Hi All,

Firstly, i would like to thank all members and moderators for this wonderful and amazing site over here. it provides so much help and information.

Problem:

we have a combined cycle module composed of 2XGE Frame 9FA+E GTs + 1 Alstom ST (Overall power is about 750 MW). few days ago we had a very abnormal behavior of both GTs.

what happened is that both GTs were running at base load about 220 MW each at rated 50 HZ and suddenly the grid frequency dropped down to 48.8 HZ (2932 rpm) and here is the recorded responses:

1. GTs MW increased to 309 MW in one unit and 310 MW in the other for 10 seconds.

2. ST MW increased about 20 MW from base load (Steam control valves were 100% open (Sliding pressure control)).

3. Exhaust temperature increased about 4 F in both GTs.

4. Exhaust mass flow decreased about 30 lbm/sec in both units.

5. during the 10 seconds MW transient Mk VI had an alarm "Backup exhaust temperature reference selected" this alarm was toggling true and false.

6. after 10 sec, MW decreased automatically till back to normal base load values.

7. No recorded alarms from G60 relays to MK VI (to be clear, no one ever tried to log in the relay and check the internal event list yet)

8. control team at site did not backup the Mk VI trip_log live files during the incident and hence we do not have any MK VI trends only we are depending on Historian trends.

9. currently i am away from site, but they confirmed that as historian trend do show the transient increase in MW but never showing any increase in FSR/SRV/GCVs feedbacks.

Technical information:
GT is GE frame 9FA+E, dual fuel
DLN2.0+
Fuel used during incident was natural gas
Control system is MK VI TMR
Rated speed is 3000 rpm (50 HZ)
GTs are running for 4 Years now

one theory i was thinking about is that, as units were at base load so FSRT was active --> when grid frequency dropped --> CPD decreased --> TTRXP increased --> FSRT increased --> more fuel injected into the turbine --> more MW.

but the drawbacks of this theory is that, Trend is not showing any FSR changes and even if the resolution of the trend was not high enough to record this, increase in MW can not be that high (almost 40 % of rated MW) and if this is true why backup ref control came into action.

second theory, there was some kind of electrical fault in the grid that caused High transient current in Generator terminals and hence transient increase in MW.

but if this is true, why G60 did not announce any kind of faults/alarms or even trip to MK VI. and also, why ST did not had the same increase in MW as the three units are connected to the same grid and should see the same fault.

i will try to upload screen shots for trends.

and in the mean time, i wish if anyone have any theories about what could actually happened to help us understand it.

and am appreciating all replies and comments in advance.

Thanks.

 

Gemi... what were 3-phase pre-fault, during-fault and post-fault electrical parameters, specifically terminal-voltage, line-current, and pf of the three generators.

This data can point to type of fault!

Regards, Phil Corso

 

Hi Phil,

For Gas turbines:
MWATT 216 before --> 310 MW during --> 215 MW after
MVAR was 87 MVAR before --> 116 MVAR during --> back to 90 MVAr after
Voltage was 15.2 KV before --> 15.2 KV during --> 15.2 KV after (almost no change)
Current 9 KA before --> 12.8 KA during --> 9.5 KA after
PF 0.92 before --> 0.93 during --> 0.92 after

For steam Turbine:
MWATT 264 MW before --> 283 MW during --> 260 MW after
Current 9600 A before --> 10000 A (this is the transmitter full scale value so we believe it actually increased over this value) --> 9600 A after

Also, for Gas Turbines:
FSR 63 % after --> 63 % during
CPD 197 Psi --> 196 Psi during
exhaust mass flow 1298 lb/sec before --> 1283 lb/sec during
speed 2994 rpm before --> 2945 rpm during
SRV 49 % before --> 49% during
also spreads were almost constant with no change

thanks and best regards

 

Gemi... thanks for quick reply. A few more questions:

1) Are there 3 GSUs? Or are all three Gens connected to a single GSU Xfmr?

2) Is there significant impedance between gens and GSU(s)?

3) Are GTGs 2-pole (3000rpm), and STG 4-pole (1500rpm)?

4) If kV, kA, and pf readings are averages, can you provide individual phase readings?

5) Did fault-event last full 10 seconds?

6) Do you know if System Stability Study was carried out?

Phil

 

Gemi,

please keep in mind that your machines like all the rest have inertia. Thus when the frequency fell from nominal 50Hz to the 48.8Hz recorded, your generator assemblies were actually braked to lower the speed. Obviously this energy has to go somewhere, and this is through the generator terminals. This will show as an increase in generator output WITHOUT a corresponding increase in fuel input. Keep in mind that the GT mass contributed to part of this energy and the generator rotor to the rest, contribution directly related to the respective moment of inertia, so through the coupling between GT and generator only GT contribution passed.

Now whether the generator can supply the additional power for the time you indicated or less is another question.

The above also holds true (and is proved by the behaviour) for the steam turbine. Given its governor valves were fully open, its governing system could not react to the frequency change. The change in boiler output, due to the slightly different exhaust conditions of the GTs, would not have occurred yet, due to the thermal inertia of the boilers. Thus the change in output of the ST was only caused by the same phenomenon, i.e. braking of the generator by grid thus availing of the rotor's kinetic energy.

