APPENDIX 31
Further Memorandum submitted by Northern
Ireland Electricity plc
Thank you for your letter of 24 May 1999 in
which you requested further information relating to:
— customers experiencing multiple interruptions
of supply;
— how circuits are prioritised for
refurbishment; and
— network performance comparisons with
other companies.
We reply as follows:
1. CUSTOMERS EXPERIENCING
MULTIPLE INTERRUPTIONS
OF SUPPLY
As regards customers experiencing multiple interruptions
of supply you requested information on the extent to which records
are kept and a statistical analysis of the frequency of interruptions.
1.1 Records
As described in our earlier submission to the
Committee, the 11 kV overhead network is principally comprised
of a number of arterial lines with normally open points running
between adjacent 33/11 kV substations together with multiple subsidiary
radial spur lines. Each arterial line is controlled by a circuit
breaker which operates automatically in response to network faults,
thereby limiting an interruption of supply to those customers
supplied via that circuit breaker.
Spur lines are generally controlled and protected
by fuses. Spur line faults do not affect the integrity of the
mainline and therefore interruption of supply is limited to those
customers connected to the faulted spur.
Where part of the network is required to be
taken out of service to facilitate maintenance work a planned
outage is arranged which requires the manual operation of circuit
breakers, switches or fuses. Where possible, the network is normally
reconfigured manually to provide an alternative source of supply
to as many customers as possible and thus minimise the number
of customers affected by the outage. The lack of a re-supply facility
on radial spurs means that there is generally little scope to
minimise supply interruptions during planned outages on those
parts of the network.
The National Fault and Interruption Reporting
Scheme (NAFIRS) is used to record historical data relating to
supply interruptions on the transmission and distribution networks.
The NAFIRS database is structured to record outages from a network
perspective and facilitates assessment of network performance
on either a circuit by circuit basis or by a comparison of the
performance of the NIE network against those GB RECs also subscribing
to NAFIRS. Supply interruptions are categorised as either fault
related" or planned outages".
NAFIRS records the number of customer interruptions
associated with each network incident, relates each incident to
the associated network device and also records the cause of the
interruption. Each network device is linked within the database
hierarchy to the appropriate circuit, the supplying substation
and district. This allows flexible and detailed analysis of network
performance.
However, NAFIRS does not record outages
against the specific customers affected and accordingly cannot
be used to provide a definitive analysis of the frequency of supply
interruptions against specific customers. Such an arrangement
would require the tracing of the electrical connectivity of each
customer through the complete transmission and distribution network
and the summation of all incidents recorded against each interruption
device. In addition, a strict account would also need to be taken
of all occasions when the network configuration was altered from
its normal running state.
While this constraint makes the analysis requested
in your letter difficult, it is possible to provide some indication
of the likely range of interruptions experienced by customers
by limiting the analysis to the 11 kV network only. The omission
of the relatively small number of interruptions contributed by
the low voltage, 33 kV and the transmission networks likely to
be experienced by the same customers should not invalidate the
general conclusions derived.
1.2 Analysis of fault-related supply interruptions
NAFIRS can provide information relating to the
frequency of faults cleared by circuit breakers and fuses and
the associated number of customer interruptions. Table 1 below
provides an analysis for the 11 kV urban and rural networks. The
figures represent annual performance averaged over the three years
from 1995-96 and the two sets of data are not necessarily mutually
exclusive.
The table shows, for example, that 292 HV circuit
breakers and 132 HV fuses experienced an annual average of one
fault over the three year period.
TABLE 1
Average annual frequency of faults cleared
by circuit breakers and fuses and the associated number of customers
affected
| No. of faults | Fault clearing device
| 11 kV Circuit Breaker | 11 kV Fuses
|
| Number | Customers
| Number | Customers |
| 1 | 292 | 184,336
| | Data |
| 2 | 76 | 45,446
| | not |
| 3 | 20 | 10,405
| | readily |
| 4 | 6 | 4,145
| | available |
| 5 | 6 | 3,164
| 5 | 244 |
| 6 | 1 | 851
| 1 | 81 |
| 7 | 1 | 350
| 1 | 110 |
| 8 | 1 | 121
| 1 | 72 |
| 9 | 1 | 108
| 0 | 0 |
| 10 | 0 | 0
| 1 | 30 |
| 11 | 0 | 0
| 0 | 0 |
| 12 | 0 | 0
| 1 | 13 |
| 13 | 0 | 0
| 1 | 41 |
NAFIRS reports customer interruptions" rather than the
actual number of customers affected. In some fault situations
where manual switching is involved two or more customer interruptions
are reported although only one customer was involved. The above
analysis derives the number of customers affected from the number
of customer interruptions reported in NAFIRS. To account for double
counting" within customer interruptions, the estimate of
customer numbers is based on 75 per cent of the customer interruptions
reported. This is considered to represent a reasonable indication
of the number of customers involved.
Although the two customer groups are not mutually exclusive
the analysis indicates that of those customers who were subject
to one or more faults around 90 per cent experienced no more than
two interruptions per annum. In respect of a "worst case"
situation, less than one hundred customers experienced in excess
of 10 faults per annum.
