Report AP-2536
-138- References
Brzozowska-Theil, und G. Theil
VERBESSERING DER SCHAL TUNG VON MITTELSPANNUNGSVERTEILUNGSNETZEN MIT
BEROCKSICHTIGUNG DER ZUVERLASSIGKEIT (Improvement of switching of medium
voltage distribution networks, taking into account the reliability; in German).
Elektrotechnik und Informationstechniek, Vol. 105 (1988), no. 7/8, p.318-326.
Bruggeman, G. and J.E. Daalder, R. Heinemeyer, R.W. Blower
MV/LV DISTRIBUTION TRANSFORMER FAULT PROTECTION.
In: Int. Conf. on Electricity Distribution (CIREDI. May 1991, Liege, Belgium.
Liege, Belgium: A.I.M., 1991. Paper 1.14, p.1-6.
Bunch, J.B. and H.1. Stalder, J.T. Tengdin
RELIABILITY CONSIDERATIONS FOR DISTRIBUTION AUTOMATION EQUIPMENT.
Canadian Electrical Association (CEA)
FORCED OUTAGE PERFORMENCE OF TRANSMISSION EQUIPMENT FOR THE PERIOD
JANUARY 1,1984 TO DECEMBER 31,1988.
Montreal: Canadian Electrical Association, 1990.
Capra, R.L. and M.W. Gangel, S.V. Lyon
UNDERGROUND DISTRIBUTION SYSTEM DESIGN FOR RELIABILITY.
Chan, F.C.
PERFORMANCE ASSESSMENT AND CONTROL OF POWER SYSTEM RELAYING.
IEEE Transactions on Power Delivery, Vol. 4 (1988), p. 986-994
Chang, N.E.
EVALUATE DISTRIBUTION SYSTEM DESIGN BY COST RELIABILITY INDICES.
CIGRE working group 12.05. (CIGRE, 1983)
AN INTERNATIONAL SURVEY ON FAILURES IN LARGE POWER TRANSFORMERS IN
SERVICE.
Electra (CIGRE), nr. 88 (May 1983), p. 21-42.
CIGRE Working Group 23.07. (CIGRE, 1989)
INSTRUMENT TRANSFORMER TECHNOLOGY AND SERVICE BEHAVIOUR FAILURE
SURVEY.
Electra (CIGRE), nr. 125 (July 1989), p.113-117.
CIGRE Working Group 13.06. (CIGRE, 1991)
HIGH VOLTAGE CIRCUIT BREAKER RELIABILITY DATA FOR USE IN SYSTEM
RELIABILITY SURVEYS.
1991.
Paris: CIGRE, 1991. Paper 2-01, p.1-6.
-139- References
Connor, R.A.W. and R.A. Parkins
OPERATIONAL STATISTICS IN THE MANAGEMENT OF LARGE DISTRIBUTION
SYSTEMS.
Proceedings lEE, Vol. 113 (1966), p.1823-1834.
Declerq, J.C. and P. van Miegroet, H. Roshsler, H.G. Schutte.
SUBSTATION RELIABILITY-COMPARISON OF TWO METHODS AND SUITABLE
PRESENTATION OF RESULTS.
Dedman, J.C. and H.L. Bowles
A SURVEY OF URD CABLE INSTALLED ON RURAL ELECTRIC SYSTEMS AND FAILURES
OF THAT CABLE.
In: Rural Electric Power Conference, 29 April - 1 May 1990, Orlando, Florida. (Ed. by
C.A. Blackmon et al.)
New York: IEEE, 1990. Paper 02, p.1-7.
Dialynas, E.N.
MODELLING THE OPERATION AND VOLTAGE CONTROL EQUIPMENT IN THE
RELIABILITY EVALUATION OF POWER DISTRIBUTION SYSTEMS.
Archiv fur Elektrotechnik, Vol. 71 (1988). p.369-380.
Dialynas, E.N. and R.N. Allan
LOCAL GENERATION FACILITIES IN THE RELIABILITY EVALUATION OF POWER
DISTRIBUTION SYSTEMS.
IEEE Transactions on Power Systems, Vol. 1 (1986), no. 4, p.62-67.
Dialynas, E.N. and R.N. Allan
RELIABILITY MODELLING AND EVALUATION TECHNOLOGY FOR POWER
DISTRIBUTION NETWORKS WITH LOCAL GENERATION.
Proceedings lEE, Part C, Vol. 134 (1987), p. 348-356.
Dialynas, E.N. and M.P. Padadopoulos
RELIABILITY ASSESSMENT STUDIES IN DISTRIBUTION NETWORK OPERATION AND
PLANNING.
In: Int. Conf on Electricity Distribution (CIRED), London, 8 - 12 May 1989. Ed. by H.
Lhussier et al ..
lEE Conference Publication no. 305.
Dialynas, E.N. and N.C. Koskolos
EVALUATING THE RELIABILITY PERFORMANCE OF INDUSTRIAL POWER SYSTEMS
WITH STANDBY AND EMERGENCY GENERATING FACILITIES.
Electric Power Systems Research, Vol. 20 (1991), p. 143-155.
Dickinson, W.H.
ECONOMIC EVALUATION OF INDUSTRIAL POWER-SYSTEM RELIABILITY
AlEE Transaction Pt II, Vol. 76 (1957), p. 264-272.
-140- References
Dickinson, W.H.
EVALUATION OF ALTERNATIVE POWER DISTRIBUTION SYSTEMS FOR REFINERY
PROCESS UNITS.
