Generator power factor fall... 5

[loadangle]

just after synchronizing our power plant's turbo generator with grid when load gradually incresed & we satarted exporting,PF was near 0.86 lag & everyting seemed stabilize, suddenly we noticed sharp fall in generator power factor & rise in generator amps.we have steam turbine,brushless excitation system(newly installed dual channel AVR abb make replacing single channel AVR).power factor dropped sharply & as a result gen amps increased.as soon as we manually tripped the grid breaker the gen amps came to normal at home load.we have electronic governor for turbine.I just want to know the probable causes of this sort of behavior of TG after synchronization with grid.

 

[ScottyUK]

If you're in islanded mode then the chances are your AVR is set to regulate the terminal voltage. If it is in voltage regulating mode when connected to a large system then unless the generator voltage matches the system voltage then there will be a large reactive power flow. You need to change the AVR operating mode to allow operation in parallel with a large system - probably by keeping the generator at a constant power factor, although other options exist - or if the system voltage is stable you could flex the machine terminal voltage setpoint to adjust the Var flow.

 

[loadangle]

our AVR alaways stays in PF mode when connected to grid,normally operating the gen at 0.90 lag, when we are connected to grid

 

[davidbeach]

I have no doubt that you believe your AVR is in PF mode, it may even be telling you that it is in PF mode; but the behavior you describe is not PF control. One way or another it is trying to regulate voltage.

Which way did the vars go? Into the machine or out to the system? Either would increase the current for a constant real power output.

Is there a voltage range outside of which it automatically leaves PF mode and enters some form or voltage control mode? What were the system voltages like? What are the system voltages like when the AVR behaves itself?

 

[Hoxton]

David Beach beat me to it!

We assume that everyting worked correctly, until you changed the single channel AVR to a dual channel.............

The AVR will be in voltage control mode from the point when excitation is applied (switched on) during the turbine start sequence UNTIL the point of synchronisation with the grid when the AVR should be switched to Power Factor Control mode.

Assuming that your synchronising breaker is the generator breaker, then there will be auxiliary contacts on this breaker that are used to switch an input to the AVR Power Factor Control module (this is probably part of your AVR unit, in the old days it was often a separate item.)

If this connection fails for any reason, or if the AVR is not set up during commissioning to accept this input, then the AVR will remain in voltage control mode. Assuming that the grid voltage is lower than the AVR voltage set point, then the AVR and generator will increase excitation and thus generator current in an attempt to raise the grid voltage.

So you will see a rise in generator current and a strange, usually low power factor.

If the generator is much smaller than the grid capacity, then it will not be able to raise the grid voltage.

If the grid voltage is high, then you will see the opposite effect, a low generator current, or a high reverse current, as David Beach says.

I agree with David Beach’s recommendation, but I notice that you have changed from a single to a dual channel AVR. This may be the cause of the problem, as the breaker auxiliary contact is not connected, or there may be some AVR setting required after the wiring change to accommodate the AVR change.

As David Beach says it may be telling you that it is in PF control, but it is not!

So check your wiring from the breaker auxiliary contact to the “power factor control enable” input to the AVR. This may have another name like “breaker status”.

If you can rack the breaker out into “auxiliary mode”, i.e. with the breaker disconnected from the busbars, but with the auxiliary contacts engaged, then operate the breaker to closed position and see if the AVR is switched from voltage control to power factor control.

There may be an indicator on the AVR to show which control mode it is in.

You say “as soon as we manually tripped the grid breaker the gen amps came to normal at home load”. This makes me think that you have some local / home load and the grid connection is a separate breaker to the generator breaker. So your control system to enable the AVR to know whether the grid is present may be a little complicated, as you may have two possible points where you can synchronise, or more likely a grid breaker that you do not close if the generator is already running in island mode.

You may have some form of interlock to prevent incorrect breaker sequence

Do this all with care! We can only give you guidance, as we do not know how your plant is connected.

 

[loadangle]

thank you so much all of you, for your kind guidence.yes after meeting the plant load extra power is pushed into grid.in this situation grid & generator breaker in closed position as you mentioned.but all you said is true.contact of grid bkr has been used for avr for letting it know that grid bkr is closed after synchronization.Actually sir let me get into more details,every time after we synch. with grid generation PF normally stays arround 0.70lag from there load is increased by our DCS operator & gen pf reaches 0.90 lag & then we push 'pf mode on switch(with led indication)' on avr panel & avr controls the pf.Now at one fine morning after usual synchronization we ware just waiting to reach tahe gen pf to 0.90 from 0.70 lag,everything seemed calm,but maybe 2 mnts after we noticed that sharp fall of PF & rise in amp which is a first time experience to us & generation voltage was at bit higher side also(around 11.10kV).Normally after synch. we dont expeperience this kind of PF falling isseue.as you are saying there may be reactive power flow but it was may be going to grid otherwise directional relays would come into action which didnt happen.Is there any chance of faulty governor action,or problem in gen bkr itself or problem in grid that could cause that behavior in generator? after that incident we again synch our system to grid without any hitch.

