Electrical distribution system fault location-How is it done?

[djsb]

After been without power for over 10 hours I am curious if we have anyone on the forum with experience of working on the UK Electical distribution network?
I learnt that the initial problem presenting itself was with the local substation. It kept tripping out and that was confirmed by my lights staying on for around a minute and going off again. Then the problem was described as being with an underground cable that had maybe been damaged due to heavy rainfall. I learnt that houses are fed by a three phase system, meaning that one phase feeds every other house etc and that blocks of flats where I live have all 3 phases fed to them but the power is still split. I'm just curious how faults are traced in the underground wiring. Is there some kind of sensor network or a computerised system or is the method more crude than this. Does anyone have any experience in this field? Thanks.

[Rerouter]

I do not know about the UK, but in Australia the pad mount and some pole mount transformers have basic telemetry in them telling phase voltages, frequency and relative phase shift, with that information they could narrow down the fault to a few KM wide area,

next up would be complaints phone calls, they have your street address on file so when people ring they log where you are on there grid and it lets them narrow it down to possibly a street or road, they could then send a tech to walk to a middle house in where they think the fault lies, open up the power meter and see if it spins on the phase that is at fault, cutting the area in half when in a residential area,

[chickenHeadKnob]

Yes I have some Knowledge, but it is out of date. Back in the mid 80's there began commercial development of microprocessor based fault locators. The basic operating principle is simple: The line is continuously monitored by the fault locator, when a fault condition occurs the traveling wave of fault conditions is accurately timed, peak determined,  with DSP  operations usually,  something  like Radamacher-Walsh transforms. By having synchronized clocks the fault locator(s) can report distance to fault, and that is relayed to the dispatched crew.

[DenzilPenberthy]

There are various time domain reflectometer techniques that can be used to find underground cable faults. Essentially you send a signal down the cable and it is reflected back from the short/open circuit. By timing this accurately, you can make a surprisingly accurate estimate of the distance to the fault.



Also there is the 'thumper'. This generates high voltage high current pulses which cause an arc at the cable fault. these arcs make a loud clicking/banging noise which can be detected acoustically or the arcs can be detected magnetically.

[Kremmen]

Not directly relevant to UK but here (Finland) all meters are nowadays required by law to be remotely readable. They report any power outages directly to the power network operator without human intervention. That of course won't pinpoint say a cable fault but it will indicate the affected area. Together with grid telemetry the operator will get a pretty decent picture of the fault location. It does work as i get SMSes of actual and planned outages.

[SeanB]

Simple way is they depend on people phonong in to say the power is out. They see an area has a few calls about power outage, and that this area is fed by either a single substation or by a larger substation. The larger substations do have a monitoring system that shows load current, phase current and voltage to the central office via a remote viewer ( either a radio link, via a phone line call or via a dedicated fibre link these days) to show thasr feeder that has a fault. then they despatch a crew to the area to fix it.

the crews go to the distribution station and see which feeders are tripped, and they check if the reclosers have tripped safe ( for overhead lines reclosers operate 3 times then need manual reset as there is a hard fault, for things like a bird or branch they trip then reconnect typically 2 seconds later) or if the ground fault sensors have tripped for a circuit. They then go out to around the middle of the line and open it with either line switches or with oil switched, and reclose the line. If it works then the fault is downstream and the do the same again further or if it trips still they work upsteam to the station till they find the faulty section. When found they might use a ring feed to restore power and work on the failed cable later at leisure, or for overhead they have to walk the failed section till they find the fallen pole, tangled wire or tree branch on the wiring, then fix it.

If it is on the output of a substation they will typically check fuses, then find the section that has blown fuses. First they replace the fuses ( using a blast shield as if there is still a fault the fuse may rupture during insertion) and see if the fault has cleared. If the fault is hard then they have to trace the cable, either by thumping it with a tester or by using a TDR to locate the short. TDR works quite well, you can see every joint, every tap to a subscriber and every change in cable type along the line with it. When you get distance to fault you pull out the cable lay maps ( always on paper and often a century or more old) and measure along it till you get to the area, then go there and set up your camp. Then you look for fresh digging, open it up and fix your cable, or if there is nothing then you use a tone locator to find the cable ( there could be a few cables in the section, and only one failed one) and start to expose cable till you find the fault ( normally very visible as a blown area) and splice in a new section with 2 Scotchcasts and after testing insulation resistance you apply power again.

[Galenbo]

There are various time domain reflectometer techniques that can be used to find underground cable faults. Essentially you send a signal down the cable and it is reflected back from the short/open circuit. By timing this accurately, you can make a surprisingly accurate estimate of the distance to the fault.

We used this for underground and air coax troubleshooting.

When shorted of broken, it did well, but when water entered the cable, giving reflections, it gave un -intuitive responses.

Worked well, occasionally also for electrical wires and taps. You could also check if someone made an electricity connection before the sealed meter, even when nothing was connected on it.