| |
|
The program eaFIRM is a system
for investigating the effect of the number of
fault response teams on the performance statistics
of an electricity distribution company. It also
provides a flexible way of evaluating the different
strategies that can be used for team deployment.
It operates by re-playing a set of fault records,
but makes changes to number of repair teams, or
to the strategy for their deployment. It answers
questions like:
- If the company was organised
differently could it have dealt better with
the faults that occurred last year?
- What would it cost, and how
much would the percentage of restorations within
3 hours improve?
|
Main Screen of eaFIRM
|
|
eaFIRM - Fault Incident Resource
Model
Logistic and resource modeling
Faults on the HV overhead network
are of particular importance as they occur very
unevenly through the year and can affect large
numbers of customers. This means that large numbers
of staff have to be available for emergencies
which may occur only two or three times a year.
The cause is usually severe weather which can
be lightning storms in the summer, high winds
in the autumn, or snow and ice in the winter.
In contrast, the underground network damage or
failures occur more evenly throughout the year,
which makes the response much easier to plan.
The appropriate number of staff to keep available
for dealing
with underground faults can be readily determined,
but for
|
The main eaFIRM screen |
overhead faults the situation is
much more difficult. On many days there may be only five
overhead faults or less, but there may be several days
per year with 150 or 200 faults in the day.
Changing the number of overhead linesmen
affects the rate at which overhead line faults can be
repaired. When there are many faults on the same day
the faults have to be queued up to wait for the teams
to become free. The Regulator creates pressure to improve
the fault repair statistics, so as to reach agreed targets
for the total customer minutes lost and the percentage
of interruptions restored in 3 hours or 24 hours. Using
fewer teams saves money but allows the performance to
worsen. Having more teams on duty improves the performance,
but on many days there may not be enough faults to occupy
all the teams.
|
The Fault Incident Resource
Model is a computer model for analysing the staffing
of the response to faults on the HV overhead network.
It uses the historic fault records from previous
years and replays them with changes to represent
alternative numbers of linesman teams combined
with alternative strategies for their deployment.
It also covers the effect of network investments
which aim to change the incidence of faults or
to change the time taken to respond to those faults.
It answers questions such as
"If the company was organised differently,
could it have dealt better with the faults which
occurred last year?" and "What would
it cost, and how much benefit would be gained
to increase the number of teams available for
faults?" Replaying the historic fault records
can show the influence of the staffing responses
to the performance statistics, but in its basic
form it assumes that
|
Staff costing sheet |
the faults occur at the same rate
as in the past. Some network investments in improved equipment
are designed to reduce the incidence of faults, and for
these cases the historic fault records can be adjusted
to reduce the number of faults.
There are other initiatives
that aim to reduce the time to travel to a fault, the
time to locate a fault, the time to perform switching
operations, or the time taken to repair a fault. In
such cases the historic fault records can be adjusted
to reduce the time taken for the appropriate stages
of restoration or repair. The result will be an estimate
of the performance that would have been achieved if
the same faults had occurred after the investment had
been made. Examples of suitable scenarios to analyse
in this way would be the installation of fault passage
indicators, and remote control operation of switches.
|
A database with all fault information
is used as the source information for replaying
the defined period of faults. The time and date
of each fault incident remain unchanged, only
the times teams are assigned to it and subsequent
stage restoration times.
To model the effect of
the number of teams on performance it is important
to distinguish between the different fault types,
this is carried out and all faults in the database
are split into the following categories, Normal,
remotely restored non damage , on site interruption
and invalid (where data is missing in the record)..
Once faults start to accumulate in a queue a strategy
is needed to decide which fault to take from the
queue first, these strategies can be replayed
through eaFIRM to establish which would result
in the best performance figures. Strategies
|
The staff scheduling screen
|
for emergencies and priority restoration
before repair can alsobe replayed by amending the number
of lines teams available and their responsibilities. Different
authorisation levels for teams and the costs associated
with the different teams can also be analysed.
|
The results from eaFIRM provide
comparison information between original data and
the simulation data, and includes customer interruptions,
customer minutes lost, and restoration times.
Other information includes average queue time
and costs incurred with the option selected.
|
The eaFIRM results file
|
|
|
