Nuovo Pignone Compressor

[bohdand]

Hello,

I am hoping that someone is able to help me. We are using a 2 stage motor driven Nuovo Pignone centifugal natural gas compressors (3x50% capacity - 2 normally running) to boost our ~190 PSI supply pressure up to about 480 PSI for a 7FA combined cycle power plant (2x1) and we are having issues with load transients when one 7FA trips. We've had a couple of 'experts' look at the problem and they have identfied some slow response loops on the recycle valve and vent valve that are used to prevent over pressurization of the discharge header when one 7FA trips (2 compressors in service).

But, another issue appears to be that the Nuovo Pignone compressor is designed for pipeline or process industry applications and not a power gen combustion turbine application. Although we know there are issues with the existing control algorithm that needs improving, my fundamental question is - is this compressor suitable for our application where very fast response is needed for unit trips? It works fine under normal load changes but not the transients.

The flow rate for one compressor is 3600 but I am not sure whatthe unis are (MCFS maybe?).

What is specifically happening is that when one of the 7FAs trips, there is an increase in pressure on the header, the anti-surge valve opens and the vent valve opens (dumping gas to atmosphere) and keeps the pressure below 525 PSI, but these two valves then are very slow to close due to ramp limiters in the logic causing the pressure to drop down to 415 PSI and intitiating a trip or liquid fuel transfer on the remaining 7FA - apparently this was done intentionally to prevent instabilities in the control response.

Any input would be appreciated and welcome.

Thanks,

Bohdan

 

[rotw]

I assume 7FA is a Frame 7 GT ? Is the compressor feeding the GT as fuel gas booster? variable speed type?

In any case, when GT trips seems the flow reduces, system pressure goes up, antisurge valve opens for protection. Strange that you blow off natural gas by the way.

When anti-surge/blow off valves open, compressor is in recycle + blow off mode - so that brings the operating point to the right of the operating curve. Possibly depending on your antisurge algorithm and how fast flow reduction occurs upon GT trip it may be the case that your flow is so low that the anti surge valve/ blow off valve is triggered to go directly "fully open", in this case the flow goes from surge point to right of curve quick, of course pressure drops but compressor is normally safe.

If the controller does not adjust then the point back to left of curve and/or valve actuator is slow when closing, it might be the case that pressure is low for a while and initiate GT trips. although as trip is not a matter of duration but of a set point signal.

I dont think this has to do with the type of compressor even not the brand; it is dynamic response problem, main aspects I could think of (just brainstorming) :

- anti surge valve and blow off valve CV (original sizing)
- Type of controller (CCC, or else) ?
- Adjustment of trip settings, can this be done ?
- Were the controllers programmed to handle the case scenario you described ? to me it looks you are in a sort of off design situation with your plant. Thus not part of the original design specification of the machine.
- Why do you have so many GT trips? Is this by design or are you facing a situation of spurious / excessive trips?
- do you have a case where two compressors are fighting each other, e.g parallel run?

"If you want to acquire a knowledge or skill, read a book and practice the skill".

 

[georgeverghese]

From the limited information supplied, it does seem that there is insufficient control buffer volume between these NP booster compressors and the power gen GTs ( GE Frame 7?? as rotw guesses)

This is a common design flaw. The buffer volume should allow sufficient response time to allow the compressor to return back to normal operating pressure before the vent valve (on the GT side) gets triggered into operation.

Without being reliant on the speed response of the antisurge control loop for these situations ( for obvious reasons), would guess you need at least 3-5seconds accumulation time between the max normal delivery pressure and say (525-10)psig to allow for the normal capacity control loop at the compressors to recover for the case when one 7FA trips and both compressors are still running at max flow into the delivery pipeline.

The buffer volume can be in the form of an oversized delivery line between compressors and power gen GT (if the line is long enough) or a vessel.

 

[bohdand]

Thank you for your responses.

Yes, this is a GE Frame 7 Combusiton Turbine application. We are using a fixed speed compressor with variable inlet vanes for flow control. Gas vent valve is based on high header pressure, but we just learned that the recycle valve is slow to open further contributing to the problem.

At this time we are not considering a buffer volume due to cost, space constraints and permitting difficulty in a US city.

We are hopeful that some tuning will fix the problem but I wanted to know if we should expect this compressor to work in our aplication.

 

[LittleInch]

Your issue sounds much more like a control system issue than a mechanical compressor issue. The only thing that might make a difference is the mas of the gas compressor, but given this is a fixed speed unit then whether the compressor is designed for pipeline use or otherwise has no difference.

If you're running two gas compressors and two frame 7 GTs why don't you trip a gas compressor as soon as one of your GT's trips??

George is correct though that the real cause s simply too little buffer volume. Normally this is an issue where you have a pressure control valve reducing fuel gas pressure very close to the consumer. the same thing happens on trip or sometimes start-p where the variable demand causes the control system to struggle. Too much volatility and the system never gets to steady state, too little and you trip or relieve on rapid change of flow. Even if you simply oversize the pipe by as much as possible, this can be enough to get you the time for the system to react.

The margins you are running at are also relatively tight. Is there anything you can do to alter the normal pressure or the short term over pressure limits to buy a bit more time for the sluggish control system to work?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

The guide vanes will help with hydraulic eff and some limited throttling.

Is this recycle PCV a fail open valve? See if a quick exhaust pneumatic relay on the PCV will help it to open faster( for a fail open valve).

There is only so much you can do to zap up the PIC / PCV - the buffer volume is really key to getting this discharge pressure to stabilise. A less elegant solution would be to add / convert a blowoff continous modulating PIC loop near the power gen GT (rather than this gap acting vent valve you have now).

We note you are using the anti surge valve as a pressure recycle valve also - that complicates controls - another solution would be to separate these two functions - and have the discharge PIC operate a dedicated recycle PCV. You then have more freedom to tune this capacity / pressure recycle loop as you need without worrying how this may affect the antisurge protection.

 

[bohdand]

Thank you all for the feedback and suggestions.