AlEE Transactions on Power Apparatus and Systems, Vol. 179 (1960), p.1 10-124.
Dickinson, W.H.
REPORT ON RELIABILITY OF ELECTRICAL EQUIPMENT IN INDUSTRIAL PLANTS.
AlEE Transactions Pt. II, Vol. 81 (1962), p. 138-151.
Dopazo, J.F. and D.E. Tuite, E.L. Stein, H.M. Merrill
OBJECTIVE CRITERIA FOR POWER PLANT MAINTENANCE SCHEDULING.
In: Reliability of Power Apparatus and Systems, 23 - 24 September 1976, Montreal. Ed.
by M.F. Hashmi.
Piscateway, NJ: IEEE, 1976. P.97-107.
Dortolina, C.A. and J.J. Porta, R. Nadira
AN APPROACH FOR EXPLICITEL Y MODELING THE PROTECTIVE RELAYING SYSTEM
IN SUBSTATION RELIABILITY EVALUATION STUDIES.
IEEE Transactions on Power Systems, Vol. 6. (19911. p. 1373-1379.
Duckett, D.A. and C.M. Mc. Donough
A GUIDE FOR TRANSFORMER REPLACEMENT BASED ON RELIABILITY AND
ECONOMICS.
In: Rural Electric Power Conference, 29 April - 1 May 1990, Orlando, Florida. Ed. by
C.A. Blackmon et al ..
New York: IEEE, 1990. Paper C4, p. 1 -7.
Duke, A.J. and J.T. Maguire, J. Flaherty
RELIABILITY ASSESSMENT OF AN ELECTRICAL POWER DISTRIBUTION SYSTEM.
In: Reliability '89 Proceedings, Brighton, UK, 14 - 16 June 1989.
Warrington, UK: UKAEA, 1989. Paper 3c/3, p. 1 -16, Vol. 1.
Farmer, W.S.
AGING RESEARCH IMPACT ON COMPONENT AND SYSTEM RELIABILITY AND
MAINTENANCE IN NUCLEAR POWER PLANTS.
In: Int. Conf. on Operability of Nuclear Systems in Normal and Advance Environment
(OPERA). Lyon, France, 18 - 22 September 1989.
Paris: Soc. Fracaise d'Energie Nucl., 1989. Paper 58, p.1-8.
Ferran, J. and M. Boudaillier, J. Jacquenet
DISTRIBUTION CABLE PRACTICES AND DEVELOPMENTS IN ELECTRICITE DE FRANCE.
In: Annual conference ofthe Electricity Supply Engineers Association of N.S.W., Sydney,
August 1986.
Ficek, J. and J. Grolich
BETRIEBSZUVERLASSIGKEIT VON DAMPFTURBINEN (Reliability of steam turbine
operation; in German).
Skoda Review, no. 1 (1989), p.22-25.
-141- References
Filter, R. and A.S. Jones
SEPARABLE CONNECTOR SWITCHING PERFORMANCE.
IEEE Transactions on Power Delivery Vol. 5 (1990), p. 435-441.
Fiorina, J.N. and M. Chevalier
A SIMPLE APPROACH TO THE RELIABILITY OP UPS ARCHITECTURES.
In: Power Quality '90 Europe. Official Proc., Paris, 13 - 15 November 1990.
Ventura, CA: Intertec Commun., 1990. P.213-222.
Franke, K.
ZUVERLASSIGKEITSKENNGROBEN FOR INDUSTRIELLEN RELAISSCHUTZ (Reliability
parameters for industrial relay protection; in German).
Energietechnik, Vol. 37 (1987), p.286-288.
Franke, K.
ERMITTLUNG VON ZUVERLASSIGKEITSKENNWERTEN AUSGEWAHLTER
STARKSTROMTECHNISCHER BETRIEBSMITTEL AUS DER STORUNGSST ATISTIK EINES
CHEMIEBETRIEBS (Determination of reliability parameters of selected power engineering
components from the failure statistics of a chemical plant; in German).
Energietechnik, Vol.40 (1990), p.256-259.
Fransen, H.
DE ENKELVOUDIGE RESERVE IN DE ELEKTRICITEITSVOORZIENING VAN EEN
CHEMISCHE INDUSTRIE (The single redundancy in the electric power supply of a
chemical industry; in Dutch).
Elektrotechniek, Vol. 67 (1989), p. 719-722.
Gangel, M.W. and R.J. Ringlee
DISTRIBUTION SYSTEM RELIABILITY PERFORMANCE.
Goldberg, S. and W.F. Horton, V. Rose
ANALYSIS OF FEEDER SERVICE RELIABILITY USING COMPONENT FAILURE RATES.
IEEE Transactions on Power Delivery, Vol. 2 (1987), p. 1292-1296.
(Government, 1989)
A STUDY OF LONG-TERM ENERGY PLANNING FOR THE ISLAND OF CURACAO, 1989-
2003.
Curat;:ao, Netherlands Antilles: The Government of the Island of Curat;:ao, 1989.
Governmental report.
Green, A.E. and A.J. Bourne
RELIABILITY TECHNOLOGY.
London: Wiley, 1972.
Grover, M.S. and R. Billinton
A COMPUTERIZED APPROACH TO SUBSTATION AND SWITCHING STATION
RELIABILITY EVALUATION.
IEEE Transactions on Power Apparatus and Systems, Vol. 93 (1974), p. 1488-1497.