If you manage (given you stated that the mark VI logs have been erased) try to plot frequency, power output and fuel flow during the event and see what comes out.

 

Thanks to you Phil for your interest, we really appreciate it.

and now regarding your questions:

> 1) Are there 3 GSUs? Or are all three Gens connected to a single GSU Xfmr?

This module is configured to have 3 GSUs.

> 2) Is there significant impedance between gens and GSU(s)?

NO, each generator is directly connected to GSU thru Iso phase bus and only one Auxiliary xfmr is connected to each bus (so we do have three aux xfmrs)

> 3) Are GTGs 2-pole (3000rpm), and STG 4-pole (1500rpm)?

No, all three generators are 2 pole/ 3000 rpm.

> 4) If kV, kA, and pf readings are averages, can you provide individual phase readings?

Actually no, as only trend provide the avg values.

> 5) Did fault-event last full 10 seconds?

Yes, this is correct for the three units.

> 6) Do you know if System Stability Study was carried out?

as far as i know, No.

also, we did further review for the ST trends as previously our focus was on GTs behavior but later we deeply studied ST. and we found out something abnormal also where, just after the speed dropped down, the HP bypass valve opened and so IP pressure increased and as a result IP Bypass valve also opened. the abnormal about this is that the settings for HPBP VV is 130 Bar and the actual pressure as per the trends was only 115 Bar which is lower than the settings. an investigation is carried out now to figure out why HPBP vv did open.

steam flow was constant all over the transient period (about 142 Kg/sec).

Thanks again.

 

Hi Jojo,

Thanks for participating in this thread.

actually, your analysis is making much sense. but if this increase is related to inertia of the Gen rotor should not ST had much contribution i mean much more MW increase as the steam rotor is much heavier than of GT's one. aslo, could this inertia result an increase up to 100 MW in each GT?? i really do not know if this is consider normal or not as we did not have this issue before.

and we already get a historian trend, and as Frequency drooped down --> MW increased --> Fuel flow/GCVs/SRV/Gas fuel pressure had almost no change. the same also for ST, Steam flow was not changed (only a 0.0x KG/sec portions)

Thanks.

 

Gemi... following are my best Guesstimates regarding the two questions you posted, (1) cause of disturbance; and (2) unexpected generator responses:

(1) Cause of Disturbance

It is my thesis that the disturbance was indicative of a switching error within the grid network! The salient indicators are:

a) GTG: current increase, 1.5 pu; MVAr increase, 1.3 pu; pf, no change; kV, no change!

b) STG: current increase, 1.1 pu; MVAr increase, nil; pf, no change; kV, no change!

c) Event duration: relatively long compared to normal clearing-time of an electrical-fault!

(2) Unexpected Response of GTGs and STGs

I strongly agree with Jojo's logic, i.e., the difference between the GTG and STG response is due to different Inertia constants and their respective braking criteria, i.e., Pe > Pm, which is 43% and 7% for the GTGs and STG respectively:

a) Although they remain synchronized during the transition from 50.0 Hz to 48.8 Hz, their torque-angle (also referred to as power-angle) excursions are very dependent on their Inertia constants.

b) Regarding the Inertia Constant, I believe the GTG's is considerably smaller than that of the STG.

c) Please note, this is not a true case of 3 generators connected to an infinite-grid. The difference is that the 3 GSUs represent a relatively large impedance (compared to the gen impedance) between the GTG-1, GTG-2, and STG sources. Had they been all connected at 15 kV, the result could be considerably different.

The above conclusions were based on the following estimates regarding equipment sizes:

a) GTG (ea) rating: 275 MVA; 220 MW; 156 MVAr; 0.8 pf; 15.0 kV; 8.8 kA; 3,000 rpm.

b) STG rating: (750-2x220=310MW); 390 MVA; 310 MW; 235 MVAr; 0.8 pf; 15.0 kV; 15.0 kA; 3,000 rpm.

c) GSU's MVA approx 25% greater than their connected Generator.

If you find there are substantial differences with actual parameters, please let me know!

Regards, Phil

 

Hi Gemi,

thank you for your comments.

Well actually your answer got me a bit confused (jokingly).

On the one hand you are stating that you find it strange that you manage to get 100MW additional output from your generators just from inertia (which in part I don't blame you as it is a hefty amount of energy), but on the other hand you state that during the period you had these additional 100MW output, there was no increase in fuel flow (or in the case of the steam turbine, steam flow), so from where did the energy come from?

Keep in mind that the power output due to inertia depends on the rate of fall of frequency, and thus the braking effect on the rotating masses. Do you know what was the rate of fall of frequency during the incident? Did the frequency level off before the end of the incident i.e. was the 48.8Hz you indicated previously a steady reading or a changing reading?

I don't know what type/make of steam turbine you have, but are you sure there is such a big difference between the moment of inertia of the steam turbine rotor and the gas turbine rotor?

