Load swinging

[ciceks]

Hello,
I am currently in a freaky trouble with the unknown reaction of the AVR-Engine Governor-and the actuator.
We are connected to the grid, with 5 step-up transformers, and each transformer is fed by 2 diesel generator(22.4MW) and engines are running in base mode, that means a power measuring relay 4-20mA output is going to governor, the governor also gets the power setpoint from PLC. With these two values, governor stabilise the power output of the engine. There is no load sharing or var sharing between each engines.

When 1 engine is running on 1 step-up there is no problem.
When we started two engines on same step-up, everything is going fine, after a day or 2 days, 2 engines that connected to the same transformer shutdown at the same time with the alarm "4-20mA failure" in governor. The reason is active power relay analog output is under 4 mA in reverse power so the governor determines this situation as 4-20mA. I just decrease the gain parameter of the governor of all engines, 8 engines are running fine but the engine 1 and 2 not. Load is swinging, also the fuel rack, and engine 1 and 2 shutdown.
All of the engines have the same settings and all of them are working fine except the engines 1 and 2.
You can see the picture as attachment;
-#1 for load swing and tripping #2 is excitation

 

[waross]

The reason is active power relay analog output is under 4 mA in reverse power

Have you considered that this component may be failing?

Load is swinging, also the fuel rack, and engine 1 and 2 shutdown.

Are both engines swinging or just one?
If both engines are swinging check the power transducer. Also all connections.
If one engine is swinging check the governor.
Gut feeling; Power transducer going bad or a bad connection in the 4-20 ma wiring.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ciceks]

I have already changed the Governor-power transducer, and wiring also..
Yes both of them is swinging.
Even when you look closer to the picture, swing peaks are going further, instead of compensating.
This is a pole slip? am i missing some thing?
Engines are totally separate, just connected to same transformer.

 

[ScottyUK]

Almost certainly not pole slip. You will feel the machine shaking itself apart if that happens.

The period of the instability is perhaps a bit quick to be governor related on a machine of that size, but the period being dominated by a rotor (field) time constant is at least credible. Do you have a commissioning manual for the AVR? Has this problem existed since Day 1 or has it developed following a period of normal running?

 

[rasevskii]

Obviously if the power transducer to one of the two engines is acting up, the other will also swing. If you have metalclad switchgear with draw out PTs then there is a possibility of the sliding contacts being dirty. A common problem which will drive any kind of fast electronic control system wild.

Try running the unit with maniual excitation if provided. You have to determine if the problem is with the excitation or the governor.

If you give the specifics of your actual system (suppliers) we have experts who can give more exact advice.

rasevskii

 

[ciceks]

It is existing from beginning, first we started engine one by one for 5 days, no problem, than both together and we faced with this.. when both engines are running for 2 days, they trips at a random period, after when we start again they continue 1-2 days, then again same thing happens. There is no period for this issue, or act? it just happens randomly.
Alternator is ALSTOM, AVR is BASLER DECS200 and governor is a useless thing called SEMT engine controller that drives ABB asac 400 via the driver ABB dego2. Active power transducer is Sfere AW1-3.
PT's are not in metal clad switchgear casette, direct installed on phase to earth.

Today i increased a parameter called tacho-gain in dego2 driver on Engine1 and 2, so the tacho feedback effects more PID, i thought this will reduce the actuator speed and it did. Engines were started so the same thing happened 2 times today, in first one only the Engine2 tripped.
In second one, both of them again.s

I can force from PLC, but;
You mean excitation in FCR mode before synchronising? or loading engine to rated power in FCR in PF mode?

 

[rasevskii]

Well, if you have FCR, there is no PF mode actually. It is a controlled Field current only. It is not advisable to run in this mode generally as there is no control of stator voltage. The protections have to be in order, the unit will go to overvoltage if it trips off the busbar while in FCR mode.

The main point is to find out if it is a governor or AVR problem, therefore while in FCR the excittaion should be stable so that any goverkor problem will be seeable. Not at all advisable to run the unit in FCR at high loads, see above...

We will wait a response from our "Catserveng" who is an expert on the DECS and other modern systems.

BTW have you contacted the OEM?

rasevskii

 

[ciceks]

Dear Rasevskii
If i would know the exact time for this issue, i would try, but the thing is i really don't know the failure factor so the exact time.. Maybe i can try to set preposition when CB tripped so FCR maintain to setpoint or stop the excitation. I should do this to both engines, but it will hard to operators to keep PF for more than 1 or 2 days.
I have contacted to the OEM and waiting for reply.

