Switched capacitors at generating plant 5

[dcdtn]

I have a situation in which we have installed switched capacitors on the bus at a generating plant. There are five generator units and 4 cap banks. Obviously, there is a conflict between the caps and generators since they are controlling the voltage at the same bus. In addition, for stability reasons, it is desirable to minimize (or eliminate) reactive power absorption of the generators (i.e. Qmin=0). My question is what is the recommended method for controlling the capacitors? One alternative I thought of was to limit the exciters on the generators so that Qmin=0 and set caps as normal. This has one minor drawback since the exciter settings are manual and cannot easily be overridden. Does any one have other suggestions?

Thanks.

 

[busbar]

Maybe I’ve misunderstood the question, but isn’t the usual practice to install {shunt} capacitors closer to loads where the reactive power is needed and not "on top" of synchronous generators? It seems like excitation {and its very controllable dynamics} would be more than adequate for conditions at a generator site, and effectively without additional costs or equipment.

 

[jghrist]

Are you sure it is a synchronous generator? It would be common to install capacitors for pf correction at an induction generator, but not at a synchronous one.

 

[dcdtn]

Yes, these are synchronous generators. Admittedly, this is an unusual case, but there are several situations where it is desirable to have capacitors close to generation. In this case, the generators are fairly small (50 MVA each) and do not produce enough reactive power under certain system conditions. Other possible situations: 1.) Use capacitors to increase the available dynamic VARs in a unit for added stability. 2.) Generator may operate at Pmax at unity power factor such that there is no reactive power output or reserve. 3.) Loss of generator leading to low voltage and need for capacitors.

 

[SidiropoulosM]

What size are the capacitor banks? What was the main reason for installing them at the generator bus? It can't be for dynamic VARS if these are just static capacitors. If these generators are synchronous, as you said, I can't think of any reason why you want the capacitors right there and not closer to the load. Please provide more information on the reasons for this set up. In our system we would do this only for wind generators, which are induction machines. But in that case we would use some type of a dynamic reactive source, not a static capacitor.



Michael Sidiropoulos

 

[dcdtn]

Just in case there is confusion, the capacitors are on the bus that the high side of the GSU's are connected to. There are 4 banks of about 17 MVAR each.

Static capacitors cannot be used for dynamic VARs by themselves, but they can free up dynamic VARs from generators that are outputing maximum VARs already. In my situation, the capacitors are needed because there is more load (both real and reactive) in the area than generation.

 

[peterb]

Just a point to add here re the generator stability limit. By operating at reduced VARs on the machine, you are in fact operating with a REDUCED stability margin. The dynamic stability limit lies in the negative VAR region, so you would normally operate the machine at a lagging power factor (positive VAR output) to establish an adequate margin to allow the generator to ride through system swings.

 

[peebee]

peterb: could you please expand your last post to provide a bit more detail?

 

[dcdtn]

peterb,

That is correct and brings me back to my original question: How to coordinate control of the capacitors and generators so that generators do not absorb VARS? I have a couple of ideas (one already mentioned) but am looking for other opinions.

 

[busbar]

One way of sensing absorbed machine vars may be a reverse-power [32] relay with the potential input connected in quadrature.

 

[peterb]

dcdtn -
Seems to me that you may need to consider operating the generators at rated power factor, on PF control if available. As a minimum, you need to ensure that the minimum excitation limiter on the AVR is properly set up to avoid absorbing VARs. Possibly tie line PF monitoring to control the capacitor bank switching? (Sounds as though what you are really trying to achieve is minimizing VAR flow on the tie lines). Re-reading the original post, you may have to upgrade the AVRs to achieve this.

peebee -
If you examine a typical generator capability curve, the stability limit line is shown in the negative VAR region. Operation beyond this line will result in the machine falling out of step. Normal practice would be to operate with an adequate margin separating the operation point from the stability limit, usually resulting in operation well into the positive VAR quadrant. This ensures that the excitation is high enough to maintain stability when swings occur.

 

[peebee]

peterb, you get a star from me. Anyone else who can shed some more light on vars & gens gets a star from me too. I still don't have a very qualitative understanding of what goes on when you start shoving vars around with generator systems, other than a vague sense that if you turn that pot just a little too far one way or the other that BAD THINGS start happening. Any further explanation of the BAD THINGS would be appreciated. Is this a resonance issue, with the plant or utility? Is this something purely internal to the gens, that they don't like to have vars shoved on to them -- and if so, why not?

I was recently on a project where they parallelled (3) 760kW, 480v generators full time with the utility, the utility tx normally only saw 100kW of load, the gens ran near rating. Due to the tx being unloaded, they would experience overvoltage. But if the gens shut down, they'd go to sever undervoltage. Someone came up with the bright idea to import LOTS of vars from the utility in an attempt to drop the voltage. And that seemed to help somewhat with the voltage issues. Someone else came up with the warning that on a similar installation, someone turned the dial a little too far and it launched the tie breaker out of the switchboard.

"Instability" sounds like a reasonable diagnosis there. Any more explanation of what's going on there would be appreciated.

I've asked this question of our local [#1 gen mfgr] dealer, seems I already knew more about this problem than they did [they were the one's that came up with the idea to suck vars from the utility in the first place].

 

[dcdtn]

Thanks everyone for your responses. I am going to look into limiting the exciters on the generators so that they do not consume VARS. They will still be in voltage control mode and will be allowed produce VARS.

 

[jbartos]

Suggestion: Sometimes, capacitors are used in passive harmonic filtering (PHF). PHF can be engineered and designed relatively high in the power distribution system, e.g. close to power source to correct the voltage waveforms distorted by voltage harmonics.
Alternately, the capacitors can be installed at the generator if the load is very inductive, high VARsind. They also cause problems, namely, heating problems.

 

[dcdtn]

jbartos, can you expand on the heating problems you are talking about?

 

[HVmariner]

Just a little suggestion:

If you know in which configuration of the network (ie which loads are connected) you really need to have the capacitors connected, why not proceed like this:

- keep the generators excitation as it is (for them to control the bus voltage).
- add some sequential devices (for exemple relays) to have the capacitors breaker closed when they are needed, and open when not.

Regards,

 

[windie]

Why not keep the voltage control with the generators and switch the capacitors in respose to the VAr loadings on the generators in order to take the reactive loading off the generators. That way you are able to use the dynamic cabibility of your generators for voltage control. For example if the combined generator output is +10 MVAr, switch a cap bank in, if they are abosbing more than 10MVAr switch a bank out.

You may have to adjust levels and time delays in order to avoid excessive capacitor switching.

Similar controls are used to extend the reactive range of SVCs - operators/suppliers of these devices may be able to provide more insight.

 

[dcdtn]

windie - Thanks for the advice. That is very similar to what we ended up doing a few months ago. We simplified the logic a little since there were 5 units and to do it right, you have to know how many are turned on.

 

[pablo02]

We have had both capacitor banks and static VAR systems located in a 4 unit 350 MW plant substation (on the high side bus) -- however, it's been 8 years since I left that plant and I can't remember what we did, although I remember that once the cap banks were energiized, we were given an alarm ... I believe the operators had to compensate for the VAR loading... the static VAR system didn't really kick in unless we had a system upset (our load center was local, the majority of generation was 150 miles or more away) -- most our plants had capacitor banks located in the switchyard...