-142- References
Guertin, M.B. and Y. Lamarre
RELIABILITY ANALYSIS OF SUBSTATIONS WITH AUTOMATIC MODELLING OF
SWITCHING OPERATIONS
IEEE Transactions on Power Apparatus and Systems, Vol. 94 (1975), p. 1599-1605.
Gusciora, R.H.
BATHTUB CURVES, WEI BULL DISTRIBUTIONS AND RELAY RELIABILITY.
In: Proc. of the 36th Relay Conference. Stillwater. OK. April 1988.
Elkhart, IN: Assoc. Relay Manuf . 1988. Paper 8. p. 1 -7.
Heising. C.R.
RELIABILITY OF MEDIUM-VOLTAGE VACUUM POWER CIRCUIT-BREAKERS.
IEEE Transactions on Reliability. Vol. 32 (1983). p. 3-6.
Heising. C.R.
SUMMARY OF CIGRE 13-06 WORKING GROUP WORLD WIDE RELIABILITY DATA.
MAINTENANCE COSTS DATA AND STUDIES ON THE WORTH OF IMPROVED
RELIABILITY OF HIGH VOLTAGE CIRCUIT BREAKERS.
In: Industrial and Commercial Power Systems Technical Conference. Cleveland. Ohio.
5 - 8 May 1986. Ed. by C.E. becker et al. .
New York: IEEE. P. 93-101.
Heising, C.R. and R.C. Patterson
RELIABILITY EXPECTATION FOR PROTECTIVE RELAYS.
In: Fourth International Conference on Developments in Power System Protection.
Edinburgh U.K . April 1989.
London: Institution of Electrical Engineers. 1989. P. 23-26.
Heiss, W. and A. Schweer
RELIABILITY ANALYSIS FOR PLANNING AND OPERATION OF MEDIUM-VOLTAGE
NETWORKS.
In: Int. Conf. on Electricity Distribution (CIRED). Liege. Belgium. 1991.
Liege, Belgium: A.I.M . 1991. Paper 6.2. p.1-5.
Horton. W.F. and S. Goldberg. V. Backman
THE FAILURE RATES OF UNDERGROUND DISTRIBUTION SYSTEM COMPONENTS.
In: Proc. of the 22nd North American Power Symp . Auburn, AL. 15 - 16 October 1990.
Los Alamitos. CA: IEEE Computing Society. 1990. P.306-316.
Horton. W.F. and S. Goldberg. R.A.Hartwell (1989a)
A COST/BENEFIT ANALYSIS IN FEEDER RELIABILITY STUDIES.
In: Power Quality '89. Official Proc . Long Beach. CA. 15 - 20 October 1989.
Venture. CA: Intertec Commun . 1989. P. 552-558.
Horton. W.F. and S. Goldberg. K. McSweeney (1989b)
ACHIEVING DISTRIBUTION FEEDER RELIABILITY USING BENIFIT/COST RATIO
ANALYSIS.
In: Proc. 21th North-American Power Symp . Rolla. MO, 9 - 10 October 1989.
Washington. DC: IEEE Comput. Soc. 1989. P. 254-261.
-143- References
(IEEE, 1974)
REPORT ON RELIABILITY SURVEY OF INDUSTRIAL PLANTS, PART I: RELIABILITY OF
INDUSTRIAL PLANTS.
IEEE Committee report.
IEEE Power Systems Reliability Subcommitte (A.D. Patton, coordinating author). (IEEE,
1975)
RELIABILITY OF ELECTRIC UTILITY SUPPLIES TO INDUSTRIAL PLANTS.
In: Industrial and Commercial Power Systems Technical Conference, Toronto, Canada,
5 - 8 May 1975. Ed. by B.W. Whittington.
New York: IEEE, 1975. P. 131-133.
(IEEE Working Group) (IEEE, 1977)
IEEE GUIDE TO THE COLLECTION AND PRESENTATION OF ELECTRICAL, ELECTRONIC,
AND SENSING COMPONENT RELIABILITY DATA FOR NUCLEAR POWER GENERATING
STATIONS.
New York: IEEE, 1977.
IEEE Std. 500-1977.
(IEEE Working Group.) (IEEE, 1980)
IEEE RECOMMENDED PRACTICE FOR DESIGN OF RELIABLE INDUSTRIAL AND
COMMERCIAL POWER SYSTEMS (IEEE GOLD BOOK). Ed. by C.R. Heising
ANSIIIEEE Std. 493-1980.
SUMMARY OF REPLIES TO THE 1982 TECHNICAL QUESTIONAIRE ON ROTATING
MACHINERY (MOTORS 1000 HP AND UP).
Unpublished report cited by [IEEE, 1990). p.72.
IEEE GUIDE TO THE COLLECTION AND PRESENTATION OF ELECTRICAL, ELECTRONIC,
SENSING COMPONENT, AND MECHANICAL EQUIPMENT RELIABILITY DATA FOR
NUCLEAR-POWER GENERATING STATIONS.
New York.: IEEE, 1983.
IEEE Std. 500-1984.
IEEE RECOMMENDED PRACTICE FOR EMERGENCY AND STANDBY POWER SYSTEMS
FOR INDUSTRIAL AND COMMERCIAL APPLICATIONS (IEEE ORANGE BOOK).
ANSIIIEEE Std. 446-1987.
