Voltage control / MVAr Mode - Generator Parallel Operation 4

[NickParker]

Three generators are running in parallel in an Islanded system.
I have been told to select one generator as "swing" and other two in "voltage control mode" or "MVAr control mode" in ETAP. I understand the selection of one generator in "swing mode".

In which situation I chose Voltage controlled over MVAr or vice versa for the other two generators?

 

[waross]

Three generators are running in parallel in an Islanded system.
I have been told to select one generator as "swing"

That is textbook advice for a perfect system.
Unfortunately that advice depends on a number of important assumptions that may be erroneous in the real world.
Run all three at 3% droop. The governors will share the kW load in proportion to the various set capacities.
Use quadrature circuits (cross current compensation) to share the VAR loading.
The final choice depends a lot on your expected loads.

By the way, I have been responsible for an islanded plant with three diesel generators.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[FreddyNurk]

waross, whilst you're usually correct, the context here is how to configure the generators in ETAP.
At a guess for an islanded network the other two sets would be in voltage control mode as there's not much point in using VAr control mode.

When the system is actually installed though, either what waross has suggested, or running sharing via controllers (my preference) would be used.

EDMS Australia

 

[xnuke]

I agree with FreddyNurk regarding the ETAP analysis setup (use one as swing and the others in voltage-control mode). There are four parameters of concern in load flow analysis: voltage magnitude, voltage angle, active power, and reactive power. For each mode, two of these are held constant and two are allowed to vary as the nonlinear solver works the problem. This is because the two state variables are voltage magnitude and voltage angle at each bus, and the two input variables are active and reactive power. The swing (slack) generator keeps voltage magnitude and voltage angle constant at the bus, while allowing its active and reactive power to vary. Voltage-control mode (PV mode) is used to keep voltage magnitude constant on the bus and have constant active power flow out of those generators, while allowing bus voltage angle and reactive power output to vary. I think Mvar control mode is used for generators that would have constant active and reactive power flow and allow bus voltage magnitude and voltage angle to vary. This would be useful for modeling sources that can have specific active and reactive power flow commanded via setpoints.

I disagree with waross slightly for what to do in practice depending on the type of generator controllers available. Many modern generator controllers can be networked and will load share and regulate frequency and voltage via network communications, so you wouldn't need to run one isochronous and the rest in droop, or all in droop, and you wouldn't need cross-current compensation for var sharing if that's the case. The method waross stated is used in older control systems where individual generator controllers can't communicate with each other.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[davidbeach]

I guess I’ll third that. In the real world Bill is totally correct. But in the model world you need one swing bus and everything else fixed.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[RRaghunath]

For an islanded system, I think Power Factor control is better suited than the MVAr control or Voltage control modes for the AVR.
In this mode, machine will be able to supply required VARs to the loads and thus maintain voltage in spite of differing load conditions.
MVAr control mode may result in over voltage when the demand for VArs is lower than the MVAr setting of AVR.
Voltage control mode is better suited for grid connected generators.

 

[waross]

The power factor is determined by the load. The generating plant has no control over the load PF.
<Anecdote Alert>
I once witnessed an operator trying to run one of three, paralleled, islanded sets at 100% PF.
He improved the PF by increasing the field slightly. The frequency, voltage and PF were all interacting.
Change one and the others changed.
He would tweak the excitation, (voltage and PF) and then tweak the speed setting.
Repeat multiple times.
He eventually achieved 100% PF on the set.
Then he stepped back and looked at the readings on the other two sets.
Kilo-WAtts, Zero.
Power Factor, Zero.
Amperes, fairly high. Generator theory tells us that the current would have been 100% reactive.
</AA>

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[FreddyNurk]

I agree with waross, the anecdote is correct for what happens in an islanded system.
The power factor of the load is what it is, and no amount of trying to run the system in VAr control mode is going to change that.
It is possible to run single sets as part of a station in VAr control mode with modern controllers, but not the whole station, and in my experience with these sorts of power stations, no one has considered it worth doing.

EDMS Australia

 

[RRaghunath]

If the load is not steady and if there is large motor starting with associated voltage dips, Voltage control mode will ensure faster response from the machines in correcting the voltage to rated quickly.
This is also beneficial in boosting the currents during a fault in the power system and thus helping better response from protective relays.

 

[rockman7892]

A bit of a sidebar to main topic here but was wondering if anyone had any good reference that summarizes the difference between voltage control and MVAR control modes?