 

[waross]

If this issue has been since new, you may consider verifying all part numbers and settings. We had a supposed rebuilt and load tested gen set that was incredibly unstable. We finally noticed that there was one digit difference between the governor electronic module part number and the part number of an otherwise identical unit that ran well. A quick call to the governor manufacturer. What model engine? What speed? What governor part number?
Sorry. There is no way that that module will run that engine at 1800 RPM.
Probably the correct parts and settings, but it doesn't hurt to double check.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ciceks]

I wish it would wrong parts or settings or both, but they are correct,s
I've totally changed everything with the running well engines, except AVRs. But still same.
Engine mech power can not fluctuate rapidly like in the picture due to moment inertia, i think both AVR and Governor acts to fluctuate PF more. When engine stopped, i will reduce the governor gain as much as i can, to see if the AVR absorb this fluctuate, I wish this will not block a regular synchronisation.s

 

[catserveng]

I'm looking at your graph and have a couple of questions, where did this data come from? The area that seems wrong to me is looking at what I think you have represented as field volts and field amps for both gens, Ifd and Vfd. Usually with an exciter field field volts and field amps follow one another, in your graph for unit #1 they are opposite, so either the data is misrepresented of you may have a problem with the exciter stator or field wiring, like an itermittant open.
Also look at your Ib (B phase current?)and compare to the kvar data, the only way this makes sense (if the data is correct) is if the current spike if current coming into the generator, the peak of the current spike is the same as the lowest point on the kVAR graph,the highest point on the field volts but the lowest ont he field amps. The only time I have seen data like this was on a machine where the field circuit was opening up.

Is there a field output breaker or set of contacts wired, like thru an 86 device or relay? On a DECS there shouldn't, if the field output opens under load the field output of the regulator can have all kinds of problems or fail outright. In my experience the biggest cause of DECS failures, at least that I see. If you need to shut off the excitation, use the STOP input or open the input power circuit, but the excitation disable input is the fastest.

What are you using for the AVR power, a PM, transformer or external supply source? Is it the right voltage for your rated field output. This has been an issue I have also seen quite a bit, usually reconnectable PM's wired for a 120 output on a unit rated at 125 field volts. Recently had a site where two or the units worked fine the other two didn't, all were miswired, all were corrected and no more problems.

I don't think you have a governor issue, based on the info so far and the data on your graph, unless it doesn't represent the actual problem you're seeing. Governor settings can make the problem seem worse, at least in units the size I deal with. Unless your seeing a load transient that is upsetting the governing, which seems unlikely.

I'm assuming your are in Power factor control mode when in parallel, what is your setting? Have you tried moving it closer to unity to see if it reduces the problem? Something else you can try is to put it in VAR control and set it for a low VAR export, if it is a governing issue the tail end is less likely to respond when the output kW changes.

Just to be clear, you essentially are in Baseload parallel to the grid, sharing a GSU with another generator? Likely a real power (kW) problem should affect the other unit, as the grid is there, unless it is a soft system. However, if it is a tail end or AVR problem, it could affect the other unit as it may be trying to supply VAR's if the other unit suddenly starts to consume rather than produce VAR's. We saw this in a number of multiple power module tied to utility grid sites when we were having rotating rectifer issues.

Hope that helps, Mike L.

 

[ciceks]

Dear Catserveng,
First of all Sorry for my special english
Ive downloaded this data from Basler records, and the trends were drawed by Bestview app.

Ive played with genset to observe some data,
-when the active power is decreased at same PF, Vfd is decreasing proportional to active power, and Ifd is increasing.
But i think when you keep the Ifd constant while you are increasing KW, PF should go to unity from lagging, and you need to increase Ifd to reach previous PF according to the Vcurve and Torq-Ifd curve of standard alternators. Why is the Ifd decreasing when you decreasing load? I think it should also be increasing.
-When PF goes unity from lagging at constant power, Vfd and Ifd decreasing at same time,
-When PF goes lagging at stabile power, Vfd and Ifd is increasing at same time.
These are Ok.

Yes Bphase current.

Also there is no KW sharing, or Var compensation between engines, also to cut the excitation, we are using stop cantact, not the actuated breaker for field or stg else.

 

[catserveng]

Yes, I agree that normally field volts and field amps should follow each other. So if field volts are going up and the current drops, the only reason I can think of is that the DECS is trying to drive up the excitation, the voltage goes up, but because the current doesn't appear to be flowing, for what ever reason, the DECS keeps pushing the voltage, then the current starts to flow, then the DECS starts to drive the voltage down. Really looks like either an intermittant open in the field circuit, a possible problem in the exciter stator.

We had a site where we were having some strange failures and operational issues, generator RTD card repeated failures, unexplained instability and finally trips and shutdowns. Turned out to be a failing exciter stator that finally got worse enough to troubleshoot and repair. Once it was repaired all the other problems went away.

Have you tried putting it in VAR control mode and just letting run there for a while? Would be interesting to see how that works and what shows up.

Hope that helps, Mike L