IEEE RECOMMENDED PRACTICE FOR THE DESIGN OF RELIABLE AND COMMERCIAL
POWER SYSTEMS (IEEE GOLD BOOK). Ed. by C.R. Heising
IEEE Std. 493-1990.
-144- References
Ishibashi, T. and A. Konemura, M. Sakurai, T. Aiyoshi
RELIABILITY OF VACUUM SWITCHING APPARATUS AND THEIR APPLICATION TO
MINING.
Mining Technology, vol. 72, no. 833 (March 1990), P. 75-79.
Kanai, Y. and T. Tsutsumi, N. Okabe
RELIABILITY OF VACUUM SWITCHING EQUIPMENT.
In: Probabilistic methods applied to electric power systems, London, 3 - 5 July 1991.
London: Institution of Electrical Engineers, Ed. by R.N. Allan et al ..
lEE Conference Proceedings 338. 1991.P.272-277.
Kawamura, T. and M. Horikoshi, S. Kabayashi, K. Hamamoto
PROGRESS OF SUBSTATION MAINTENANCE BASED ON RECORDS OF OPERATION
AND MAINTENANCE.
In: Int. Conf. on Large High Voltage Electric Systems (CIGRE), Paris, 26 August - 1
September 1990.
Kevers, J.
DE BETROUWBAARHEID VAN DE ENERGIE VOORZIENING TEN ZUIDEN VAN
EINDHOVEN IN DE HUIDIGE SITUATIE EN IN DE SITUATIE OAT HET 150 KV STATION
VELDHOVEN EN DE NIEUWE 150 KV VERBINDING GEREALISEERD ZIJN (The reliability
of the electric power supply south of Eindhoven in the current situation and in the
situation for which the new 150 kV substation will be in operation; in Dutch).
Den Bosch, The Netherlands: N.V. PNEM, 1987.
Internal report .. PNEM (Electricity supply company)
Kj0l1e, G. and K. Sand
RELRAD - AN ANALYTICAL APPROACH FOR DISTRIBUTION SYSTEM RELIABILITY
ASSESSMENT.
In: IEEE Power Engineering Society Transmission and Distribution Conf., September
1991, Dallas.
New York: IEEE, 1991. P.729-734.
Kloeppel, F.W. und G. Adler, W.W. Sorin, W.W. Tislenko
ZUVERLASSIGKEIT VON ELEKTROENERGIESYSTEMEN (Reliability of electric power
systems; in German).
Leipzig: Deutscher Verlag fUr Grundstoffindustrie, 1990.
Koglin, H.-J. and H.-J. Richter, W.H. Wellsson
COMPOSITE RELIABILITY TECHNIQUES IN EHV SYSTEMS IS THERE A NEED IN CASE
OF HIGHLY MESHED SYSTEMS!
1991.
Paris: CIGRE, 1991. Paper 1 A-02, p.1-6.
-145- References
Koval, D.
IMPACT OF DISTRIBUTION SYSTEM ISOLATION-RESTORATION PROCEDURES ON
INDUSTRIAL CONSUMER RELIABILITY LEVELS.
In: Industrial and Commercial Power Systems Technical Cont., Milwaukee, WI, 9 - 12
May 1983.
New York: IEEE, 1983. P.71-76.
Kula, M. and H. Rohster, V. Vachek
COORDINATION OF SUBSTATION MAIN CIRCUITS AND PROTECTION FORM THE VIEW
POINT OF THEIR RELIABILITY.
In: Electric power systems reliability Symp., Montreal, Canada, 16 - 18 September 1991.
Lau, P.C.K. and M.S. Grover, W.H. Tanaka
RELIABILITY ASSESSMENT OF SPECIAL PROTECTION SYSTEMS.
In: Electric power systems reliability Symp., Montreal, Canada, 16 - 18 September 1991.
Paris: CIGRE, 1991. Paper 3A-11, p.1-6.
Lee, N.
HOW RELIABLE IS YOUR UPS?
Australian, Asian and Pacific Electrical World, Vol. 56 (1991), no. 2, p.14-15.
Londsdale, J.G. and G.B. Hitchen
RELIABILITY EVALUATION IN THE PLANNING OF DISTRIBUTION SYSTEMS.
Eggleton et al..
Magnon, B. and J. L'Henoret, Y. Reyraud
RELIABILITY OF STAND-BY DIESEL GENERATOR SETS IN NUCLEAR PLANTS.
Eggleton et al..
Mazza, G. and R. Michaca
THE FIRST INTERNATIONAL ENQUIRY ON CIRCUIT BREAKER FAILURES AND DEFECTS
IN SERVICE.
Mcnab, W.A.
URBAN ELECTRICITY SUPPLY SYSTEMS.
In: Reliability of Power Supply Systems, London, 21 -23 February 1977. Ed. by M.N.
Eggleton et al..
-146- References
Mohan Rao, N.S. and P.V.S. Sekhar
COMPARISON OF RELIABILITY OF ALTERNATE SYSTEMS OF DISTRIBUTION.
In: VI NPSC 90. Power Systems for the Year 2000 and Beyond. Proc. of the 6th Nat.
Power Systems Conf., Bombay, 4 - 7 June 1990.
New Delhi: Tata McGraw Hill, 1990. P.573-580.
Morrison, W.G. and R.J. Arhart
USING NEW TECHNOLOGY TO IMPROVE RELIABILITY OF AN INDUSTRIAL CABLE
DISTRIBUTION SYSTEM.
