Intriguing DG Deloading problem

Last August (2017), we came across a very intriguing (on the verge of being unintelligble as the problem was staring at us point-blank) problem onboard a container ship.

What was this problem? We had these two DGs , when operated in parallel, which could not deload fully. i.e. Either of the two DGs could not deload to the setpoint for the Deif PPM controller to open its respective breaker. The deloading consistently got stuck at 50~70kW.

Some key points that we considered during the troubleshooting process:

1. Problem occurred in both Automatic or Manual mode. This ruled out the respective PPM controllers being responsible for this problem.

2. Both DGs were able to take on load normally. Both were able to shed load normally (till the magical 50~70kW level that is). This implied that there was no issue with the governors too.

3. We set out to scrutinize the subsystems in between the deload (governor speed down) relay and the DG governors, namely the governor speed control card and the speed electronic potentiometer.

4. Both these items also appeared to function as expected during the loading and deloading process.

5. That's where we hit the dead-end. Everything seemed to be working fine. What's responsible for this... enigma?

After the death of untold number of brain cells, my brilliant colleague, Banyar, had his Eureka moment. (somehow he has such moments frequently enough to solve every problem) He was staring at the LED indicator on the EP300 Comap electronic potentiometer when the revelation came.

Banyar noticed that the EP300 appeared to be at its lower output limit when the deloading was 'stuck'. What did this mean? The potentiometer might not be able to lower its output further to further reduce the DG frequency to shed load further.

Ah ha... an interface problem. A devil in the details... that monster under the bed... From that point, we went about our work to fix the problem. (Adjusted the Governor control card's reference frequency so that for the same potentiometer output, the frequency of the DG was lower) .

Brilliant thinking on the feet. Simply astonishing.