Are Utility Power Quality Issues Impacting Genset Stability?

[rodney1981]

Lil background that may support the best answer to my question.

I'm a plant manager for a small landfill gas power generation facility 3.6MW, 2ea RICE CAT 3520'S, which are paralleled to the grid. I believe my site is experiencing utility power quality issues that are effecting genset stability. Some known facts is the site is at the end of the utility distribution line in a rural area. The bus or utility frequency seems to deviate even with the genset offline more than other sites I operated in the past, for example 59.97-60.3ish, I'm not so sure if this swing is acceptable or not, and this swing moves relative fast - within minutes.

Also when the site gensets load increases the generator/line voltage increases, our generators are 4160v with a step up transformer to line voltage of 12480v, generator voltage increases up to 200-250v from nominal start voltage most of the time. Reactive Power also seems to erratic at times, we run a PF .99 and put out about 250vars from each genset.

My background is mechanical so before I start making phone calls to our power purchasers (utility company) I like to have the decent knowledge on what might be contributing this issues, and not sound like a complete idiot. We have multiple devices monitoring our generator and line power quality and I'm seeing the same issues, so its highly unlikely its instrumentation. We also recently load banked our gensets to see the gensets stability and no anomalies were found. Because the gensets are unstable at load we are experiencing cylinder detonation due to unusual loading and unloading of engine.

I have exhausted all my tricks with tuning the switchgear VAR/load PID loops along with engine governor system and generator voltage regulator gain settings, and unfortunately the local CAT techs have less technical knowledge in this department then I do.lol

Any insight will be greatly appreciated

 

[magoo]

My key notes:
- end of the utility distribution line in a rural area

- 12.47 kV on high side to 4.16 kV

- Frequency swings - +/- 0.03 Hz typical (60.3 looks like a typo)

- Voltage increases by 200-250 V with generators running

3.6 MW at 12.47 kV is a significant change in load for a rural distribution feeder. My experience with 12.47 kV utility feeders is that the typical load is maybe 400 A at peak, which is about 8.6 MVA. By starting up your gen sets, you change the feeder load significantly and that change results in the higher voltage that you are seeing. I assume the 200-250 V change is at the 4.16 kV side. Is that correct? If so, is that the phase to neutral or phase to phase value?

I would ask the utility what is the expected voltage when the generators are off-line and when they are on-line? This would save me guessing at it and providing you with some bogus numbers.

 

[anorlunda]

Are you on an isolated island, or a very small grid? Those frequency changes sound huge.

The utility engineers are best equipped to do the calculations to predict power quality at your llocation. It is their responsibility. In fact in regulated areas like I am used to, these calculations would be done when your plant was first proposed, long before it was connected to the grid. In many cases, you would need the services of a consulting engineer to help design your system and to work with the utility for them to verify grid performance or perhaps to make changes. Your plant might be causing power quality issues for other customers on the same feeder.

How is it that you are dealing with those issues at such a late stage?

 

[Rive]

... the site is at the end of the utility distribution line in a rural area. The bus or utility frequency seems to deviate even with the genset offline more than other sites I operated in the past, for example 59.97-60.3ish, I'm not so sure if this swing is acceptable or not, and this swing moves relative fast - within minutes.

Just by curiosity: are there other generators on the same distribution line//nearby?

 

[rodney1981]

Magoo,

-Yes the frequency was a typo.
-The voltage increase is viewed on the 4.16kv side -switchgear and generator protective relay- and the voltage increase is viewed phase to phase.
-Ok I can def find out what the nominal line voltage is for a reference for this area and expected voltage with gensets loaded.

Anorlunda,

- There was a field study for this site, but the owners do not recall much, I was told by utility line worker he believes that the power we are producing is maxing out what the lines can handle, which doesn't tell me much.
- Our generators are paralleled to the grid, however the utility company just requested reviewing our switchgear equipment/operations/programing to verify that the generator breakers would open in pre determined time if the generators some how got islanded. Maybe this test is indicative to the area?
- "How is it that you are dealing with those issues at such a late stage?" my previous operators had no idea of these issues! they thought the genset swing and voltage increase was normal, same for the frequency shift.

Rive,
-No other known generators near by.

Doing some laymen leg work, it sounds like line capacitors could help? but then again I'm sure there much more to this problem.

I like where this is going, thanks!

 

[magoo]

rodney
from what you said, I would be looking to change the power factor setting from 0.99 to around 0.85 lagging. The lagging setting will compensate for some of the voltage rise you are experiencing. A 0.99 PF setting might be OK if your facility were close to the distribution substation.

I don't think capacitors would help. They would actually make things worse.

 

[rodney1981]

Magoo,
Changing the PF increases reactive power and moving that setting will increase the VARs alot, should I be concerned with Stator Winding temps doing so? We do have winding temp probes so I will monitor closely. So the distribution of reactive power may help stabilize some of these issues? I guess I try and see and report back.
Thanks
Rodney

 

[magoo]

I don't expect you'll have a problem but I didn't have a chance to look up your machines. What is the nameplate say about MW and PF?

