Rapid Output Fluctuations of an ABB-13E2 Gas Turbine

T

Thread Starter

Thomas

The output of an ABB-13E2 gas turbine installed in Malaysia seems to fluctuate randomly by +/- 1.5 MW on short time scales. Output goes from 141 MW to 144 MW and back to 141 MW is about 20 seconds. This is about +/- 1% of output. Can fluctuations this large and rapid be real or are they instrumentation error? If they are real, what is the mostly likely cause? Can grid frequency vary plus/minus 1% on short time scales?
 
B

Bruce Durdle

That depends very much on where you are and how big the connected electrical system is. In New Zealand we have effectively 2 separate electrical systems (1 in each island, connected by a DC link with no frequency-varying behaviour). Loss of generation in the North Island will cause frequency to fall significantly, depending on how much is lost compared to the total connected. The South Island is smaller so the frequency swings are greater. A few years ago, a major cause of problems was due to the DC link getting hiccups and missing a beat - the resulting frequency deviations were up to 2 or 3 % and essentially peaked after about 4 seconds and returned to normal over about 20 seconds. The plant GT load increased from around 50 % to well over full-load as a result.

Our North American and European colleagues have trouble believing that this type of upset can occur, as their systems are massively interconnected. However, I have seen small variations in frequency in the UK, which also gave a measurable change in GT output.

Cheers,
Bruce.
 
Hello Thomas

Fluctuations as you mentioned in your email can be real. For a start, you can trend your MBP41 & 42 control valves, presuming the unit was running on gas and on baseload during the time of the event. If so, we can forget about MBP43 (the pilot CV) at the moment as it should be off during baseload operation, or.. if you are already operating the unit after the switchover point.

Trend these control valves along with the units MW output (the signal MKA10CE601, active power). Make the trend during the time of the event. From the trend, you of course you should be able to see the MW fluctuating as you mentioned. But as at the same time, if any of the control valves contributes to the fluctuation, you can see the control valves fluctuating too.

Do try this as a first recommendation, and see what we can find from this troubleshooting exercise.

Best regards
 
What "mode" are you running the unit in? Base (fixed MW output), droop (aka speed load), or some other style?

Secondly, what is the system frequency doing during these fluctuations? Is it rock solid or moving up/down?

This information matters.
 
I would ask when this problem started. Is it a relatively new phenomenon, or something that has been happening for some time, say, since commissioning?

I would also ask what the grid frequency fluctuations were during the time of the event, and whether the output oscillations corresponded with the frequency oscillations?

Is this a small grid, or on the Malay peninsula where there are many units all synchronized together?

I wouldn't call +/-1.5 MW in 20 seconds very "rapid".

To answer your first questions, they can indeed be real. If the fuel flow is changing then the torque produced by the machine will be changing and that will cause the generator output to change (generators convert torque to amps).

I sometimes find it very difficult to trace control valve position and identify it as the source of load oscillations when these kinds of events occur. If the control valve is oscillating at the same rate, is it the control valve causing the load oscillations or are the load oscillations causing the valve oscillations?

If the grid frequency was oscillating at the same rate, and the unit was operating at Part Load on "straight" droop speed control then the load would go up as the frequency went down, and the load would go down as the frequency went up.

There's too little information about the circumstances of operation to make much of an educated guess about what might be the cause of the problem. Most notably, when this problem started, and what was the grid frequency doing when the load oscillations started.
 
Thanks to everyone for all the great information.

We're testing an inlet fogging system (for power augmentation) at this site. We measured a power boost of about 9 MW but the fluctuations impose a large uncertainty on that number. I suspect that the fluctuations in output are normal but I only have the information for a few hours of one day so I can't say for sure. The turbine OEM says such fluctuations are normal if the turbine is in auto frequency response mode. I'm trying to get more information from the operator about grid frequency, turbine rpm, etc.

The power plant is on the mainland and, according to Wikipedia, Malaysia has large interconnected grid with plenty of excess generation capacity.
 
Hello Thomas

First of, I'd like to correct myself by quoting CSA:
"I sometimes find it very difficult to trace control valve position and identify it as the source of load oscillations when these kinds of events occur. If the control valve is oscillating at the same rate, is it the control valve causing the load oscillations or are the load oscillations causing the valve oscillations?"

I agree with CSA in the sense that, looking at the CV trends and the MW, may not tell us what the root cause of the problem is. Particularly, as mentioned by CSA, we really can't tell actually, if the CV fluctuation is causing the MW fluctuations, or vice versa.

But now that you have provided us with an important piece of information, i.e.: you were running the fogging system when the fluctuations occured, it's best to start your troubleshooting from there. From this point onwards, more information may be required from u.

A few questions i have at the top of my head, after looking at ur reply are:

1. when u started the fogging system, what was your unit's load? was it on baseload? I assume it was on BL when u started the fogging system, cause then, the fogging system can augment the maximum power output of your unit when it's already on BL.

If so, what was your MW setpoint when you turned on your fogging system? Is it more than the actual MW output of the machine, say perhaps 10 or 20 MW more than what your unit's actively producing? Perhaps you do know that standard operating procedure is to set the MW setpoint more than the actual maximum output of the unit, so that your unit can automatically give the extra MW output when grid conditions (frequency for example) allows it.

2. is this power fluctuating problem new? i.e.: did it happened only recently? or, it has been happening for a while? for e.g.: ever since you installed the fogging system, i.e.: e'time you start the fogging system, you will see this same problem, is this the case? and how long have u installed this fogging system anyway? is it a retrofit to your existing GT system? or was it installed along with your GT when it was first installed and constructed?

Assuming this fogging system is a retrofit, and it has been installed in your plant for a while, and e'time you start the fogging system, you see this problem, then i would say perhaps you have to start looking into the operation concept of the fogging system, and understand the operation concept of your unit. Assuming you are also using ABB's product to control your unit (namely Advant Controllers or Procontrol), definitely there are parameters, settings, and of course logics that exist in your controller that controls the fogging operation. On top of that, there are also logic, parameters and settings in the fogging system's PLC that control the fogging system. These settings and parameters particularly, need to be set accordingly and correctly according to the operation concept of the unit. If somehow you do have have access to both of these systems (the GT ctrl system & the fogging system PLC), i would suggest to study/check these parameter & settings so that it conforms to your units operation concept & requirements.

Best regards
 
This is good information.

This is not atypical behaviour for many combustion turbines when operating at "base load." Also, if the unit was operating with a load setpoint, could the governor have been hunting a little trying to maintain the setpoint?

The next question is how does the machine behave before the fogger is in service, and then about half-an-hour after it's shut down? This presumes the unit is operating at "base load" (some kind of exhaust temperature control, and not at part load while on droop speed control).

Does the unit have nearly the same fluctuations (the same period, 20 sec's, but possibly lower excursions, say +/- 1.0 MW, or +/- 0.75 MW)?

If the unit doesn't have the same period or no load fluctuations when the fogger isn't in service, could it be that the increased mass flow through the compressor is affecting the unit?

Is the axial compressor inlet temperature constant during the time the fogger is in service?

How much carryover is there from the fogger?

Are you certain that the fogger pumps are working properly and have stable pressure and flow?

Are you measuring the inlet humidity? Is it steady during fogging?
 
Top