We have an Alstom 620MW, 689MVA, 20kV generator on our power station. We are experiencing the following challenges:
1. High hydrogen seal oil flow; about 220l/min whereas we used to run at 90l/min in the past and on all other units. The oil flow rate is about 90 - 110 l/min on barring, as opposed to 20l/min in the past.
2. Our seal oil vacuum compartment runs empty when we run the turbine up. This cavitates the AC main seal oil pump, but we miraculously keep the turbine from tripping on low differential pressure by switching on the emergency DC pump.
3. So we have a situation were now we keep the generator on load by running the AC and DC pump. The discharge pressure on the AC pump is about 1400 kPa as opposed to 900 kPa under normal conditions.
4. The Hydrogen/seal oil differential pressure is unstable and fluctuates around 30kPa as opposed to being stable at 50 kPa.
We suspected that the hydrogen seals are damaged and shut the unit down for repairs. We replaced the seals and the seal springs and came back with the same issues. We have turned the seal oil plant upside down but still can point to the source of the problem.
Has anyone experienced this problems and how did you sort them out? Can anyone help with a proper diagnosis or lead me to a source that can help?
First, I am not use to metric so the values don't mean any thing to me, so I an just looking at the ratios you provided.
I had the a very similar incident a few years ago. However in my case I did not have an inservice flow meter, but the vacuum tank running low with turbine speed increase and needing the DC did happen.
But just with your measured flows you KNOW you have a problem and it is either the seal rings or the housing assembly resulting in the excessive flow.
Mine was with a GE generator and the seals had test throttling valves that allowed me to determine which end was my problem.
the mechanics were positive they had things correct and I had to argue to have them open the bearing/seal.
observing the seals with upper halve bearing removed there was a gusher of oil coming out instead of the expected thin low flow sheet of oil. In this incident there was an assembly error that had the seal rings pushed away from housing face
Does your flow monitoring provide flow used by either seal so you could determine which is the reason for the excessive flow?
I don't know the alstrom's system... Is there any bypass for seal oil flow? for example on the GE design the collector end seal flow could be dumped with an equalizing valve for regulator sensing. I would expect to find similar equalizing valves for a delta pressure gauge or sensor.
I try to rely to maintenance that making arrangements so that the seal oil can be applied while the upper half bearing is off will provide some ideal that seals are assembled correctly by observing only the least amount of equal flow around the shaft. ( the sheet of oil may not extend over the journal an inch with good seals at 10 psig).
the one coincidence I observed with the event I mentioned was that when the flow was near pump name plate flow, the vacuum tank fill valve was also at maximum flow and thus the tank level began to decay. Not sure that was an intentional design, but it makes sense not to allow makeup to exceed seal flow capacity to bring this adverse condition to light.
could you provide a drawing of your arrangement
I have checked the flow with an ultrasonic meter on both lines and there is a small difference of what is going on either side. Say about 10 l/min difference.
We do have a bypass line for what we call differential pressure regulating valve on our system.
Will edit the drawing and post
Does the ultra sonic indicate only 40-50 l/min per seal. does you Ultra sonic confirm the total flow of 220 l/min.
I would want to verify flow meter readings on sister unit.
Here are some thoughts:
I assume you have verified that the seal oil differential pressure instrumentation is correct. To do this you need to measure the generator gas pressure with a calibrated gauge, and the seal oil header pressure at each end of the generator at or close to the shaft centerline and subtract the two readings. Water that collects in seal oil sensing lines can cause errors.
I'm only familiar with large GE generators, not Alstom. Is the seal oil compartment the same as GE's seal oil vacuum tank - and the AC seal oil pump simply takes oil from this tank? I will assume that this is the case.
Is the AC seal oil pump cavitating because it runs out of oil? Is the supply of oil back to the tank not capapable of keeping up with the pump?
GE seal oil pumps tend to cavitate even with adequate supply when there is a good vacuum and the oil is cold. Typically once the oil warms up to normal temperature the cavitation stops. In this case the oil is just too thick to pump.
Vacuum leaks in the pump suction or seal oil pump shaft seals can also cause cavitation - as well as contaminate the hydrogen in the generator with air and moisture.
Our seal oil flows to each end of the generator are monitored with flow meters and continuous montitoring. The seal oil flows always do some very wierd things and get very high during rollups to full speed during startups. They always settle down once full speed is attained. I have always taken this to mean that there are resonant frequencies that caused the segments to dance differently depending on the speed. Rotor vibrations probably have a large influence on this. Variations in flow can also be seen on turning gear.
Generator casings do distort with increased hydrogen pressure and this can impact seal oil flows.
What is your H2 leak rate? Are you blowing H2 past the seals when flows are high?
