Generator Syncronisation

[ruggedscot]

Wonder if anyone has any thoughts on this one....

Two generator sets are used together to support a critical load. These are set so that both run together and when the utility supply fails the generators start and supply the load. The main load itself is protected by UPS units and there are other non critical loads supplied aswell.

Now this is the interesting bit, The generators seem to be having a loading issue as when both are running one shows a kW rating that is roughly equal to the other set. The other however has kVAr and although the kW is the same the amps differ by quite a considerable bit - even more than the load share set up.

I dont think that this is a normal state to operate in.

My initial thoughts are that the reactive power is being supplied by one set and the other is supplying the unity load. The reactive one also supplying some of the unity load at the same time. Hence the differing currents but similar Kw readings.

Should I be worried about this and if so what would the suggestions be that I should start to look at

 

[dpc]

http://www.basler.com/html/dwnapps.htm#Excitation

Go to Basler website and download Application Note 126 on reactive power sharing.

There should be cross current compensation CTs on each generator to help keep the reactive power equal. A common problem is the polarity of the this CT.

The governor droop settings will control the real power sharing, and the voltage regulator controls the reactive power sharing.

 

[powerwagon75]

Load sharing line polarity reversed?

 

[waross]

Your governors should both be in droop mode.
Turn your parallel switches off.
Balance your kW with the governor controls.
Balance your reactive currents with the voltage adjustments.
Turn your parallel switches on.
If the reactive current does not stay balanced check your quadrature circuits. (Cross compensation circuits)

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

 

[Colinjohnson]

Ruggedscot,

I agree with dpc and others suggest. Check that CT and let us know what happens,

Regards

 

[catserveng]

The CT suggestion is a good one, and a cause of many problems like yours.

For better suggestions you might need to provide more info.

What kind of load and VAR share controls do you have? Isochronus load share like Woodward, Power Command, Powerlynx, etc? Or is load share in droop as Bill suggests?

What was design intent of the VAR share system? Droop, cross current compensation, or active VAR share control?

In percentage of rated current, what is current differences you are seeing?

Is one of the units going into as leading power factor?

Did the system ever work properly? If so, was change the result of some recent work?

Lots of different systems out there, depending on cost, application, and design, the problem could be minor or an indication of deeper problems.

 

[hasi]

The real power is adujusted with the speed governor and the reactive power can be adjusted by varying the voltage of the alternator. Kindly check if the exhisting voltage regulator has quadrature droop control facility inbuilt. If it does not have replace the voltage regulators with the ones having this facility.
If the voltageregulators have this facility follow these guidelines:
1.Start Set No. 1, adjust it's voltage to the rated level.
Load the set with the rated load. Adjust tha QDC pot so that the voltage drops by 2 percent. If the voltage is increasing [instead of decreasing] on turning the QDC pot, reverse the polarity of the QDC CT.
2. Now stop the Set No. 1 and start the Set No. 2 and repeat the step no. 1.
3.Now syncronize The set no. 1 with it and check the ammeter readings. And if fine tuning is still required, yuo can do it with varying the voltage of set no.1
[Increasing the voltage will increase reactive load on it and vice - versa]

 

[ruggedscot]

Hi

Just to say thanks - were currently working on this one - The load itself is a bit of an issue we can take chances so will be seeing about doing it with a dummy bank (inductive I know...) and see if it can be bottomed out and the load properly shared again.

Can say that I value all the answers and hope that we will work this one out quite quickly.

Rugged