Unequal Reactive Loading

Give us a description of your load control center.

Are you referring to the Load sharing modules of the generators?

In any case please find below the following details:

The generators are FG wilson

Engine: Perkins

Alternator: Leroy Somer

Control Panel (LSM): Power Wizard 1

There is no central Control module, but each and every generator has its own.

Thanks

RC

Okay, not familiar with your configuration.

We utilize a centralized generator control center which synchronizes and automatically adjusts the load sharing. It makes it easy to operate and requires monitoring only, unless there is a fault of some type, then manual control can be effected from the front panel.

thanks but is this a problem from wrong settings of the AVR? i mean because of the unbalance in reactive loads.

thanks:)

It was discussed before, time and again.

Excitation (voltage produced) against a common real load establishes load sharing of the real part.

Phase leading (or lagging) of the individual generator's rotor establishes the inductive or capacitive component produced by the same generator. That is motion, drive.

The excitation settings cannot do both.

Sit down, and make drawings of the cases, to learn. It is not rocket science.

Leveles thanks for your response. I agree with the above and in the perspective that the terminal voltage Vt is the same (correct) for the three generators, the only plausable remaining parameter is the induced voltage E of the generators.

My question is rather more on the practical side than theoretical. Is this unbalancing in the reactive component be resolved by tuning the AVR (Automatic Voltage Regulators) of the generators?

I thank you for your responses and patience

Regards

RC

Because the generators are in parallel they have the same terminal voltage Vt.

Reactive power is primarily dependent on induced voltage and the equivalent generator impedance and terminal voltage.

Thus for the 3 generators we have the following 3 equations with a common terminal voltage are applicable,

Eth1 = Vt + [R1+jX1]I1

Eth2 = Vt + [R2+jX2]I2

Eth3 = Vt + [R3+jX3]I3

Because the active power is shared equally , the phase angles of Eth1, Eth2, Eth3 must be nearly same with respect to Vt on the assumption that generator impedances are identical.

Because it is assumed all 3 generators are identical in design, R1+jX1 = R2+jX2 = R3+jX3

this leaves only Eths are unequal or an excitation problem.

Real life is not equations.

Magnitude = voltage /power = real part

Phase = imaginary part (math expression) = inductive / capacitive

In real life application phase relations are real. As in leading or lagging.

Sit down, and work it out with pencil and paper.

Thanks leveles I will do so and let you know the outcome

Thanks RC

Unfortunately we design and consulting engineers use only the equations for explanations and specifications in this real world.

Assuming that generator impedance to common bus-bar is purely inductive, the phase difference between Eth and Vt will determine mainly active power flow only.

The reactive power flow is mainly dependent on the magnitude of the thevenin's voltage Eth and the terminal voltage Vt.

The terminal voltage Vt is a function of external load resulting in the generator current I.

This can be verified for typical generators where aramanture resistance will be almost negligible and synchronous reactance [either sub-transient, or transient or steady state ] is significantly higher.

Mathematical equations are the basis for almost everything we design and specify and is practiced universally in this real world.

When the 3 Gensets are paralleled, Load Sharing should be done by whatever system you are using:

The automatic paralleling equipment must have Load Sharing capability which needs to be properly calibrated and balanced at commissioning.

To spread the kVAR between the 3 generators, the fuel control is also involved.

The Voltage will be the same since they are on the same busbar and paralleled.

The Speeds will also be the same.

The difference will be the amount of mechanical power being put in by each generator: i.e. whether one is being pulled along or pushing the others, And the excitation current required to produce the voltage (or maintaining it).

If You consider the Apparent Powers S1, S2 and S3, then S1 will be the highest kVA while S2 and S3 will be the same kVA but opposite signes. This means that G2 (-15 kVAR) is capacitive with a negative p.f.

To try and adjust manually the excitation and the Fuel Governor is difficult and needs constant attendance. I suspect that the Load Sharing settings have not been properly adjusted from the start or that you are paralleling without a load sharing module. 3 Gensets need a proper load sharing module and settings.

thanks LAA I concur with you explanation above. this what i was after, i will let you know on the outcome :)

Lucke came the closest to the right explanation.

For the discussion let's assume, that the paralleled generators are identical, same make and model. In that case they have to HAVE parallel perfectly. You need to control two different factors: real and reactive. For that you need two different controls. Trying to control two things with one control is a fool's errand.

As the generators are tied together in parallel, they output voltage is forced to be the same. The excitation of the magnetic field is then setting the real power part via current. It does NOT set reactive (imaginary) part.

What is reactive, to start with? In a pure resistance fed from a voltage source the current is exactly in phase with the applied voltage: hence it is purely real power. When the PHASE of the current deviates from the applied voltage, you have reactive component. In inductors current is LAGGING. In capacitors current is LEADING over the voltage on the device.

Now, what happens when you reduce the fuel to one generator? It will try to lag behind. Unless you do something crude, it will not stop. The others will drag it along. Guess what? It still produces some real power, but became lagging current producer (inductive) too.

This explanation is for the phenomenon, and understanding.

And, no, dr. I am not interested on stepping on someone's living. You will have plenty of consulting to do, noticing the (lack of) speed of learning on the matter.

--------------------------------------

Measure, and the sit down to draw the cases for yourself. That is where real learning takes place.

