Tracking down rainy-season earth leakage trips: how to locate insulation degradation

The plant's wiring had been in service for many years, and the residual-current device tripped frequently in the rainy season, affecting production. The earlier response had been to step up the RCD's rated value progressively - a danger signal - and the first thing I did on taking over was restore the original parameters.

Random tripping meant there was more than one point of insulation degradation, and that it was humidity-related.

We used sectional insulation testing: dividing the circuits into sections by distribution board, measuring and logging insulation resistance section by section, and re-measuring after rain for comparison. Three sections showed a sharp drop in insulation value after rain, identifying the wall-penetration run and the buried run.

Excavation and inspection confirmed it: aged sealing at the wall-penetration conduit letting water in, and standing water in the buried cable joint pit. We redid the waterproof sealing, replaced the joint and raised its position in the pit, and the tripping went to zero.

Earth leakage investigation narrows the range by sectional measurement and confirms the correlation by re-testing in different weather. Raising the RCD settings only hides the problem, and leaves the hazard for the next electric shock.

In older plant buildings, wall-penetration and buried sections are the weak points of the insulation; the moisture-proofing workmanship is worth getting right in one go during a retrofit.

Rainy-season leakage trips are the classic "fine as soon as the weather clears" fault: on a sunny day every parameter measures normal, the maintenance worker came twice and said there was no problem, and the next rain tripped it all the same. The breakthrough in this case was method, not instruments: sectional insulation testing plus a comparison re-test after rain - divide the circuit into sections by distribution box, measure and record the insulation resistance section by section, re-measure after rain, and which sections plunged is plain to see. The wall-penetration and buried sections were pinpointed, excavated to verify, and re-waterproofed, reaching zero in four days. For environment-related faults, comparison data is a hundred times more useful than a single measurement.