Hello,

In the summer 2024 our 7F gas turbine hydrogen cool generator (7FH2) had oil leak into the generator. We estimated about 200-300 gallons of lube oil leaked into the generator. We tried to drain out the oil at low point drain line but believe some oil might migrate into the stator or rotor also.

We also removed the coolers and clean them during an outage.

We did not note any issues during a normal running condition (100mw - 150mw(baseload)). However, at the summer VAR testing per NERC FAC-008 in June 2025, we received an alarm volt per hertz, which was a new alarm. The requirement of this test is to increase mvar until the unit power factor reaching +0.85 pf, at about +0.865 this volt/hertz alarm came in.

My Question is, could the lube oil in the generator cause this alarm?

Our next major outage will be in 2028, and would like to know if any other plant had the same problem and experienced any issues.

Thank you

 

@albertc70,

Volts per hertz is a measurement meant to indicate excessive voltage for the operating frequency, or in some cases much lower than normal voltage for the operating frequency. We don't know what the system voltage was at the time of the test. If it was high and the generator terminal voltage had to go very high in order to "produce" the MVArs required to reach a power factor of +0.85 then that could be the reason. But, again, you didn't tell us what the conditions were at the beginning of the test (system voltage level) and we don't know anything at all about the type of system the machine is synchronized to--if any.

While many things are possible, it's unlikely the lube oil which was drained from the generator casing is related to this problem.

BUT, 200-300 gallons of lube oil is a LOT of oil and it is possible that some oil may have made its way into the generator rotor windings and when the excitation to the generator rotor was increased during the test that some previously unknown weak insulation in the generator rotor "broke down" (meaning the insulation allowed current to leak through where none should have leaked) and resulted in this problem.

All, I can add is the generator should have been inspected when the oil was found inside initially--ESPECIALLY after the quantity of oil was known. And, with this latest issue it certainly seems like if it wasn't inspected then it should be now. And I sincerely hope the hydrogen purity is being very carefully monitored and maintained after such a quantity of oil was drained from the belly of the generator.

 

@albertc70,

Volts per hertz is a measurement meant to indicate excessive voltage for the operating frequency, or in some cases much lower than normal voltage for the operating frequency. We don't know what the system voltage was at the time of the test. If it was high and the generator terminal voltage had to go very high in order to "produce" the MVArs required to reach a power factor of +0.85 then that could be the reason. But, again, you didn't tell us what the conditions were at the beginning of the test (system voltage level) and we don't know anything at all about the type of system the machine is synchronized to--if any.

While many things are possible, it's unlikely the lube oil which was drained from the generator casing is related to this problem.

BUT, 200-300 gallons of lube oil is a LOT of oil and it is possible that some oil may have made its way into the generator rotor windings and when the excitation to the generator rotor was increased during the test that some previously unknown weak insulation in the generator rotor "broke down" (meaning the insulation allowed current to leak through where none should have leaked) and resulted in this problem.

All, I can add is the generator should have been inspected when the oil was found inside initially--ESPECIALLY after the quantity of oil was known. And, with this latest issue it certainly seems like if it wasn't inspected then it should be now. And I sincerely hope the hydrogen purity is being very carefully monitored and maintained after such a quantity of oil was drained from the belly of the generator.

Hello WTF,

Thank you for your help and explaining about the volt per hertz issue.

Here is the timelines of the event,
-The oil leak happened in Sep 2024 at a short scheduled shutdown
-Before restarting we drained the oil out of the generator belly.
-In Nov 2024, we had another scheduled outage and the generator was de-gased and oil was cleaned (removed hydrogen coolers and cleaned them, the generator end shields were removed (the rotor was still in place) and cleaned, GE generator specialist inspected and performed an offline test at 5k volt and did not find any issues. And recommeded to continue to operate until next major outage 2028.
-In June 2025 we performed the annual Summer VAR test (required by NERC), the test was to start at power factor around +1 or 0 MVARs then creased MVARs until pf reached 0.85 (previous VAR test (Aug 2024) MVAR was recorded at 85 MVARs unit was at baseload and pf was at 0.85 pf and no issue with volts/hertz alarm), but this year test we had a volts per hertz alarm came in at around 80 MVARs and pf was at 0.865. We stopped and reduced VAR down to 76 MVARs and the alarm was clear (pf was about .875).
-After that our relay engineer reviewed the data and informed that the volts per hertz alarm came in at 109%, trip set point is at 115%. This is a 60 hertz system and the generator rated voltage is 18kv.

My next question is, should we limit the test at 0.87 fp, not at 0.85 pf?

Thank you

 

@albertc70,

I deliberately put this in the very first paragraph of my initial response:

"... We don't know what the system voltage was at the time of the test. If it was high and the generator terminal voltage had to go very high in order to "produce" the MVArs required to reach a power factor of +0.85 then that could be the reason. But, again, you didn't tell us what the conditions were at the beginning of the test (system voltage level) and we don't know anything at all about the type of system the machine is synchronized to...."

The system voltage the generator might have had to reach in order to achieve the 0.87 pf may have been (probably was--since you say this alarm had never been annunciated before) high for the load and the operating conditions at the time. Compare the voltage of the system the machine was synchronized to at the time of the test to some other days at a similar load (try to choose several days when the ambient temperature was higher or lower than normal as well as a "normal" ambient temperature day or two to get some good data; in other words, don't just look at the day before or two days before) and you will probably see the system voltage at the time of the test was at least a little higher than normal--forcing the excitation system to have to get to a higher voltage (at the system frequency--that volts/hertz thing!) to try to achieve the desired power factor during the test.

We also don't know if there are any tap-changers in the connection to the system ("grid") the machine was synchronized to and what positions they were in at the time of the test--versus other days (as above).

Unless the system and machine location are very unique in the world, grid voltage can vary depending on many factors (load; types of loads; other machine operating conditions on the system your machine is synchronized to, etc.).

 

WTF,

Thank you for your advice, I will monitor the system voltage at the next test.

 

albertc70

Could you please share what the cause of the oil in the generator was? we're running into a similar issue and are suspecting an increased clearance on our h2 seals.

 

Competitive_End_8637

It was an operator error during a planned shutdown for an outage. The plan was to degas (venting hydrogen) from the generator and take unit off turning gear. When all hydrogen was replaced with CO2, then CO2 was replaced with air, our operator forgot to follow the procedure; normally the differential pressure between the seal oil and the gas pressure inside the generator should be maintained between 5-5.5 psid, if let outside of that range oil could enter the generator. We had that happened and it was caught after that, but we estimated about 200 gallons of oil already entered generator.

We spent about 1-2 days to drain out the oil before restarted the unit, and after a few months we had a hot gas path inspection on the same unit, we hired the OEM crew to remove the generator coolers and cleaned the oil and also wiped and cleaned the end windings when 2 endshield covers were removed for the inspection.