On GE machines you can check the hydrogen side seal oil flows by isolating the return H2 float trap and watching the rate of level rise. You can also check the total H2 seal oil flow by isolating the supply to the seal oil vacuum tank and watching the rate of level drop.
On our GE seal oil system, you could gradually close the discharge valve on the DC seal oil pump (for a very short time) to see if the AC can supply enough oil, and then shut the DC pump down. However, you still will not know whether the DC pump alone could supply enough oil to feed the seals during the worst case flow.
I assume the unit had been running well at one time. What changed? Did you have an overhaul? Were there some modifications? Was there an event? Or did this problem build over a long period of time?
I'd recommend hooking up a high speed recorder and adding instruments to continuously monitor seal oil DP, speed, seal oil flow to each end, and any other parameters of interest.
Yeah this was does, calibrated gauges were used for verification. All instrument lines were flushed to remove air and water. What is worrying is that the pressure at the generator is about 500 kPa, whereas the pump discharge pressure is 1400 kPa 16m below
Yes correct, the seal oil compartment is the same as GE's seal oil vacuum tank - and the AC seal oil pump simply takes oil from this tank?
Is the AC seal oil pump cavitating because it runs out of oil? This is correct. We then allow oil to flow from the pressure compartment into the vacuum compartment to supply just enough to keep pump from cavitating.
"The seal oil flows always do some very weird things and get very high during rollups to full speed during startups".
This is also a case at our station, but we cannot establish a pattern
>What is your H2 leak rate? I don't get this? because of cavitations on the AC pump, the vacuum pump is switched off resulting with deteriorating hydrogen purity, therefore we are continuously venting.
>I assume the unit had been running well
>at one time. What changed? Did you
>have an overhaul? Were there some
>modifications? Was there an event? Or
>did this problem build over a long
>period of time?
We came back from a short outage and everything just slowly started to go haywire. No modifications were done at the time.
>>Yes correct, the seal oil compartment
>is the same as GE's seal oil vacuum tank
>"The seal oil flows always do some very
>This is also a case at our station, but
>we cannot establish a pattern
assuming your seals are the same as GE's, do you have the quadrant segments or the bolt together seal rings?
If the quadrant segments, they have a tendency to get out of step during the restart of seal oil.
If the bolt together, it was the installed depth of the access plug that was the reason for the high flow event I mentioned
Quadrant segments. What factors would cause the seals to go out of step? We went on an outage and replaced the seals sometime ago. we noticed some rubbing marks on the seal housing. so when we experienced the same problems after the outage we thought that the seals are out of step. we put the turbine on barring and operated the jacking oil pump a number of times. Suppose the segments were out of step, how would one restore them without opening the generator?
the most likely is the sudden rush of air past them and shock of oil impact when the system is returned to service. with seal oil down, the regulator will go wide open, then, even with just the TGOP starting, there will be a rush of air then oil to the seals.
the seal oil return to service procedure I developed includes isolating seal oil prior to any oil pump starts. start the seal oil pump and lift its relief. open regulator sensing bypass to establish flow path from CE seal to detraining section. very slowly open regulator bypass. take minutes to get least pressure indication. build pressure to design seal dP. allow to fill sensing lines and detraining section for about 30 minutes. put regulator is service. verify total seal flow at stop rotor/no gas pressure is less than 25% of normal at running
>What factors would cause the seals to go out of step?
I have heard "talk" about "shocking" the seals to fix out of step. mainly bypass regulator and increase seal oil pressure X2 (X3, etc), and dropping back to regulator. this procedure has been a topic several times on Yahoo group GETURBGEN with the consensous of worth a try, but no luck even happened
>how would one restore them without opening the generator?
did you join IGTC yet?
where did you measure that you had 200 l/min seal oil flow? if you think you only have the normal 45 l/min at two seals, look for where the oil could tee off. having a system schematic would help.
not to get side tract form your main concern, but if your design is similar to GE, the main seal oil pump relief normally at 900kpa, now being 1400 is a little concern. I understand that it should be higher since the ESOP is also running, but I would not expected that high.
You know that running of ESOP will be suppyling untreated oil to the seals so you can expect your purity to start dropping. most vacuum design GE do not have provisions for sampling purity at the seals, but only once air migrates into the main casing. also the scavaging is the main casing and not the seal drain sections where the purity will be the worst
you could try posting on the International Generator Technical Community's Forum.
this site has capability to share attachments which would be benrficial for those that are familiar with similar, but not necessarily the Alstrom's design
trying to locate what category to post in my be a challange.
I would suggest
..Inner hydrogen cooled stator windings
Is the unit still in service? I would have expected it to be in outage by now