Generator Reactive Power at Minus Values

Because of minus temperature in europe now.

You did not give generator rating. Not possible to know if -2000 kVA is big % of rating. You did not write how long to go -25 to -2000 kVA or -2000 to normal, such information helps. A generator will always go into -ve (leading current) reactive power as engine throttle is opened - unless generator excitation (field current) is increased. Usually, an automatic circuit raises the excitation as the set loads-up - on multiple parallel sets a "Busbar Voltage Control and Reactive Power Sharing" system is commonly used. When a set is stopped, it is often switched to "voltage regulator/AVR" "power factor" control while it is unloading (this ensures that both reactive and active power are reduced to small values before main breaker to busbar is opened - consequently, breaker opens at low current), if stop is cancelled power sharing is reconnected. "...suddenly turns to positive" suggests a fault like a failing relay/contact exists in circuits from stop or start/run or "breaker closed" relays to generator AVR. Not knowing particular model of set or make and type of AVR makes it difficult for other posts, who may have experience of them, to help.

Generator's rating is 8.73 MW it happens during the start the engine. When it synchronaizes with the grid it starts to absorb reactive power and we see on the digital and and analog meters - values. we see -25, -100, -250 .... -1000, -1250 kVars in a 5 -7 minutes until it goes to -2000 kVar. when it gets -2000 kVars.By the way I know the generators absorbs reactive power during the first start but very few times and it is momentary and takes just seconds. And we check all the wires, relays, cables everything even on the generator. we retightend them all as well but the problem is still exist. AVR's model is ABB Unitrol 1000-15

Unitrol 1000-15 seems quite complex and little can be done with it without reading manual!! Simple suggestion is to exchange a 1000-15 from a good machine with that on your problem machine. If problem moves with AVR, then trouble is in AVR (could be software settings, not hardware). If problem stays on problem machine, it is not AVR but the external instructions during start up or a defective input signal. Have you checked that the input signals to the AVR are not fluctuating during period of "wrong" kVAr?? That could indicate a heat or current sensitive fault in external CT/PT/transducer. Is there a facility to manually control voltage during loading - if one can load-up kVAr smoothly in manual, this points to trouble with automatic path?? The basic problem is that the machine voltage is not being raised after connection to bus. Is the AVR or overall control giving error messages like "excitation limiter" operating??

Is this a new installation, or has it been in operation longer time? If existing, has it always done this, or did it just start happening recently?

On what are you reading the kVAR - a panel meter, or other instrument? Analog or digital meter? Where is the source of the meter currents and voltages? Maybe something is mis-wired to that meter, or there is some damage somewhere that is affecting its input. Physically trace its wiring from the CTs and PTs, and compare it to one of the gens whose meter reads normally. You have 11 other ones that are correct - you should be able to find the difference.

it has just started to do this for 2 months we didn't have this kind of problem before. we see the reactive power both analog meter and on the PC at minus values. we did all inspections with the wires,relays etc. but the problem is still same. we read in the first start around - 50,-60 kVAR then it goes to -2000kVAR in five mins, but active power is at normal values just the reactive power value goes to -1500 - -2000kVar when it sees -2000 kVar the reactive power turns to the normal positive values i.e. 1860 kVa, 1950 kVar etc. during the operation

Post #6 suggested swapping the Unitrol on problem unit with that on a good unit (with a spare Unitrol). Did you do this?? Thinking further, I guess AVR control could have 3 phases 1) Voltage control for synchronising 2) Power factor control for Loading-up 3) Load (KVAr) Sharing Control when unit is loading. It is possible command 2) to Unitrol on serial link from Master Control is corrupted, Unitrol ignores it, stays in voltage control, unit progresses into leading VArs as MW increase. When command 3) comes the annomally is corrected. Try re-loading software on problem unit.

I think I m goona swap one of the well Unitrol with this. If I don t get any result, I m gonna dense on software issue. I ll share the results, thank you

Further to post #9. It is possible that set-points, like power factor, are entered into the Unitrol AVR. It should be possible to view them via the Unitrol key pad on problem unit. Then compare values got same way from a good unit. In any case, -2000 KVAr with 9000 kW unit is 0.98 power factor [leading], not a problem - but it is a possible value with a setting of unity [ 1.0 ] power factor.

Today we had a new experince with this machine. When we gave the start it absorbs reactive power and stucks at -2000kVar and active power was 2700 kW.After a while machine tripped and gives fault messages as "Generator Protection underexcitation Q<Trip" and "Generator breaker trip circuit indication" after these breaker is opened.Then we resarted it again. We measured all output signals of the Unitrol with changing the pf values. Everything seems OK(during the machine at normal operation, reactive power at positive values). I read the Unitrol Manual, there is an Kq which is reactive current influence , it has a setting range between -20 to +20, it is set +3 in the AVR. Does it make any sense if we play with this number via increasing or decreasing?

I wrote in post #10 that you must compare AVR parameter value on your problem unit with that on "well" unit. If value is same on "well" unit, there is no benefit from changing - all you do is risk adding to trouble or making a new one to confuse diagnosis. You report you now get "underexcitation" trips - the fault is getting worse. I suggest you swap the problem AVR, after recording its parameter settings and those of "well" unit. Are settings in "Commissioning Report" or "Schedule of Settings"?? If not, best to record what you have BEFORE any change. I tried to find Unitrol manual on BBC website, but only found "sales" brochures, which give poor information. I can only guess it is a "loop" stability parameter, bad idea to change. If you could point me to manual on web, I will try to help - but I think you must now change AVR to make progress!!! Going now, New Year's Day music concert from Vienna [TV] calls!

Further thought, after re-reading your latest post.

1. You write "We measured all output signals of the Unitrol with changing the pf values." - Did you switch Unitrol into Power Factor control to change pf??
2. If you can select PF control or manual excitation current control can you use this [after connection to bus" to get unit up to full load with acceptable power factor, avoiding trip?? This might be a strategy to "keep operating unit when essential" until problem is found and fixed.
3. Usually, a control system will give an alarm (warning) for under-excitation before the trip level is reached. Say "underexcitation limiter working". Did you get an alarm??
4. The only "output signal" of the Unitrol that matters is the DC field current to the generator. Can you measure this with instrument independent of Unitrol??
5. If you can use independent ammeter, how does reading compare with Unitrol or control VDU/screen?? How does it compare with value on "well" machine??
6. Purpose of field current check is to identify if Unitrol's own measurement of field current is bad. Also, to be sure generator itself is OK.
7. Check manual, Unitrol may apply limiter by field current value, but trip by generator output current and voltage values.
   

Please see the generator power factor and adjust field current excitation because,

E = Va + jX*I
Since the magnitude of E is proportional to field current excitation, the leading power factor condition is sometimes referred to as under excited operation and lagging power factor condition is referred to as over excited operation for the synchronous generator.

For the under excited case, the generated reactive power is negative. This means that the generator machine is aborbing reactive power and is effectively operating as an inductor. For the over excited case, the generated reactive power is positive meaning that the machine is supplying reactive power and the machine appears as a capacitor.

we tried this at first but we couldn't get any result