It is not possible to provide a similar breakdown for planned
interruptions. As for faults, planned outages are recorded in
NAFIRS against the relevant circuit. Unlike a main line fault
which causes a source circuit breaker to open affecting all customers
on that circuit, supply interruptions arising from planned outages
are normally restricted to the customers connected to the line
section on which work is being carried out. This is achieved through
the manual reconfiguration of the circuit prior to the commencement
of the planned work. Also, increasingly, mobile generators and
live-line working are used to avoid any customer outages during
programmed work. Therefore, although the number of customer interruptions
is recorded in NAFIRS there is no correlation between the planned
outages of a particular circuit and the number of supply interruptions
experienced by individual customers.
2. CRITERIA FOR
ASSESSING REFURBISHMENT
PRIORITIES
Circuits are ranked according to a composite performance
index which takes account of the three main performance indices
of each circuit:
— Reliability: annual faults (per 100 km) on each
circuit.
— Security: annual customer interruptions (per 100
customers) on each circuit.
— Availability: annual customer minutes lost (per
connected customer) on each circuit.
The composite index gives the greatest weighting to circuit
availability (customer minutes lost) and therefore emphasises
the number of customers affected on the circuit, whilst also taking
account of the duration of outages and the number of faults.
The ranking of circuits by this method provides a priority
list (the rogue circuit" list) of circuits to be refurbished.
The priority list may be subsequently modified to take account
of local knowledge of other relevant aspects which is not immediately
apparent from the statistical analysis.
3. NETWORK PERFORMANCE
COMPARISONS WITH
OTHER COMPANIES
We can confirm that the performance figures for NIE's network
in the table attached to your letter are correct. As requested
we have provided further information which facilitates comparisons
with other companies on the basis of fault outages and pre-arranged
outages. (See Table 2 below.)
The fault outage CML figures for NIE are given on two basis—inclusive
and exclusive of the effect of exceptional circumstances. We do
not have sufficient data to allow us to separate out the exceptional
faults for the GB companies. We have included data for Scottish
Hydro Electric and we have added a column showing average annual
performance over the seven year period.
TABLE 2
Availability of supply: Supply minutes lost per connected
customer per annum
| 1991-92 | 1992-93
| 1993-94 | 1994-95 | 1995-96
| 1996-97 | 1997-98 | Average
|
| NIE | |
| | | |
| | |
| Fault | 149 | 177
| 184 | 131 | 145
| 122 | 118 | 146
|
| PA | 104 | 91 |
79 | 82 | 95 | 90
| 90 | 90 |
| Exceptional Faults | |
| 1181 | |
| | 1112 |
|
| Total A | 253 |
268 | 381 | 213
| 240 | 212 |
319 | 269 |
| Total B (excluding exceptionals) | 253
| 268 | 263 | 213
| 240 | 212 | 208
| 237 |
| Hydro-Electric | |
| | | |
| | |
| Fault | 200 | 288
| 181 | 161 | 293
| 124 | 152 | 200
|
| PA | 70 | 68 |
73 | 72 | 72 | 82
| 67 | 72 |
| Total A | 270 |
356 | 254 | 233
| 365 | 206 |
219 | 272 |
| SWALEC | | |
| | |
| | |
| Fault | 133 | 134
| 156 | 160 | 189
| 137 | 132 | 149
|
| PA | 192 | 78 |
44 | 52 | 44 | 52
| 51 | 73 |
| Total A | 325 |
212 | 200 | 212
| 233 | 189 |
183 | 222 |
| SWEB | |
| | | |
| | |
| Fault | 101 | 120
| 113 | 94 | 85
| 75 | 93 | 97 |
| PA | 75 | 64 |
54 | 39 | 26 | 28
| 15 | 43 |
| Total A | 176 |
184 | 167 | 133
| 111 | 103 |
108 | 140 |
| MANWEB | |
| | | |
| | |
| Fault | 62 | 83
| 78 | 66 | 64 |
62 | 84 | 71 |
| PA | 46 | 46 |
43 | 36 | 24 | 16
| 13 | 32 |
| Total A | 108 |
129 | 121 | 102
| 88 | 78 | 97
| 103 |
Exceptional Faults
1. 118 CML due to major loss of generation (3 February 1994).
2. 61 CML due to transmission system capacity restrictions
following a transformer fire (4 November 1997);
32 CML due to storm damage to the distribution system (24 December
1997);
18 CML due to loss of generation capacity (13 June 1997).
In making comparisons please note that NIE's exceptional
events include transmission incidents and major generation losses,
whereas the figures for Manweb, Sweb and Swalec do not include
transmission incidents, and the potential for a loss of generation
to result in customer interruptions is much lower on the GB system.
It is also worthwhile noting the increase in NIE's prearranged
CMLs since 1994-95. This reflects the extent of the very significant
ongoing overhead line refurbishment programme that has been in
place since that time.
4 June 1999
| 