One final question, what is the scanning rate of the instrumentation from where you are picking the data? If it is a DCS, there could have been some 'overloading' on the system if there was an increase in events being picked up by the DCS due to the disturbance. Can you check this out. Are the 10 secs really the actual duration of the disturbance?

 

Hi Phil, Jojo,

Just for clarification purposes, actual configurations are:
[1] GTG:
Turbine rated Power is 258 MW.
Generator: 375 MVA, 0.85 pf, 15.8 KV, 2 Pole, 3000 rpm

[2] STG:
Turbine rated power is 280 MW
Generator: 375 MVA, 0.85 pf, 18 KV, 2 Pole, 3000 rpm

and regarding Jojo comments:
i thought that as steam turbine is much heavier than Gas turbine then the moment of inertia of its rotor sure is more than of a gas turbine.

also, we are using MK VIe DCS with historian resolution of 480 mSec so, i believe it is enough for such diagnostics.

am in instrumentation and controls division, so i have a little bit difficulty to understand the "INERTIA" of the rotor and its effect during such abnormal transient conditions.

Please, could you kindly spend few moments to explain more what actually happens in a typical synchronous generator when connected to the grid and frequency dropped so badly as in our case.

Really appreciating your interest and kindly support.

Thanks and best regards

 

Gemi... responding to your 18-Aug-10 (20:24) comments:

1. I believe that the grid's frequency excursion was not instantaneous, but instead contained two time periods: the first, a decreasing magnitude having a negative slope; and the second, an increasing magnitude having a positive slope for the remainder of the incident. The total duration being 10 seconds!

2. What I observed to be unusual about your dip in frequency incident is that for the duration given it lacks the expected load characteristic response, i.e., a proportional decrease in MW, accompanied by an almost proportional increase in inductive MVAr!

3. Can you provide some info about the grid? Is it at least 10 times as large as your plant's contribution?

4. I don't agree that a Historian with a 1/2 sec resolution is adequate!

5. The Inertia-Constant, H, is related to the Moment-of-Inertia of not only the generator rotor, but all rotating assemblies connected to it, such as the prime mover, an exciter if shaft-mounted, gear-box, etc!

6. A steam-turbine drive has a an Inertia-Constant larger than a gas-turbine driven one!

7. If you would like references on Power System Stability Analysis contact me off-forum.

Regards, Phil Corso

 

Gemi... I failed to include the most salient point in my earlier response to your 18-Aug-10 (20:24):

8. Because frequency changed, the system your plant-module is connected to fails to meet the definition of an "infinite" grid or bus!

Phil

 

Replying to Gemi's request to have the concept of Inertia explained, I will try to do it in a simple way:

Imagine you are driving a car on a level road at a constant speed, so your gas pedal is pressed just the right amount. This feeds the required fuel to the engine to overcome air drag loss, friction, etc. You are in a dynamically balanced situation. Energy in from fuel is energy out as losses.

Imagine now that you reach a hill, BUT you do NOT change the position of your gas pedal so you did not increase fuel flow, so energy going into the engine is the same. Your car will start to slow down. Why? Now you have added an additional energy loss in your car in the form of having to go uphill, which is not compensated by energy coming from the fuel. So the car starts consuming its kinetic energy (i.e. the energy coming from the inertia of the car) and the car slows down.

The same situation occurred to your power plant. The generators started to slow down because more load was put on them (increase in MW out), but you did not increase the flow of fuel to the prime mover (GT or ST as the case may be). So the additional energy was taken from the kinetic energy of the rotating masses (inertia) slowing them down.

Hope this gives you an insight of inertia and its effects.

I agree in part with Phil's arguments, however I believe that given your grid is finite (given you got such a frequency excursion), the grid suffered a loss of generation somewhere else, and the lost generation ended up being shared onto the generators that remained on line, until some form of load shedding occured on the grid, which returned the frequency close to nominal. This I am interpreting from your earlier post which stated that GT power went from 216MW to 310MW to 215MW (i.e. frequency went from 50Hz to 48.8Hz to 50Hz). Trip of generators elsewhere occurred in phase 50Hz to 48.8Hz. Load shedding occurred in phase 48.8Hz to 50Hz. Equivalent MVAR shift confirms this.

 

Gemi... Jojo's reply triggered another thought:

Is your plant-module under control of a an "Authorized" Grid Power Flow Control Organization?

Phil

 

Hi Phil, Jojo,

Actually, the grid is large and serving about 80 M people. so i believe it considers an infinite Grid. The problem always occurs during summer period where the load demand increases so badly even over the generations capabilities.

And we also have an authorized dispatch center that controls both generation and distribution over the grid. actually. we had been told that, two of the biggest steam turbines on the grid have been tripped and that caused the sudden frequency decrease.

Jojo..., Thanks for your explain about Inertia. i think somehow i can imagine what happened during the drop.

Thanks.

 

Gemi... Not to belittle the point but a "grid" that is subject to a frequency change of the magnitude you experienced is not considered to be an "infinite" grid.

Regards, Phil Corso