 

[Hoxton]

Let us think about this. I have a feeling that David Beach is on the right track

Where in the world are you?
Scotty and I are in the uk England

 

[ScottyUK]

Do you have any trends from the plant automation system which show real and reactive power flows, breaker statuses, etc? They would be immensely helpful. Do you have a single line diagram of your system showing your AVR sensing locations, any major plant such as a GSU transformer, that sort of detail?

My first guesses would be you have a bad status signal driving the AVR, or a signal which isn't doing exactly what you think it's doing, or an AVR which is not responding to that signal exactly the way you expect it to. Daft thought: you do definitely have the right breaker drving the PF select input of the AVR (i.e. the grid tie, not the generator main output?

 

[davidbeach]

This seems to me to be the type of problem that will be rather difficult to sort out over the internet but could be relatively quick and easy for someone on site who is familiar with the equipment. AVRs aren't my thing, but I've had long discussions with folks in the field about relay issues that become nearly trivial to sort out once the right info from the relay becomes available. The easiest way to make the right info available is to be directly connected to the relay. I don't see why the AVR would be much different.

 

[loadangle]

sir I am from India.no reactive power flow status available in our DCS.trends of power,frequency,turbine speed,voltage,current,pf are available.the falling of Pf didnt start instantly after synch,it took few time then suddenly fell.the machine behaved like this for the first time in the same condition thats why m worried.

 

[ScottyUK]

Do you have an on-load tapchanger in the area, either yours or belonging to the utility? The change in power flow may have caused the OLTC to operate with a resultant effect on reactive power flow. In larger generating stations it is common to use the GSU transformer's OLTC to flex the reactive load on the machine to meet the reactive despatch order.

 

[redlinej]

Trend the voltage at the time the PF start falling and see how low did the voltage drop and also look at the point of the lowest PF was is the voltage.

The ABB AVR is it a UNITROL 1000.

 

[davidbeach]

Can you get field current? Seems to me that the first thing you need to know is whether you're going over- or under-excited.

 

[redlinej]

It seem to me that the genset was over-exited because he said that the PF fall.If I am interpreting PF falling correctly, this means that the genset is supply VAR to the grid (over-excited).

 

[Hoxton]

So the normal start sequence in brief is:

1. Start turbine
2. Synchronise to grid
3. Power factor normally 0.70
4. Operator increases turbine power manually
5. Power factor reaches 0.90
6. Operator switches AVR to power factor control
7. AVR controls power factor to 0.90

So from your post, one time during stage 4, the power factor reduced and the generator current increased? This was during the period when the AVR was in voltage control?

Your generator voltage was also a little high before synchronising? Your post says this, but we want to be sure. This is because there must be a reason why the AVR is controlling the voltage to a higher value than usual, when the generator is not connected to the grid.

Please let us know if these statements are correct.

If the above is correct then I draw either of two conclusions:

A) There is a loose connection in the AVR wiring, possibly in the voltage sensing circuit. The AVR detects a lower voltage than expected, and increases the excitation to restore the voltage to the correct value.

B) The grid voltage reduced during stage 4 and this causes the AVR to increase excitation, increase generator current and thus change the power factor.

I agree with David Beach - these problems are often difficult to solve remotely, by phone or email, you have to be there!

I had a look at the ABB Unitrol details. Like a lot of modern devices it is programmable and very flexible, the downside is that a minor error n settings can result in unexpected events. You also have a twin channel version, so more than twice the complexity.

Someone just sent me one of those Murphys Law lists that engineers delight in sending each other, one seemed most appropriate: inside every large problem is a small problem trying to get out! I think that is what we have here, a loose wire or an incorrect commissioning setting, this simple fault produces a dramatic effect which seems to reveal a complex issue.

Please let us know the answers to my questions above.

