Riser Fuse Blown on an Old Underground Loop

First, treat everything as energized and rule out the non-cable causes.

1. Confirm the riser fuse is blown and on which phases, and inspect the cutout and the riser arrestors. A single blown phase looks different from a three-phase cable fault.
2. Check the padmounts for obvious transformer failure - bulged tank, oil, burnt arrestors, a blown bayonet fuse. A failed transformer blows riser fuses and is much cheaper to find than a cable fault.
3. Isolate in the right order. At PMT-1 and PMT-2, the elbows can only be pulled under load if they are marked load-break. If they are not, de-energize from the riser first. Pulling a non-load-break elbow under load is an arc flash in a confined space.
4. Separate the transformer from the cable (open the bayonet fuses or the internal switch), then test for voltage at the test points with a proven detector and apply grounds before touching anything.
5. Sectionalize the loop, ground both ends of the suspect section, and run the TDR first - it gives a distance to the fault without putting a surge into the cable. Only then consider the thumper, and only with the section isolated, both ends grounded, transformers disconnected (the surge damages them), radio checks, and the dispatcher's agreement.
6. The fault is between the two points where the TDR or thumper responds. Record the distance and compare it with the cable route and the last dig-in location.

Isolate the faulted section at both ends, then close the normally open tie at PMT-4 to feed PMT-2, 3 and 4 from the east feeder. The gas station comes back as soon as the tie is closed, provided the isolation is complete so the backfeed does not energize the fault.

Before closing the tie:

• Feeder capacity. The east feeder is at 65% of 600 A, roughly 390 A. Add the loop load at 03:00 and check the total against the feeder rating and the tie switch rating. At minus 18 degrees the greenhouse and strip mall heating load is near maximum, so use the actual load, not a summer average.
• Phasing. Confirm the tie was built with matching phase order. If it has never been proven, phase out with a phasing set - a wrong phase order will destroy every three-phase motor on the loop.
• Clearances and grounds. Confirm no crew is working and no grounds are applied in the section that is about to be energized.
• Protection. Note that the protection arrangement has changed: the east feeder breaker now sees the loop, and the loop transformers are being fed from the other direction. Tell the dispatcher, and check the transformer fusing is correct for a backfeed.
• Cold load pickup. Everything on the loop has been off in extreme cold; the inrush and heating load may push the east feeder harder than its normal peak. Consider restoring the loop in two steps rather than closing the tie onto everything at once.

Then notify the customers, keep the faulted section tagged, and log the switching.

Locate and permit first, then dig, then repair, then test.

1. Locate all buried plant in the area, raise the ground disturbance and locate tickets, and control traffic. The driveway is private property, so the homeowner's permission and a restoration agreement come before excavation.
2. Excavate by hand near the marked cable. At typical burial depth this is an excavation that needs shoring and a spotter, and the exposed cable needs protection from the bucket teeth.
3. Cut out the damaged section and install two splices with the separation the utility's standard requires, commonly about 3 m. Check the cable ends for moisture before splicing - 1987 XLPE in wet soil can be water-treed, and a splice on wet cable fails again.
4. Test before re-energizing: insulation resistance and the utility's acceptance test (DC high potential or very low frequency) on the repaired section, with the transformers isolated and the far end grounded as required by the procedure.
5. Re-energize from the riser, check voltage and phase rotation, then open the tie and return the loop to its normal open point.
6. Record the fault location, the test results and the repair so the asset record shows whether this cable is developing a pattern. One dig-in is bad luck; a second fault on the same 1987 run means plan a replacement rather than another splice.