Reactive Load Sharing of Generators in Parallel

If you increase the engine power, the generator watts output will increase. This will change both VAr and power factor if excitation is unchanged.

Increasing excitation will increase generator emf [it is NOT a back emf]. The effect on output VAr and power factor depends on the the generator internal and interconnection impedance.

With "constant VAr" control, the reactive VArs will stay constant as watts increase (and power factor must become higher).

With "constant power factor" control, VArs will increase magnitude in proportion to watts, so power factor remains constant.

- and there is a third mode, voltage control.

If you are on an isolated system, with only a single generator, you can only change the voltage (assuming constant frequency) - the Watts, VArs and power factor will be decided by the load.

If your generator is connected to an "infinite Grid", roughly speaking, changing the excitation will change the VAr output, with very little effect on watts.

Watts, VA and VAr are interconnected via a geometric triangle based on the angle between voltage and current and the definition that reactive current is at 90 degrees to the active watt producing/absorbing current in phase with the voltage.

If the VArs are to be kept constant in a real Grid system with fluctuating loads, the human operator must keep adjusting the excitation or an automatic feedback control loop be provided, using a VAr indicator or measurement.

VArs may be regulated in proportion to Grid voltage to give a system voltage regulating effect.

In most practical systems, the generator has an automatic voltage regulator (AVR) which can keep an almost constant terminal voltage on an isolated load, with a current feedback into the voltage reference which causes a voltage fall (or droop) as reactive current increases making parallel operation with other generators controllable.

The power factor or VA or Var control is an additional control loop which varies the voltage reference with a slower response than the voltage control. There are are also faster feedbacks called "PSS - Power System Stabilizers" which attempt to damp out oscillations of system voltage or frequency.

By increasing the engine power, I hope you mean increasing the fuel by governor control... this would only increase the frequency. How will it increase the VAR and PF? Can you please elaborate.

By increasing the engine power, I hope you mean that as the load increases on the generator, frequency comes down and by supplying more fuel by governor control we maintain the frequency more or less constant.

So I dont undestand when you say that it increase the VAR and PF if excitation is held constant Can you please elaborate.

The purpose of excitation voltage is to maintain the terminal voltage constant under all conditions of generator operation. As far as excitation and governor control are concerned they are totally independent.

At no load, the excitation system should only supply that much amount of volts necessary to maintain the terminal voltage of generator constant. When the generator is loaded, it's terminal voltage decreases slightly, therefore an excitation system should sense this voltage dip and increase the excitation volts immediately and thus maintains the terminal voltage.

My question was while maintaining terminal voltage - why do some AVRs say they can operate in VAR or PF control mode.

Appreciate if you can elaborate.

Be clear in your mind - if a 25 MW generator is working into a "20 GW load" Grid, it can have very little effect on frequency and connection point voltage (its connection point voltage will usually be regulated by Grid step-down transformer automatic tap-changers anyhow).

Do not get confused between one genset working to an isolated load and parallel operation of gensets. Your question about VAr or PF control can only apply to parallel operation. As I wrote, a generator isolated on a load cannot change power factor/VA (except insofar as the load VA and power factor change with voltage - basic case power factor of a resistive load will always be unity and so VArs = 0, in that case VAr/PF control would just cause over or under voltage).

Or bring in the governor! On a Grid, sets below a certain power, e.g. 10 MW, are not required to have governors.

On a steam set, one can just open the steam valve and increase power at constant speed.

I did not write that increasing engine power would increase VAr and PF - I wrote "change"!

In parallel running, increasing generator drive torque advances the phase of the generator emf in the direction of movement (angle between emf and terminal voltage changes). Effect on VAr and PF depends on whether PF was lagging or leading before change.

You need to look at the theory of current flow between two ideal sinusoidal voltage generators connected by an impedance, the basis of synchronous machine theory.

If you connected two perfect voltage regulated machines in parallel, current flow would be infinite for any small voltage or phase difference. Even two real machines with internal reactance would have uncontrolled current flows between them, dependent on difference between voltage set-point and regulation between them.

For parallel operation of alternators on same busbars, it is necessary to have voltage regulators with feedback of lagging current to cause voltage "droop" proportional to lagging current [usually about 3% droop for rated current lagging] or interconnection of AVRs to provide a reactive load-sharing feedback.

Imagine a 100 kW AC generator in direct parallel with a 1000 MW AC generator. With the 1000 MW set in AVR control, change in voltage will be unmeasurable as the 100 kW machine increases its kW. So how can the 100 KW AVR see any voltage change to alter excitation?

A basic AVR just takes in voltage. An increase in complication and cost gets simple current feedback (needs current transfo or shunt etc = more cost) for parallel operation voltage droop. More cost gets a cosine or sine "calculation" which enables a VAr or power factor feedback control.

Excitation current and governor are not independent in practical operation, increase in Watts needs an increase in excitation. When shaft power is increased, the typical large generator will go "out of synchronism" with parallel sets at about 1/3 rated watts, if excitation current is left at the initial current (no-load value).

Why operate in VAr or power factor control? Real power systems cannot control the watts and VAr their consumers put on - load varies second to second, with daylight, weather and TV programmes!

But there is an obligation to keep consumers' frequency and voltage within narrow limits. Practically, modern Grid systems buy watts and VArs separately off independent generators. Operationally, it is a "nightmare" if all generators can generate whatever watts or VArs they want at any time or place, on top of the unpredictable load.

