Gas Turbine: GE frame 7 Hydrogen Leaking

History FYI:
* Field went to ground three months after a major overhaul on turbine/generator
* unit repaired - Unit passed leak test -Unit ran several times
* hydrogen started leaking
* hydrogen seals found kicked
* repairs made to seals
* leak test passed but not great
* noted seal oil flow meter running 1 GPM
* past history reports shows 30 GPM for several years
* try to perform SHAFT SEAL FLUID CHECK per book but oil and hydrogen started blowing out around the seals abort test
* engineers stated normal seal oil flow due to machine tighter than before
* ok to run - repaired some small leaks on hydrogen system - run machine several time
* H2 leakage started to degrade

hydrogen leakage:
* 6.5 lbs in 9 hours
* note hydrogen sense in the main lube oil tanks and in the cooling water head tank with detector
* valve out all the coolers to verify as major leak - still leaking

Today we have the seals apart looking for concern.

Question:
Is it normal to have hydrogen detected in the main lube oil tank?

What should the seal oil flow be?

Should the air side seal have a bigger gap than the H2 side?

During the SHAFT SEAL FLUID CHECK is oil and hydrogen blowing out around the seals normal?

Any input would be greatly appreciated.

 

Inst_KY,

Is there something "new technology" about the hydrogen seals that would cause the required seal oil flow rate to be less than previously (when the seals were in new and clean condition)? Because, a lot of times when questions like this get asked it's learned that some new, improved technology/design was installed during the outage and that wasn't make known when the original question was asked.

If new seal technology/design was used, what does the provider say the flow-rate should be?

6.5 lb. drop in 9 hours is pretty HUGE. And, one would think it should be relatively easy to find. There are simple hand-held detectors which can be purchased or rented which could be used to check the obvious areas (shaft seals; generator bushing flanges; cooler flanges; end-shield flange joints; etc.)--but with a leak that large one would think soapy water sprayed on flanges would show a leak very quickly. (It's presumed all the obvious culprits have been checked.)

It's not unusual for a small (very small) amount of hydrogen to be detected in the L.O. reservoir, especially if the unit is at zero speed and the Mist Eliminator is off (the Aux. Seal Oil pump will be running), and the seal oil drain enlargement tanks don't eliminate all of the hydrogen which is entrained in the oil. You didn't indicate how high the hydrogen concentration was in the L.O. tank...

BUT, there should be a large, stainless steel working loop seal between the generator and the L.O. tank/return system which should pretty much isolate a lot of hydrogen from getting into the return system/tank. I'm wondering if the extremely low seal oil flow is causing the loop seal not to be working correctly.?.?.?

I would expect if the same type of hydrogen seals as have always been used were installed properly in the system and that the seals were new and the shaft wasn't excessively worn the resulting seal oil flow-rate would be slightly less than before the outage--but not significantly less.

As for the test you mentioned, I would NOT think oil and/or hydrogen would leak out of the seals if the test were performed correctly, and the seals were installed correctly. Something sure seems amiss, that's for sure. If it's the test I'm thinking of, it's done to try to determine actual seal oil flow-rate by using containers of known volume to drain the oil into to calculate flow-rate by monitoring the time required to fill the container(s) to a particular level. That's a difficult test to set up for--but again, if the seals were installed correctly and the hydrogen components were all working correctly (the trap and the loop seal, in particular) that neither hydrogen or oil should be seen leaking from the shaft seals during the test; you're only re-directing the drain flows into a bucket instead of into the main loop seal for a brief period of time (as long as it takes to fill the container(s)).

Again, that's a pretty significant leakage rate, and the low seal oil flow-rate is suspect--but one should be able to check/replace the Seal Oil Flow-rater (as I think it's called) to determine if the gauge is working correctly or not. Also, if OEM-type seals were used in the reassembly (new or re-used) then the air gap spec's should be the same as shown in the generator drawings (which are NOT usually included in the Unit Service and Parts Manuals

Please write back to let us know how you fare in diagnosing and resolving the problem!

One last suggestion--have you asked this question in the 7E Users Group forum? If so, what have you learned?

 

Troubleshooting:
Hydrogen Seal assembled on bench today and the clearances are tight on both air and H2 side. Engineering thinks only rough cut performed not final cuts during outage. Brass Seals had a lot of rub marks. Turbine end bottom seal kicked and stuck out on place.

Hydrogen Seal have been shipped for repairs and will machine to spec per OEM generate assembly drawing. Planning on double checking OEM spec before installing seals.

As far as seal oil flow, we should see an increase in flow after the brass seals are spec out correctly. I will update the post on the new seal oil flow values soon enough and any other changes.

Thanks for the reply.