Please, what could be the cause of a unit which is on base load tripping on compressor Bleed Valve fail to close?

 

Mr. Ray,

Pray tell, sir: What unit are you working on, and when was it installed/commissioned?

And, has the turbine control system been upgraded since installation/commissioning?

What is the current version of turbine control system in use on the turbine-generator?

Does the turbine have a conventional combustion system or a DLN combustion system, and if the latter, which version of DLN combustion system does the unit have?

Most importantly, though, which Process Alarm--specifically--is annunciated when the unit trips?

Finally, what have you done to troubleshoot and resolve the issue--and, what were the results?

You want an answer; we need actionable data.

 

Dan, Like CSA has said the more info you provide, the better the answer we can give. We can assume a whole lot, and probably make up some really good sounding stuff, but that doesn't help anyone.
In the past we could assume you are working on a GE heavy gas turbine, maybe the common 7FA or maybe a little newer 9E. We could assume it has a MKVI or maybe even a MKVIe. If that was the case I could suggest that either a compressor bleed valve opened while the unit was online, or maybe the switch that indicates the CB is closed failed or the wire broke. The control system thought one of the CB's was open, which with newer application code would cause the unit to runback and open the breaker. But all of this is just guessing, more specifics from you would be a good start before we spend more time here.

 

Have you looked at your P&IDs ?

 

WOW! The old-timers are coming out of the woodwork! Good to see you're back, MIKEVI and long time no hear, glenmorangie.

Unless the original poster can tell us why the unit tripped, we are all just speculating. EVERY condition that causes a GE-design heavy duty gas turbine to trip has an alarm associated with it. Now, often the alarm text messages are cryptic or don't make a lot of sense--and as all we old-timers know when a GE-design heavy duty gas turbine trips (meaning and emergency shutdown) often there can be several trip conditions that get annunciated--but only ONE is THE cause of the turbine trip.

So, I am getting and have gotten away from my standard practice of asking for alarm information when someone says, "Why did my turbine trip?" We all have to be firm on this--because I know I, and others, waste a LOT of time trying to write responses that can cover as many of the common reasons why this or that problem or turbine trip may occur. In this case I suspect we are dealing with a unit under the responsibility of GE Belfort, and we all know they do some odd things when it comes to control and protection (because they can). So, I will not be surprised to find this unit has DLN-I combustors, is a Frame 9E GE-design heavy duty gas turbine (or a Frame 6B--because GE Belfort is responsible for both these frame sizes these days), and it has a Mark V or Mark VI or Mark VIe that was configured and programmed by GE Belfort. Because I have never personally seen a unit trip at base load because the compressor bleed valve(s) failed to close. They either closed around generator breaker closure or they didn't, in which case the alarm would be 'Failure to Close' and may result in a trip 11 seconds after being commanded to close. But, to get loaded to Base Load with an open compressor bleed valve (or more than one) and then trip on 'Failure to Close'--no way, not now how, uh-uh. Not happening. (Well, except if GE Belfort did the programming, then it might be possible.)

Anyway, good to see some familiar faces! Hope everyone is safe and healthy! Flying before the Delta variant of COVID-19 was scary; now it's really scary. So be safe, and stay healthy!!!

 

Thanks CSA, safe, healthy and bored.
I feel sorry for people who have been put in situations where they are unqualified for what they are doing, it is the main thing that nearly drove me away from the business a good few years ago. Managers want megawatts and as cheaply as possible then wonder why they run into problems.
Let's wait and see how our friend replies !!

 

PUT in situations, or did they take the position thinking they would get some OJT (On-the-Job-Training) and experience and didn't know what they'd gotten themselves into?

I would posit that a big part of the problem is that Managers expect too much from automation, because salespeople and other less-than-knowledgeable people say automation is the answer to everything. The automation will protect the machine under all conditions and scenarios. But that's not true in ALL conditions and scenarios, and it presumes people will try to actively investigate alarms and resolve them.

Of course, an alarm troubleshooting guide would be very helpful. But, that is difficult to produce--not impossible, because the US Military gets VERY GOOD manuals for equipment just exactly like turbine-generators that most people with an ability to read can use to troubleshoot and understand the equipment and keep it running. The difference is that this costs a LOT of money to produce such documentation, and most commercial companies won't pay for such information--and in many cases don't even know to ask for it during the purchase negotiating, they just ASS-U-ME it is available and will be provided by the equipment manufacturer. The military, however will (and does) pay for this level of documentation.

People just think that every GE-design heavy duty gas turbine is just exactly like every other GE-design heavy duty gas turbine (there ARE a LOT of similarities!) and that they can just "propose" a problem and provide zero details and get an answer to their query.

