He was trying to keep the chiller from overheating, but the real problem was a condenser that kept shutting down. When electrical equipment shuts down, measuring current is a good place to start. However, the disconnect on the condenser did not have a "cheater," so I would have to shut down the unit to take measurements.

"We can't shut it down," the maintenance man said. "We've had so many outages this summer, we can't tolerate another."

Instead, I measured voltage and current for each fan and the whole unit at the condenser's control panel. The currents were in line with the ratings, but I noticed that Phase B was half an amp higher than the other two phases. Small amperage differences are not unusual because voltage and wiring are never exactly the same. The voltage caught my eye. On every fan, Phase C volts to ground were 3 to 4 v less than the other phases.

Curious about the voltage, I wanted to go to the main switchboard to see if the difference existed on the incoming service. There was no good place to measure at the main, but the disconnect for the idle chiller was available. It was fed directly from the main panel. Without a load, the voltage would be the same as at the main. The voltage was the same across all three phases, so the problem was not a service imbalance.

Next, we went to the panel to check the voltage leaving the breaker that fed the condenser. Once again, all three voltages were equal.

"We have no choice," I said. "I don't see a problem on the line side of the disconnect, but there is clearly a problem on the load side. We have to shut the condenser down so we can open that disconnect." What I saw nearly caused a fire. The Phase C fuse ferrule was discolored, the wire had melted insulation, and there was a little red glow where a damaged line-side lug had arced through an insulator to the ground lug. Current leaking off Phase C to ground was like a large single phase load. It not only accounted for the lower voltage on Phase C, but the heat it generated likely was the reason for the blown fuses that plagued the system. A new fused disconnect would solve the problem.

Ken Elovitz
Energy Economics Inc.
Foxboro, Mass.