In: Annual Pulp and Paper Industry Technical Conf., Montreal, Canada, 3 - 7 June 1991.
New York: IEEE, 1991. P. 93-99.
Morzelle, M. and M. Donnelly, M. Kauffmann
THE RELIABILITY OF EMERGENCY GENERATOR SETS. STUDY, EXPERIMENTATION,
OPERATIONAL EXPERIENCE.
In: Int. Conf. on Operability of Nuclear Systems in Normal and Adverse Environments
(OPERA), Lyon, France, 18 - 22 September 1989.
Paris: Soc. Franc;;aise d'Energie Nucl., 1989. P. 393-403.
Nelson, J.A. and E.F. Johnson
POWER SUPPLY TO GAS CENTRIFUGE URANIUM ENRICHMENT PLANT.
Norris, A.T.
REPORT OF CIRCUIT BREAKER RELIABILITY SURVEY OF INDUSTRIAL AND
COMMERCIAL INSTALLATIONS.
In: Industrial and Commercial Power Systems Technical Conference, Chicago, Illinois, 7-
11 May 1989.
New York: IEEE, 1989. P. 1-16.
O'Donnell, P.
REPORT OF SWITCHGEAR BUS RELIABILITY SURVEY OF INDUSTRIAL PLANTS AND
IEEE Transactions on Industry Applications, Vol. 15 (1979), p. 141-147.
O'Donnell, P.
REPORT OF GENERATOR RELIABILITY SURVEY OF INDUSTRIAL PLANTS AND
In: Industrial and Commercial Power Systems Technical Conf., Houston, Texas, 12 - 15
May 1980.
New York: IEEE, 1980. P. 40-44.
O'Donnell, P. and R.N. Bell, C. Heising, C. Singh, S.J. Wells
REPORT OF LARGE MOTOR RELIABILITY SURVEY OF INDUSTRIAL AND COMMERCIAL
INSTALLATIONS.
In: Industrial and Commercial Power System Technical Conf., Atlanta, GA, 7 - 10 May
1984.
New York: IEEE, 1984. P. 176-185.
-147- References
O'Donell, P. and R.N. Bell, D.W. Mc Williams, C. Singh, S.J. Wells
REPORT OF LARGE MOTOR RELIABILITY SURVEY OF INDUSTRIAL AND COMMERCIAL
INSTALLATIONS.
In: Industrial and Commercial Power System Technical Conference, Milwauki, WI, 9 - 12
May 1983.
New York: IEEE, 1983: P. 58-70.
Parascos, E.T. and J.A. Arceri
RELIABILITY ENGINEERING AND UNDERGROUND EQUIPMENT FAILURE, COST AND
MANUFACTURER'S ANALYSIS.
In: Reliability of Power Apparatus and Systems, 23 -24 September 1976, Montreal,
Canada. Ed. by M.F. Hashmil.
Patten, A.D.
DUTY CYCLE EFFECTS ON GENERATING UNIT AVAILABILITY.
In: IEEE Industry Applications Society Annual Meeting 1992.
Patterson, R.C. and J.E. Teague
FIELD EXPERIENCE - SOLID-STATE PROTECTIVE RELAYS.
In: Proceedings of the American Power Conf., Vol. 36., Chicago, 29 April- 1 May 1974.
Chicago: Illinois Institute of Technology, 1974. P. 1074-1079.
Pijls, R.M.M.
BACK-UP BEVEILIGING IN MIDDENSPANNINGSNETTEN (Back-up protection in medium
voltage networks; in Dutch).
Heerlen, The Netherlands: HTS (Technical College) Heerlen, april 1988.
Bachelor's thesis.
Prescott, W. and A.J. Rapson, C.D. Schofield
PROBABILISTIC EVALUATION OF POWER SUPPLIES FOR LARGE COMPUTER AND
COMMUNICATIONS INSTALLATIONS.
London: Institution of Electrical Engineers, Ed. by R.N. Allen at al..
lEE Conference Proceedings no. 338.1 991.P.224-228.
Radwan, R.M. and M. EI-Marsafawy
FAILURE ANALYSIS OF CIRCUIT BREAKERS IN THE UPPER EGYPT ZONE OF THE
EGYPTIAN UNIFIED POWER SYSTEM NETWORK.
In: Third Int. Conf. on Future Trends in Distribution Switchgear, London, UK, 23 - 25
April 1990. Ed. by R.N. Allan, et al ..
London: Institution of Electric Engineers, 1990. P. 79-83.
Roos, E.
VERZORGUNGSZUVERLASSIGKEIT VON MITTELSPANNINGSNETZEN DER
OFFENLICHEN ENERGIE-VERSORGUNG (Supply reliability of medium voltage networks
in the public energy supply; in German).
Elektricitiitswirtschaft, Vol.88 (1989), p.316-329
-148- References
Ruoff, J.T. and H.P. van Meeteren
BETROUWBAARHEIDSANAL YSE VAN DISTRIBUTIESYSTEMEN: EEN EENVOUDIGE EN
EEN GEAVANCEERDE METHODE (Reliability analysis of distribution systems: a simple
and an advanced method; in Dutch).
Elektrotechniek, vol. 61 (1983), p. 770-779.
Sallam, A.A and M. Desouky, H. Desouky
EVALUATION OF OPTIMAL-RELIABILITY INDICES FOR ELECTRICAL DISTRIBUTION
SYSTEMS.
IEEE Transactions on Reliability, Vol. 39 (1990) p. 259-264.