I would suggest that the "Compressor Bleed Valve Failure to Open" trip message was annunciated AFTER the actual trip condition, but because it was the "last" trip message annunciated it is being presumed to be THE trip condition, when, in fact the FIRST trip message annunciated is the actual cause of the trip. GE should be MUCH better at this (annunciating turbine trip conditions and blocking subsequent, erroneous trip indications (a first-out trip indicator has been produced and provided by Nuovo Pignone on their steam turbines...!--it wouldn't take that much effort, really!). So, after the trip, one or more of the compressor bleed valves failed to open--or the limit switch failed to detect a fully open compressor bleed valve 11 seconds after the bleed valves were commanded to open. But, the turbine had already tripped....

As you suggest, glenmorangie, we will--hopefully--hear more from the original poster.

 

Looks like we won't hear back from the original poster....

I still fail to understand how a unit could get to Base Load with one or more compressor bleed valves open. If that were to happen, Base Load would be less than usual--because a portion of the axial compressor flow/pressure was being dumped to the exhaust through the compressor bleed valves. So, there's something very wrong here.

Something that a chronological list of Process Alarms and some more information about the unit in question would answer.

 

Interesting discussion thread. I have a few 7EA peaker sites I'm responsible for - 3 sites (all 2X 7EA with DLN1 MarkV, each) have similar logic as described - unit enters a runback at full load if CBV (VA2-1 or VA2-1) don't indicate closed after a given timer. A 4th site (6X 7EA with DLN1 MarkV) has dissimilar logic - no issues if the valve doesn't close.

The last seems correct to me - risk is loss in performance, it seems. In practice, reportedly there are no other issues. I'm curious why the other 3 sites have the "odd" logic - thoughts? I thought there could be a potential impact to DLN1 equivalence ratio which we'd never see since we have no emissions or dynamics monitoring, but I'd think not enough to actually impact operability.

 

Interesting dilemma--the difference in logic. It would be very interesting to understand why, from the OEM. Have you tried looking at the Control Specification drawings for the various sites to see if they have any "information"?

For DLN-I systems, I think the risk of stable operation is much higher than you are assuming. I don't know if dynamics would be so much of an issue if the compressor bleed valve(s) opened while running in Premix Steady State, but I would think it would have an effect on emissions and maybe even result in a primary zone reignition and "extended lean-lean." You MIGHT be able to re-load to Premix Steady State to get into low emissions mode--and you might not, if the compressor bleed valves remained open (it would probably depend on if the units had IBH (Inlet Bleed Heating) or not).

For conventional combustor-equipped units you are correct--there would be a decrease in output if the compressor bleed valves opened while operating. Power output would also decrease for DLN-I units--but it would change the air flows through the combustor, and with DLN being an "on-the-edge of stable combustion" system it would probably cause knock-on effects, such as primary zone reignition. Perhaps with faster processors the control system could respond quicker--who knows (certainly not me).

 

Interesting dilemma--the difference in logic. It would be very interesting to understand why, from the OEM. Have you tried looking at the Control Specification drawings for the various sites to see if they have any "information"?

The sites are old, GE doesn't support anymore - and with the state of GE today, it seems hard to find good motivated help to dig into anything. These are purchased units and we can't find complete manuals including control spec - I have one from the units WITH the logic. Those units do have IBH - they units with allowed continued operation with CBVs open do NOT. I agree with you on the risks for combustion operation - BUT the site (allowed to run) argues: transfers from primary to lean-lean to secondary transfer to Premix baseload ops is smooth. Good point on primary zone reignition - instead of absolutely trusting the site is correct, I'll hunt through PI and validate with flame eyes indication and exhaust T/Cs.

Thanks for the response!

 

Is this a "thing"--compressor bleed valves opening while operating at load (in Premix Steady State)?

To my way of thinking, if the compressor bleed valves are opening intermittently under load and while in Premix Steady State--then the solution is not worry if the logic will take some action, but rather get to the root cause of the compressor bleed valve's inability to remain closed under load. I've LONG advocated for the use of clean, dry instrument air for compressor bleed valve actuation. It's cheap and reliable. And keeping moisture out of the bleed valve actuators (and solenoid-operated valves) is usually the best remedy for bleed valves which won't remain closed, or won't fully close (or fully open).

CPD (and CPR) is critical to DLN-I operation; it's key to combustion reference temperature calculations. And CPD (and CPR) is going to suffer if the compressor bleed valves won't stay closed.

Anyway, yeah; the OEM is a shell of its former self. And, true to form, all they care about is the newest, biggest machines. I'm kind of hoping MIKEVI will chime in on this thread.... I'm sure he has something to offer.