Mark-VI system TCPD, 5 Nos of TBCI, 1 No of TREG & 2 Nos of TRLY card component (MOV - Metal Oxide Variastor of +VE 125VDC) Failed with no alarm for any ground fault or any other alarm & unknowingly during shutdown in GE FRAME 6 Gas turbine. The reason for such huge failure is still unknown & we are trying to find it out if anybody had faced same type of problem in past. than pls help to find out the exact reason & solution to avoid such type of failure in future.

 

Blown MOVs are not necessarily related to grounds. There is likely some problem with voltage spiking on the power input to the TCPD. If you have a DACA then it could be that the AC input to the DACA is not properly grounded (it should NOT be a floating AC that powers the DACA).

Ensure the incoming power is clean (free from spikes) and as described in GEH-5421, Mark VI System Guide, Vol. I. There should be a section that describes the various power supplies and the limits for their excursions.

It's been said before many times on control.com: The output of many inverters (UPS; Uninterruptible Power Supply) is not very clean, and the DACA module is NOT a filter for dirty incoming power, and in fact, it almost makes spikes on AC input worse on the DC output of the DACA. If you are using an inverter/UPS for supplying the DACA, ensure the output is properly grounded (if it can be! check the manufacturer's documentation) and clean.

 

you have mentioned that the failure of the
a. 5 nos of TBCI
b. 1 nos of TREG
c. 2 nos of TRLY

MOV's failed during SHUTDOWN of frame 6 GT. i am assuming the following:

when you say 5 nos of TBCI / 2 nos of TRLY, do you mean 5 relay contact inputs (out of the 24) / 2 relay outputs (out of 6) in the respective boards have failed?

if this is the case then it is a field problem , a high voltage spike had caused the failure of the MOV's.

was any welding work done inside the turbine compartment during the shutdown? and importantly was the processor ON during such welding work.

as a precautionary measure , the mark 6 supply should be isolated and the processor turned off before any welding work is carried out inside the turbine/axillary compartment area.

the TBCD of the PDM has MOV's only in the input end , up to three redundant power supplies can be given 1 DC and 2 AC. did those MOV's also get damaged ??

this is clearly a problem with the 125V DC distribution network in mark VI , as PDB supplies 125 V DC to all the boards mentioned , the problem can be either with input supply end , or as i have mentioned voltage spike from the field. but as you have said that the failure took place during a shutdown where normally welding works are carried out , the best bet would be that the origin of the voltage spike would have been in the field. check this possibility.

 

Thanks for ur reply

Yes you are right, MOV of individual channel for single Channel of 5 diff TBCI cards & single channels of 2 diff TRLY cards failed with TCPD card's main +ve 125 VDC Input supply MOV failed.

It was asked to mechanical dept & operation for any welding work going on during incident or not? but they told that there was no single job was going on during incident & we r following the same practise to power off the control system during any hot work like welding.

Regarding fault from field side we have checked each n every channel input for any ground related or any other fault but every thing found normal

so this possibility is also getting eliminated.

If any further possibility is there than kindly suggest

 

Ok things are getting interesting now. i am sure that the MOV's can only get damaged during a voltage spike , the problem is to find how it originated and where. can you please provide me with the following data.

a.How mant tbci cards are there in your system. please give the data in the following order <pre>
VCRC slot Terminal Board and Terminal board no tag connected Bool value during shutdown
Slot # TBCI input # / TRLY output # tag no / description 1/0 </pre>

give the above data for the failed channels. i need to know the following for accessing the following

did the failures take place in different TBCI/trly cards or just one. if different did the failure occur on just one VCRC slot or different ones.

the tag no and description is required to visualise in field where the failure probes were , may be there is a connection between them , in the same place may be ?? may be the same type or some other corellation between the failed channels.

the bool value is required to ascertain weather only the open channels got faulted or not. please give that information also. ie in the failed channel , were there other switches whose values were always one.

you will need to slog out on this one , but hey the answer to this might just be worth it