"As the generators are tied together in parallel, they output voltage is forced to be the same. The excitation of the magnetic field is then setting the real power part via current. It does NOT set reactive (imaginary) part."

Dear leveles,

Pl note that when gen are tied together in island mode (without grid), output voltage will increase/decrease according to excitation increase/decrease.

When gen are tied together with grid, excitation voltage has no effect no effect on output voltage (bcos it remains the same as that of grid). Now, in this case when we increase/decrease excitation voltage of a gen, reactive power on this gen increases/ decreases.

Thats what Dr. is talking about.

Hope you have got it now.

Wow, amazing!

A single control setting two different things at the same time. Wow. Please tell me in larger detail, how such a miracle can happen?

Increasing / decreasing reactive part? What is that? I do not mind at all disagreement, but at least stay with understandable nomenclature: inductive / capacitive, or lagging / leading current phase. That is standard language since my great grandfather's time.

The equation for the power flow in a two bus sytem with voltages V1 and V2 and angular separation of theta radians between V1 and V2 and reactance of X between the two buses is given by

P = [V1*V2 * Sin(Theta)] / X

the corresponding equuation for reactive power is

Q = [V2*V1*Sin(Theta) - V2*V2]/X

It is clear that both voltage magnitudes, voltage angle separation [Theta] and the reactance between the two buses influence both P and Q.

However, for small values of Sin(Theta) will be nearly equal to Theta.

In other words Sine(Theta) = Theta.

Thus from the equations P increases linearly with Theta. Till Theta reaches 90 degrees. P is more sensitive to Theta value rather than to Voltage magnitude.

Thus P is controlled more by prime mover - which is the true active power contributor.

In fact, in swing equation of generator we only use turbine input and generator active power output as the variables to assess the generator swing with respect to a reference frame.

In the same way, for typical values of power flow and angular separation betwen the two buses, it can be determined Q is largely influenced by the voltage magnitude difference between the two buses and less by the angle Theta.

Consequently Q output is mainly controlled by Excitation control and P output is mainly controlled by prime mover - governor control. Even though to lesser extent these controls influence the other variables under consideration. For example Prime mover speed influences the induced emf. However, for normal operation, at near rated frequency, variation in excitation control influences Q more than prime mover as the generator is running at constant synchronous speed.

There are evidently two control loops, popularly known as torque angle loop and the excitation loop. The mechanical torque prime mover loop and another one is electrical torque via excitation control and external load. For steady state operations these torque must balance each other. Since for any disturbance, these two controls have different response times and response characteristics, the generator will normally swing for several seconds before settling to new stable operating point. Such oscillations may be observed when the generator excitation is adjusted manually in the present problem.

Interesting, dr.

So, one setting controls real power,

And the other setting controls torque whatever, what is called phase in normal electrical engineering., Good improvement.

As I said before, I do not step on another man's living.

BUT, do not even try to hold me for a fool!

Leveles

that is the way our present systems are working.

Active Power sharing control / phase control / by the prime mover-governor controls

Reactive power sharing control / by excitation controls

These are explained in all standard text books on power system stability and control

Even when the external active and reactive power demand from loads and transmission and distribution systems remains same, the sharing of these demands are determined by the relative phase angle difference among generators for active power and relative excitation among the generators for the reactive power.

do it on your own generators leveles..

then you will know it..

and do burn some midnight oil reading text books.

there is no meaning explaining someone who doesnt want to understand..

1) The unbalance in VAr can it be resolved by adjusting the generators field voltage i.e. excitation (in view of the large discrepancies) and is this accomplished throughthe generators AVR?

2) Is the 100A increase in G1's load current be attributed to this unbalance?

3) Are there any other possible issues that i need to consider in order to resolve

1) the reactive unbalancing

2) the 100A load current discrepancy of G1 as opposed to G2 and G3?

Your all questions have a same answer:

All same make gen sets running in parallel with grid have their a load sharing unit which shares active and reactive load b/w gen themselves and b/w grid and generators.

Now, what you need to check patiently is setting of avr of all of your gen which are getting command from load sharing module. or you can put your excitation in manual mode of all gen while running with grid, change the excitation voltage of all gen and see whether KVar of each gen changes or not.

This will sort out your problem. You are on right track.

thanksssss:)

Muditmah so my understanding was correct.....in order to rectify the problem (in practical means) we should tweak the AVR in order to get reactive balancing...

Yes. try it out in manual mode, by increasing/decreasing your generators excitation voltage. If, they respond to your increase/decrease pulses, then you can pin point that the problem is in AVR.

Yes

Try decrease the G2 excitation (Voltage Potentiometer) very slowly. This should move it into + kVAR. Also, you should monitor the other Gs and change their Pots accordingly.

If G2 becomes +15kVAR, with G3 the same, then G1 will be ~40kVAR as an example. That is assuming that the kW on all the 3 remains equally distributed. The tweeking will probably force you to agjust the spedd pots also.

I thank you all :) I will do so soon and will revert back with the outcome.

Thanks once again

RC

Dear All

Just to keep you in light of what was done on this issue, we have successfully managed to calibrate the set-points of the three generators by tweaking the AVR and governors ultimately obtaining approximately same current loading from the three generators.

I thank you very much for your help.

Cheers:)

Ryan