 

[waross]

It depends what you need.
Voltage control is the basic control for a system.
MVAR control is to control MVAR sharing in side a system.
The load sets the MVAR demand on the system.
There are a number of reasons to control MVAR sharing between the different generators in a system.
<Anecdote Alert>
A real world example of MVAR control.
A city was originally supplied by a diesel generator plant.
A large hydro-electric facility was constructed in the country and the city was supplied by a long transmission line.
As the city grew, the capacity of the transmission line was reached.
In this case the capacity of the line was limited by the ability of the OLTCs to compensate for transmission line voltage drop.
The transmission line voltage drop was mostly reactive.
The old diesel plant was started up and the sets run under MVAR control during peak periods, to pump MVARs into the line to offset transmission line voltage drop.
With the diesel sets pumping out MVARs and little or no real power, the fuel costs were negligible.
This scheme resulted in a substantial increase in the transmission line capacity.
</AA>
There may be other reasons for MVAR control.
In system with generators of different types and different fuel efficiencies, MVAR control may be used to shift The MVAR demand to the less efficient sets to allow the more efficient sets to generate more real power.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[rockman7892]

waross - Thanks for the example. In the example that you described wouldn't the generator in that case be the same a sync condenser providing just MVAR support to the system?

 

[waross]

Yes, very similar to a synchronous condenser. They didn't have synchronous condensers but had the diesel sets. The diesel sets were installed and connected. They just needed to be serviced and started to go into operation.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[FreddyNurk]

Great anecdote Bill. Thanks.

EDMS Australia

 

[crshears]

With the diesel sets pumping out MVARs and little or no real power, the fuel costs were negligible.

Hey Bill, did not minimum load have to be applied to the Diesel to prevent cylinder/injector fouling and - what was it you called it elsewhere - "slobbering"? If so the fuel cost of using this unit to provide MX would not have been trivial. I ASSume of course that there was no means of disconnecting the generator from the Diesel engine once the unit was synched, and that if not there was also no ready means of "unloading" the engine at zero fuel rack.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[waross]

Hi CR. I was not able to get close to the plant and I don't know the KW loading on the sets.
I don't know the numbers but I suspect that the fuel cost was a small investment compared to the added KW that could be delivered over the transmission line.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[redlinej]

The term swing generator in island mode can operate in the following scenario:

Scenario #1 The swing generator speed governor is set to control the frequency in isochronous mode and AVR set to voltage control mode ( voltage droop control). The other two generator speed governor set to baseload mode (fixed kW control) and the AVR in PF or Mvar control mode.

Scenario #2 The swing generator speed governor is set to controlled the frequency in droop mode and AVR in voltage control and the other two set speed governor to fixed KW mode and AVR set to PF or MVAR mode.

Scenario #3 The swing generator speed governor is set to controlled the frequency in isochronous mode and AVR in voltage control and the other two set speed governor set to droop mode and AVR set to PF or MVAR mode.

These scenarios above are for swing set because alternatively all three generator can be set to KW loading mode and KVAR sharing mode. Or speed droop and voltage droop load sharing control.

 

[waross]

Using a swing set is a little more involved than just setting one set to isochronous mode.
There must also be provision for adding or removing units of baseload generation as the system load changes.
If a system load change exceeds the ability of the swing set to compensate, the system reverts to droop mode with the swing set pegged at either 100% or 0% of rated output.
Historically the load dispatch center would adjust the baseload via telephone.
As the system load increased or decreased, the load dispatcher would place a phone call to a baseload generating station with instructions to either increase or decrease their baseload contribution.
I'll leave it to someone else as to how to program that phone call into ETAP.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[redlinej]

@Waross, I understand that provision have to be made for adding and removing generator, but I was answering the question post by OP with 3 generator scenario. Also, in island mode with 3 generator more likely there will not be a dispatcher.

 

[waross]

Also, in island mode with 3 generator more likely there will not be a dispatcher.

That is one reason why, when I was involved in discussions on the desirability of implementing a swing set in our three generator islanded plant, we opted to remain on droop control.
Our operators still added and dropped sets as the load varied.
While 3% droop may imply a possible frequency error of 1.8 Hz, our operators checked and trimmed the frequency every 15 minutes (more or less).
Our frequency error was seldom greater than 0.5 Hz.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!