Practically, it is better if watts and VArs are contracted at known times, levels and locations. Whatever the commercial arrangement, it is more energy efficient to generate VARs in a big generator than a small one. A limited number of generators contract for Watt or VAr outputs which fluctuate to meet the short-term fluctuations of demand and act to keep frequency and voltage steady.

Unity power factor or fixed VAr outputs have least disturbing effect on voltage, so most smaller generators are required to operate in these modes - hence the need for AVRs to have these modes.

"If your generator is connected to an "infinite Grid", roughly speaking, changing the excitation will change the VAr output, with very little effect on watts."

Again, how it effect the Watts?

"Excitation current and governor are not independent in practical operation, increase in Watts needs an increase in excitation. When shaft power is increased, the typical large generator will go "out of synchronism" with parallel sets at about 1/3 rated watts, if excitation current is left at the initial current (no-load value)."

Can you please explain, why increase in watts needs an increase in excitation?

The briefest explanation is that the the strength of the rotor electromagnet affects how much torque it can produce to oppose the torque of the engine.

Exceed the the torque the rotor magnet can give and rotor angle advances continuously relative to Grid - this causes torque cycling from negative max to positive max with corresponding forces on rotor/stator windings, generator-engine flexible coupling, foundations, along with VAr cycling round the clock.

Not surprising that automatic relay protection against "loss of sync" is fundamental to getting a Grid connection permission.

A generator has to be connected initially with low KW, low frequency slip, voltage and phase exactly matching the Grid - excitation much less than necessary for full kW.

Loading a generator manually one must advance governor and voltage in steps to keep a required power factor, avoiding low leading power factors which risk loss of sync.

"If you increase the engine power, the generator watts output will increase. This will change both VAr and power factor if excitation is unchanged."

As I understand PF = kW /(kW+kVAR), therefore only PF gets changed, how it affects VAR? as if the excitation is unchanged, so is the VAr.

In reply to post #7....

Oddly, CR4 has since some months ago, started serving up posts from years ago - it seems to be sorting posts oldest first, not newest first!! This "Reactive Load..." post is dated 11/01/2011 in USA notation! If you mouse click on the post date title the list changes to proper date order.

You misunderstand, PF = kW/(kW + kVAr) is only true if kW and kVAr are combined (added) respecting their angular difference in a graphic triangle in which kW and kVAr sides are at right angles to each other and longest side is kVA. That is vector addition not arithmetic addition.

Fundamentally, PF = watts absorbed by load/(volts x amps into load) = kW/kVA

A typical generator can be represented as a constant voltage (e.m.f. = electromotive force) feeding through its winding inductance to bus or Grid voltage which is constant - unaffected by the amount of current/power generator delivers.

Using the triangle theory on a voltage diagram triangle, if the emf and grid sides stay constant you can increase the voltage across the inductance & generator current by changing the angle between emf & grid. That changes both power factor and VArs, but if the engine keeps the same throttle position the kW does not change.

When the fuel is increased to a genset (kW increase at constant speed of rotation), a spot on its shaft viewed using a stroboscope flashed in step with the Grid volts zero instant moves in the direction of rotation - the direction of the rotor magnetic field and the emf it produces is attached to its physical angle relative to the stator field angle.

I hope, you do not apply this new found knowledge to an asychronous generator.

GENERATOR SET-KW & KVAR LOAD SHARING (IN A PARALLELED SYSTEM)

LOAD SHARING

Load sharing is defined as the proportional division of the kW and kVAR total

load between multiple generator sets in a paralleled system.

Load sharing is essential to avoid overloading and stability problems on the

systems' generator sets.

ACTIVE POWER (KW) LOAD SHARING

When generator sets operate in parallel, the engine speed governor of each

generator set determines the proportional sharing of the total active power

requirements (kW) of the system.

The kW load sharing is achieved by increasing or decreasing fuel to the

systems' engines.

As the fuel to the engine of one generator set in a group is increased it will not

lead to an increase in speed and hence frequency (as it would if it were

operating alone) but it will lead to an increase in the proportion of the total kW

load that it will deliver.

As the fuel to the engine of one generator set in a group is decreased it will not

lead to a decrease in speed and hence frequency (as it would if it were

operating alone) but it will lead to a decrease in the proportion of the total kW

load that it will deliver.

The control system of the generator sets (via the engine speed control system)

monitors and controls the sharing of the total kW load in proportion to the

relative rating of the engines on the systems' generator sets.

Created by J.M.J. Lloyd Uncontrolled once printed

REACTIVE POWER (KVAR) LOAD SHARING

When generator sets operate in parallel the alternator field excitation system

of each generator set controls the proportional sharing of the total reactive

power requirements (kVAR) of the system.

The kVAR load sharing is achieved by increasing or decreasing the field

excitation to the systems' alternators.

As the field excitation of one generator set in a group is increased i.e. overexcited

it will not lead to an increase in voltage (as it would if it were operating

alone) but it will lead to an increase in the proportion of the total kVAR load it

will deliver and a decrease in its power factor.

As the field excitation of one generator set in a group is decreased i.e. underexcited

it will not lead to a decrease in voltage (as it would if it were operating

alone) but it will lead to a decrease in the proportion of the total kVAR it will

deliver an increase in its power factor.

An undesirable circulating reactive current (cross current) will flow in the system

if the excitation of the alternators is not matched.

The voltage control system of the generator sets (via the alternator voltage

control system) monitors and controls the sharing of the total kVAR load in

proportion to the relative rating of the alternators on the systems' generator

sets.