Schilling, M.Th. and J.C.G. Praca, J.F. de Queiroz, C. Singh, H. Ascher
CHARACTERIZATION OF THE AGING OF THERMAL GENERATORS FOR RELIABILITY
ASSESSMENT.
In: Probabilistic Methods Applied to Electric Power Systems, Toronto, 11 - 13 July
1986.
Oxford, U.K.: Pergamon, 1987. P. 185-193.
Schilling, M.T. and J.C.G. Praca, J.F. de Queiroz, C. Singh, H. Ascher
DETECTION OF AGEING IN THE RELIABILITY ANALYSIS OF THERMAL GENERATIONS.
IEEE Transactions on Power Systems, Vol. 3 (1988), p. 490-499.
Schneider, F.
RELIABILITY OF UNINTERRUPTIBLE POWER SUPPLY SYSTEMS.
In: Int. Telecommunications Energy Conf. (lNTELEC), Florence, Italy,15 - 18 October
1989
New York: IEEE, 1989. Paper 14.4, p.1-7.
Selseth, N.A.
RELIABILITY OF CIRCUIT-BREAKERS. OUTAGE DATA EVALUATION.
In: Int. Conf. on Large High Voltage Electric Systems (CIGRE), Paris, France, August
1972.
Paris: CIGRE, 1973. Paper 13-04.
Simpson, W.C. and H.G. Stoll
INFLUENCE OF UPGRADING AND MAINTENANCE ON GENERATING UNIT
AVAILABILITY PERFORMANCE.
In: Proc. of the American Power Conf., Chicago, IL, 24 - 25 April 1989.
Chicago, IL: American Power Conf., 1989. P. 947-955.
Smith, C. and H. Donovan, H.J. Bartos
RELIABILITY SURVEY OF 600 TO 1800 kW DIESEL AND GAS-TURBINE GENERATING
UNITS.
IEEE Transactions on Industry Applications, Vol. 26 (1990), p. 741-755.
-149- References
Smith, M.J.
RELIABILITY, AVAILABILITY, AND MAINTAINABILITY OF UTILITY AND INDUSTRIAL
COGENERATION PLANTS.
In: IEEE Industry Applications Society Annual Meeting, San Diego, CA, 1 - 5 October
1989.
Piscatataway, NJ: IEEE, 1989. P. 1783-1787.
Snaith, E.R.
RELIABILITY EVALUATION OF AN ELECTRICAL SUPPLY SYSTEM FOR A NUCLEAR
POWER STATION.
In: Reliability of Power Supply Systems, London, 21 - 23 February 1977.
Stanek, E.K. and S.S. Venkata
MINE POWER SYSTEM RELIABILITY.
IEEE Transactions on Industry Applications, vol. 24 (1988), p. 827-838.
Suntio, T. and J. Uusitalo, l. Jonsson
AC-UPS RELIABILITY AND AVAILABILIITY PERFORMANCE: COMPARISON OF
AVAILABLE SOLUTIONS.
In: Int. Telecommunications Energy Conf. (INTELEC), Florence, Italy, 15 - 18 October
1989.
New York: IEEE, 1989. Paper 14.1, p.1-10
Suntio, T. and K. White
RELIABILITY AND AVAILABILITY PERFORMANCE AS A TOOL TO OPTIMIZE POWER.
In: Int. Telecommunications Energy Conf. (INTELECI. San Diego, CA 30 October - 2
November 1988.
New York: IEEE, 1988. P. 49-56
Todd, Z.G.
A PROBABILITY METHOD FOR TRANSMISSION AND DISTRIBUTION OUTAGE
CALCULATIONS.
Ugokwe, E.C.
OFFLINE 1774 PLC RELIABILITY CASE STUDY.
van Amelsfoort, B. and J. Kevers, F. van Loon
BETROUWBAARHEIDSANALYSE VAN HET 10 kV-DISTRIBUTIENET IN RELATIE TOT
HET AANTAL VERMOGENSSCHAKELAARS EN DE PLAATSING HIERVAN IN HET NET
(Reliability analysis of the 10 kV distribution network in relation to the number of circuit
breakers and their position in the network; in Dutch).
's Hertogenbosch, Netherlands: PNEM, 1986.
Internal report. PNEM. (Electricity supply company)
-150- References
van der Vaart, P. and K. Bouwknegt
SOME ASPECTS OF THE IMPROVEMENT OF THE MTBF OF THE POWER SUPPLIED TO
COMPUTERS.
In: Int. Conf. on Electricity Distribution (CIRED), London, 8 - 12 May 1989.
London: Institution of Electrical Engineers, 1989: P. 402-405.
lEE Conference publication 305.
VDEN
RAPPORTERING VAN DE NIET-BESCHIKBAARHEID T.G. V. STORINGEN IN NETTEN VAN
0,4 kV ... 30 kV, ANALYSE 1987 (Report on the non-availability due to failures in 0.4 ..
30 kV networks, analysis 1987; in Dutch).
Arnhem, Netherlands: VDEN, 1988.
VDEN Report nr. 00043-VB 88-850.
Verplanke, F.W.R.
BETROUWBAARHEIDSONDERZOEK (Reliability research; in Dutch).
Vlissingen, Netherlands: Hogeschool (College) Zeeland, 1991.
Bachelor thesis.
Vesely, W.E.
AGING EFFECTS IN COMPONENT FAILURE AND DOWNTIME DATA AND IMPACTS ON
PLANT RISK.