 

[loadangle]

Again thanks all of you for analysis.I am triyig to answer all the questions mentioned above in this segment.
1)YES WE HAVE AN OLTC OF GENERATOR-TRANSFORMER UNIT OF 15MVA(11/66kV)LOOP-IN-LOOP-OUT.but the RTCC panel is located near switch-yard at 150 meters distance from power plant.we dont maintain any reactive power despatch,but we maintain the primary voltage of x'mer between 11.06 kV to 10.86 kV by using controller which signals oltc.But during every desynchronization we operate a bypass toggle switch to bypass controller & after synchronization we take controller in line just to maintain the grid voltage during synchronization.
2)Voltage didn't drop,it was on the higher side.the exact figure I will give you in next post surely,but at lowest pf voltage was high & we tried to decrease the voltage from synch panel but pf kept falling & gen amps was shooting high within just three minutes after synch. the AVR is UNITROL 1020 dual channel replacing old unitrol 1000.
3) No sir I dont have any records in of field current that point of time as that problem came suddenly & gave us less than 12 sec to take decision to manually trip the grid bkr & save gen trip on over current. & I am not sure, if unitrol 1020 records this sort of phenomenon or not.
4)Mr. Hoxton you have done a nice sketch,i must say.kindly follow my sequence in your sketch mentioned below,
i)turbine rolling ii) turbine attains rated 6280 RPM iii)on the excitation from avr & buildup voltage till 11kV iv)then synch with grid v)when gen pf reach 0.90 lag then on PF MODE at avr panel vi) take oltc controller of gen-x'mer unit in line to maintain export voltage & PF.
in above seq. we reached the stage (iv),u also mentioned that,PF 0.70 & VOLTAGE OF GEN DURING SYNCH WAS NORMAL 10.96kV.Now our operator was increasing the load.
& pf was 0.84 lag till now every thing was normal & as a usual practice(written in our ISO work instruction of synchronization or SOP) we were waiting to reach it to 0.90 lag.NOW the DRAMA started.
we got merely 10 sec. to think & take action on what actually was happening.all happened so quickly & our DCS operator was screaming & we were stunned to see this.BECAUSE EARLIER WE DIDN'T FACE THIS TYPE OF PROBLEM USING OUR STANDARD OPERATING PROCEDURE OF SYNCHRONIZATION.EVEN AFTER THAT INCIDENT DESYNCH HAPPENED & WE SYNCHRONIZED IT WITHOUT ANY ABOVE SAID PROBLEM.
m quite convinced with the thought of VAR flow thing towards grid due to grid voltage disturbance because that day when desynch happened GRID voltage & frequency was unstable a bit,but really wanted to know all probable causes.As I mentioned in earlier post the grid disturbance & woodward 505 governor & even VCB of gen bkr are my suspects also.Kindly light my thought on how grid disturbance can cause above problem ,coz that dat during synch we realized grid is more unstable than usual.Thanks a lot & merry christmas to all.

 

[rcw retired EE]

VARS flow from high voltage to low voltage. Generator MVAR output is dependent only on the relative voltage between the system and the generator. MW output has almost no effect on MVAR output. Generator VAR output increases if the generator voltage is raised or utility voltage is lowered. A system disturbance that drops system voltage will suck VARS out of the generator unless the generator also lowers its voltage in response to commands from a constant VAR control or a Power Factor control. But that response creates system problems.

Running generators in power factor control is prohibited in most areas of North America because the control reduces voltage & VAR output during a low voltage disturbance making the situation worse. A large US West Coast blackout was caused in part by multiple generating plants running in PF control and reducing generator VAR output until the system collapsed.

You might check your AVR droop settings and CT polarity into the AVR. That is the function that reduces generator voltage setpoint proportional to VAR output. If VARS increase, the voltage is reduced. Depending on your set up, it may not be active in power factor control.

I never liked power factor control, because we are trying to control a complex ratio of two variables, watts and vars with a single output that only affects vars. Any change in MW, VARs or system voltage requires a response from the PF controller to maintain the MW/MVA ratio.

In my experience, the power factor control is usually an outer feedback loop that automatically adjusts the Voltage Regulator's set point. The AVR response time is typically faster than the power factor control for stability reasons. If there was a sudden, large system voltage drop, it is possible that PF control response was too slow and the AVR tried to maintain the setpoint causing VARS to flow out to the system while MW output stayed the same. PF would drop since pf = MW/SQRT(MW^2 +MVAR^2).

In my experience, the power factor control is usually an outer feedback loop that automatically adjusts the Voltage Regulator's set point. The AVR resposne time is typically faster than the power factor control for stability reasons. If there was a sudden, large system voltage drop, it is possible that PF control response was too slow and the AVR tried to maintain the setpoint casuing VARS to flow out to the system while MW output stayed the same. PF would drop since pf = MW/SQRT(MW^2 +MVAR^2).