Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Generator tripping while paralleling 4

Status
Not open for further replies.

Ulterior24

Electrical
Mar 28, 2023
4
0
0
IN
Hi, I have three identical generators of 80 KW each. The generator have electronic governor and an AVR. Generator 1 can be paralleled with generator 3. Also generator 2 is running well when paralleled with generator 3. But when we parallel generator one with generator 3 the power factor is hunting and generator 1 gets offloaded. Unable to understand the issue
 
Replies continue below

Recommended for you

Can you confirm which parallel configuration you are having trouble with? Is it generator 1 and generator 3? It’s a bit confusing in your post.
I personally would look for the settings in the AVR for reactive droop compensation or reactive load sharing features. If you have two or more units on the same bus with minimal impedance between them the AVR functions will fight each other unless you have some sort of droop or reactive load sharing. Alternatively you can place one machine in manual regulator mode as a test, but leaving it in manual has addition operational issues to contend with.
I’ve set droop in machines at 5% with good results, but these are grid connected machines and 5% would likely be too large for good local voltage regulation. 1% would probably be more appropriate.
 
Two issues.
1. The universal setting for islanded generator droop in that size is 3%.
Grid connected generators the droop is 5%. This is often an interconnect requirement.
2. Quadrature circuits.
The quadrature circuit interconnects the AVRs so as to properly share reactive current.
This requires an additional CT and a quadrature dropping resistor.
It is common for AVRs intended for parallel operation to have the quadrature resistor build in and to have a terminal for the connection of the CT.
Common also is a panel mounted switch to select Stand Alone or Paralleled.
And, by the way, don't try isochronous control.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!
 
Ooops, maybe it wasn’t clear, I was referring to the reactive droop compensation function in the automatic voltage regulator, not the governor droop. It’s a software feature used in place of the quadrature dropping circuit you referenced I believe.

As an aside, the Governor droop on those particular units I mentioned (two 50MW hydro units) is 3%, which we drop to 1% for islanding.
 
Ooops, maybe it wasn’t clear, I was referring to the reactive droop compensation function in the automatic voltage regulator, not the governor droop.
Is that a stand alone feature or does it depend on an input from a CT?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!
 
It a functionality built into the regulator as all the regulators I’ve installed always have 3 phase PT and CT sensing, as all the machines I’ve worked on need Power System Stabilizers functionality (I’m in the Western US Interconnection where PSS is required for all synchronous machines greater than 30 MVA per NERC VAR 501 WECC). I’ve tested and did troubleshooting on a few other types, but I’ve only installed Basler DECS 250N, 400, and 450 regulators and they all have 3 phase voltage and current sensing.
 
At that size there's multiple ways of doing load and var sharing.
Once such way is as waross has stated, Quadrature sharing, another is cross current compensation (which also requires a separate CT), the third way with modern controllers is to let them do the VAr sharing, no CT required but each unit needs it's AVR bias connected and set properly.

Since there's no indication on which particular scheme is in place, it's hard to suggest what troubleshooting options might be available.
Initial sites I did were done with controllers, later sites all used cross current compensation and all sorts of fun in making sure the appropriate number of CTs were switched in and out for the number of sets running.





EDMS Australia
 
OP said:
Hi, I have three identical generators of 80 KW each.
At that size, forget three phase PTs an CTs.
The AVR will most likely be single phase sensing.
That makes reactive compensation easier.
The basic compensation is voltage sensing across phase A to phase B.
Compensating CT on phase C.
A quadrature ready AVR will have a two sense terminals and a compensation terminal.
There is a resistor between one sense terminal and the compensation terminal.
The sensed voltage flows through the resistor.
The CT is connected across the resistor.
Any real current from C phase will be at 90 degrees to the voltage sensed across A and B phases.
The voltage drop caused by the CT current will have negligible effect on the sense voltage.
However, reactive current will develop a voltage that is in phase with the sensed voltage and will either add or subtract from the voltage seen by the AVR.
If one CT is connected in reverse, you may expect issues when the power factor is other than unity.
A basic AVR may be used for reactive compensation with the addition of an external resistor and a CT.
It is common to have a switch that shorts out the CT when the unit is running alone.
Sets in that size range will have 3% droop.
Set your no-load frequency at 51.5 Hz or 61.8 Hz.
Freddy: I am under the impression that "cross current compensation" is another name for "quadrature compensation".
I may be wrong.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!
 
waross, you are correct that they do the same thing, it appears the difference is quadrature droop allows the voltage to droop as the power factor gets further away from unity, cross current does not.

I've managed to locate the document I was using for reference in the difference between the two schemes and it's attached. The implementation between the two is rather different hence knowing what system is actually installed is critical.

EDMS Australia
 
 https://files.engineering.com/getfile.aspx?folder=7190c0d7-2241-44de-960b-fe3dd7dc6358&file=5_Voltage_Regulator_and_Parallel_Operation.pdf
Status
Not open for further replies.
Back
Top