In: Proc. 17th Water Reactor Safety Information Meeting, Rochville, MD, October 1989.
Washington D.C.: Nuclear Regulatory Commission, 1990. P. 473-496.
NUREG/CP-0105.
Vlutters. H.
VINGER AAN DE POLS VOOR VEILIGHEID (finger on the pulse for safety; in Dutch).
Stroomlijn, Vol. 12 (1991). no.3, p. 3-9.
Volkman, C.A. and S. Goldberg, W.F. Norton
A PROBABILISTIC APPROACH TO DISTRIBUTION SYSTEM RELIABILITY ASSESSMENT.
London: Institution of Electrical Engineers, 1991. P. 169-173.
lEE Conference Proceedings 338.
Warren, J.C.
INCREASING UPS AND COMPUTER RELIABILITY IN LARGE INSTALLATIONS.
Piscateway, NJ: IEEE, '1992. P.1981-1985.
Waumans, R.J.R.
OPENBARE NETTEN VOOR ELEKTRICITEITSDISTRIBUTIE (Public networks for electricity
distribution; in Dutch).
Deventer, The Netherlands: Kluwer, 1986.
Also: Arnhem, The Netherlands: VDEN, 1986.
-151- References
Wahlstrom, B. and W. Lord, K. Petterson, R. Alvinsson, M.A.S. Leijon
EXPERIENCES AND CONSIDERATIONS REGARDING ERECTION TESTING AND
OPERATION AVAILABLE OF GIS IN SWEDEN.
In: Int. Conf. on large High Voltage Electric Systems (CIGREI, Paris, 28 August - 3
September 1988.
CIGRE. Paris, 1988. Paper 23.07, p.1-8.
Whiting, J.P.
RELIABILITY OF POWER SUPPLIES AND SYSTEMS AT BROADCAST TRANSMITIING
STATIONS.
in: Int. Conf. on Electricity Distribution (CIREDI, Brighton, UK, 8 - 12 May 1989.
Williams, W.P. and S.G. Mudge
RELIABILITY ASSESSMENT FOR INDUSTRIAL POWER NETWORKS.
In: Int. Conf. on Reliability of Power Supply Systems, 19 - 21 September 1983, London.
Ed. by H. Hammersley.
Wiseman, N.E.
RELIABILITY TESTING.
In: Reliability of Power Apparatus and Systems, 23 - 24 September 1976, Montreal,
Canada. Ed. by M.F. Hashmi.
Yaguchi, T. and Y. Oura, A. Tsuboi, F. Andow
IN-SERVICE EXPERIENCE AND RELIABILITY EVALUATION OF PROTECTIVE RELAY
SYSTEMS WITH BUILT-INN AUTOMATIC TESTING AND SUPERVISING DEVICES.
In: Int. Conf. on Large High Voltage Electric Systems (CIGREI, August 1984, Paris,
France.
Yotsumoto, K. and S. Muroyama, H. Watanabe
DESIGN CONSIDERATIONS OF A DISTRIBUTED POWER SUPPLY FOR
TELECOMMUNICATIONS NETWORK USE BASED ON RELIABILITY ANALYSIS.
Transactions of the Institute for Electronics, Information, and Communication Engineers,
Part E, Vol. 73 (1990), p. 395-400.
-152- References
iOin<:ihoven Universitv of Technology Research Reports ISSN 0167-9708
[acul!.L of El8ctrica I Enqineerinq
12511 EIJndhovell. J.T.J. van and G.6. de Jonq. L. Stok
THE ASCIS DATA FLOW GRAPH: textual format.
EUT Report 91-E-251. 1991. ISBN 90-614H5H
(2521 Chen. J. ,nd PJ.I de Maag!. M.H.U. Herben
Coden: TElTEDE
WIDE-ANGLE RADIATION PATTERN mCULATION OF PARABOLOIDAL REFLECTOR ANTENNAS A compmtm
study.
EUT Report 91-E-252. 1991. ISBN 90-6144-252-4
(25J) Haan. S.W.H. de
A PWM CURRENT -SOURCE INVERTER FOR INTERCONNECTION BETWEEN A PHOTOVOLTAIC ARRAY AND THE
UTILiTY
EIlT Report 9H-253. 1991. ISBN 90-6144-253-2
(254) Velde. M. van de and U.K. Clultmans
EEG ANALYSIS FOR MONITORING OF ANESTHETIC DEPTH.
EUT Report 9H-254. 1991. ISBN 90-6144-254-0
(255) Smolders, 1.8,
AN EFFICIENT METHOD FOR ANALYZING MICROSTRIP ANTENNAS WITH A DIELECTRIC COVER USING A
SPECTRAL DOMAIN MOMENT METHOD.
EUT Renort 9\-E-255. 1991. iSBN 90-6144-255-9
(256) Backx, A,e,p,M. and A.A.H. Damen
IDENTIFICATION FOR THE CONTROL Of MIMO INDUSTRIAL PROCESSES,
EUT Report 9H-256. 1991. ISBN 90-6144-25(,-7
1257) Maaqt, P J! 1\, and H,G. ter Morsche, J.L.M. van den Broek
A SPXmL RECONSTRUCTION TECHNIQUE APPl.ICABLE TO MICROWAVE RADIOMETRY
EUT Report 92-E-257. 1992. ISBN 90-6144-257-5
!25GI Vieeshoums. J.M
0ERIV!TION OF A MODEL OF THE EXCITER OF A BRUSH LESS SYNCHRONOUS MACHINE.
EUT Report 92--250 1992 ISBN 90-6144-258-3
(259) Orlov, V.B.
DEfECT MOTION AS THE ORIGIN Of THE IIF CONDUCTlNCF. NOISE iN SOLIDS.
WT Report 92-E-259. 1992. ISBN 90-6144-259-1
(2601 RODI!"kers. J.E.
ALGDRITIIMS FOR SPEECH CODING SYSTEMS BASED ON LINEAR PREDICTION.
EUT Report 92-E-260. 1992. ISRN 90-6144-260-5
(261) Boom, LJ.J. van den and A.A.H D,men, Martin Klompstra
IDENTIFICATION FOR ROBUST CONTROL USING AN H-infinlty NORM.
EUT Report 91-E-261. 1992. ISBN 90-6144-261-)
12621 Grot,n. M and W. van Etten
LASER LINEWIDTH MEASUREMENT IN THE PRESENCE OF RIN AND USING THE RECIRCULATING SELF
HETERODYNE METHOD.
WT Report 92--262. 1992. ISBN 90-6l44-262-1
(2631 Smoiders. t
RIGOROliS ANALYSIS Of THICK MICROSTRIP ANTENNAS AND WIRE ANTENNAS EMBEDDED IN A ,lISmATE
EUT Heport 92-E-263. 1991. Isr,N 90-6144-161-j
EHQhoven UniversitY of R-es8arch_R-eoorts 01fi7--9708
Coden, TEl JEDE
!264\ Frenks, MId U M,j von Gils
THE ADAPTIVE RESONINCE THEORY NETWORK IClusterillg-1 behdVIour In re'dilon with hrmst,,,
audltory evoked D"tent"l pdtterllS,
WT Report n-E-164, 1992. ISBN 90-6144-1648
1165i Wellen, J,S. and F. M.r.e. Schem"ann, E. Sm,l\lruqqe, L N,F. KeulIDdfln
MANUFACTURING AND CHARACTERIZATION OF GA!5mGlAS MULTIPLE QUINTliMWELl. RIDGE mEGUIDE
LASERS
WT Report 92+20j 1m. ISBN 90-6\44-165-6
12661 Cluitruans, U.M
USING GENETIC ALGORITHMS fOR SCHEDULING DATA FLOW GRAPHS.
Wi Report n+M \99, 9H14H66-'l
12671 J6zmk, L, dnd A,P.H. van DiJk
12681
12691
mOl
i27l!
( 273)
( 274J
! METHOD FOR GENERAL SIMULTANEOUS FULL DECOMPOSITION Of SEQUENTIAL MACHINES:
Algorithms and lmplement,tlon
EUT Report 9H-267 1992. ISBN 90-6144-267-2
Boom. H. Vdo den and W VdO Etten, W,H.C, de Krom. P VdO Benneko . f. HUl)skens,
L. NJessen, F de =
AN OPTICAL ASK AND fSK PHASE DIVERSITY TRANSMISSION SYSTEM,
RUT Report 92+266, 1992. ISBN 90-6144-168-0
Putten, P H,i, Vdn der
MUlTlOISCIPLlNAIR SPEClFICEREN EN ONTWERPEN VAN MICROELEKTRONICA IN PROaUKTEN 110 Dutch I
EUT Report 1993. ISBN 90-6144-269-9
StnKs, R,HJ
PROGRIL, A lancru,o, for the defloJtion of protocol grarurudrs.
WT Report 93+270. 1993 ISBN 90-6144-270-2
Bloks, R,H,J,
CODE GENERATION FOR THE ATTRIBUTE EVALUATOR OF THE PROTOCOL ENGINE GRAMMAR PROCESSOR UNIT.
WT Report 93+271. 1993 ISBN 90-6144- 27 HI
1,n, KeplnQ dOli V'II Vcldhumn
flUE GAS clEANING BY PULSE CORONA STREAMER
fUT ReDort 9J-E-2J2 1993, ISBN 90-6144-2'12 9
Smolders, 11,.
FINlTE STACKED MICROSTRiP !RRAYS WITH THICK SUBSTR!TES.
EUT Report 93-E-273 1993. ISBN 90-6144-273-7
Bollen, M,H.J, and M I van Houten
ON INSULAR POWER SYSTEMS DraWlng up an inventory of phenomena dOd research POSSibIlities.
EUT Report 9J-f,-274. 1993. ISBN 90-6144-274-5
i27SI O(,I1[:;,n I.P.J van
rIl:CTROMIGNETlC COMPATTBILITY' Part 5, lust,ll,ilon and mitIgatIOn '!llldelJOes, scellOn 3,
,a\llilla dnd w'rllO
EliT Report 9H-175. 1993, iSBN 90-614075-3
12761 Bollen. M.H .1.
LITERATURE SEARCH fOR RELIABILITY om OF COMPONENTS IN ELECTRIC OISTPI8IITlON NFTWORKS,
EUT Report 93+276, 1993, ISBN 90-6144-276-1

Literature Search For Reliability Data of Components in Electric Distribution Networks

Uploaded by

Copyright:

Available Formats

Literature Search For Reliability Data of Components in Electric Distribution Networks

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Literature Search For Reliability Data of Components in Electric Distribution Networks

Uploaded by

Copyright:

Available Formats

Literature Search for Reliability

Data of Components in Electric

You might also like