 

[jim hardy]

Is there a single stepup transformer or one for each generator ?

Just some things you could try to become more conversant:

Most voltage regulators have a setting for reactive droop. It may be called line loss compensator, or reactive something, or most anything else depending on what was in the mind of design guy.
It probably came from the factory set for droop so you can operate the machine in parallel with a stout bus..When operating with a transformer in between a machine and its bus,
the transformer's impedance gives you some droop, maybe more than you want and that might be part of your voltage swings.
In that case the regulator's droop setting should be changed from droop to boost or rise, to cancel out some of the transformer's impedance.
Of course one can get too aggressive with that setting and make his machine unstable.
And be aware when you make that change the machine becomes a lot more sensitive to voltage regulator voltage tweaks. What used to be to the operators "Just a pinch of vars" becomes a double handful. So warn them.

First thing i'd do is find out from your voltage regulator guru
what is that droop/rise/compensation adjustment set to ? How much droop or rise? Why ?

next thing i'd do is find out from the utility
what is system impedance to infinite bus from your point of interconnect ?
And from Rice what is the inertia constant of your generator?
Then ask your voltage regulator guru to have his summer intern EE student calculate natural frequency and damping for your machine. That's mainly an academic exercise, i wouldn't expect trouble but then, on the other hand, we had a 1000 mw machine at end of a long line slip a pole... so check before something gets broke. That's paying attention to detail.

Then i'd take a stroboscope out to the machine and power it from as close as i could get to machine terminal volts. I used an old fashioned "Strobotac" on our 900 mva steam turbine.
Observe shaft, you should see sort of random small hunting as its power angle follows system conditions. Any sustained regular oscillations at low frequency like 1 hz are cause to call in a power system specialist like @anorlunda .

You'll probably find some days it's stable, others it's not , depending on settings of other voltage regulators out on the nearby grid.
Switch regulator from auto to manual and see if oscillations go away.
Hopefully all you need is some attention to the droop/rise setting.
We set all ours to give ~7% impedance counting the transformer,
IOW if a generator was connected through a 12% transformer we'd change the reactive droop to reactive rise and set it for 5%.

my two cents, probably overpriced at that , please pardon my 'old firehouse syndrome'...

good luck, and keep us posted ?

old jim

 

[rodney1981]

name plate doesn't mention anything about power factor, the genset is 1.6MW, I am sure I am not understanding some here.

 

[rodney1981]

wow Jim that was alot, I always had a hunch that the digital voltage regulator could use some finer tuning to compliment the bus. Good suggestion to find a guru for particular equipment and VR.
Rodney

 

[rodney1981]

Single step transformer for both gensets- 3750kva. The Caterpiller Digital Voltage Regulator does have some droop settings but I have to do some research on the specifics of each droop settings. Here is pic of the VR settings on one of the units, this might shed some light

 

[magoo]

The CDVR sheet shows 0.8 PF for the machine so my suggestion will work without harming either machine.

I should have asked early on, but is this a new installation? Sounds like it must be or you would have had these problems for some time.

 

[rodney1981]

the problem has been present for years, super green horn operators had no idea of the genset operating conditions, and mostly thought it was fuel related issues causing the genset instability which its definitely not.

 

[anorlunda]

OK, I think your problem is voltage, not frequency. Your mention of frequency in the first place is a red herring and confused the issue.

It is possible that the distribution feeder is too high impedance for the amount of load and load variation you demand. That's why I asked about the history. Issues like this can cause voltage regulation problems and they are normally sorted out in advance by the utility engineers, so that complaints are not generated. Something went wrong in your history.

If you are a normal retail customer, then it is the utility's duty to regulate voltage to within the limits specified in their tariff. So this is one of the rare occasions that I disagree with my friend @jim hardy. His advice was how to tune a voltage regulator. Jim is technically correct, but I think that the fact that you have a voltage regulator at all, muddies the water. It makes a case of 100% clear utility responsibility into a case of mixed responsibility.

Industrial customers, as opposed to retail customers, typically have some explicit technical requirements about customer's responsibilities built into their contract.

I think what you should do is to put a data logger on to record the voltage at your service entrance 24x7, with measurements of one per second, or one per minute. Plot those results versus time for several days when you experience the problem. Then send documentation of those results with a letter stating that you have a power quality problem to the utility with cc to the Public Utilities Commision in your state. Do not work through 800 number customer service channels. That should trigger an investigation by them.

Hopefully, they'll determine that it is their fault and correct it, but even if it is your fault, the investigation should point out what you need to do. As I said before, they have the resources and the talent to solve voltage regulation problems much better than you and they have the legal responsibility to maintain power quality.

 

[rodney1981]

Anorlunda,
I completely agree a power quality test needs to be conducted by a third party to have sufficient documentation to present. I still will look into some VR tuning which may or may not help stabilize the genset.
I don't know how this comes into play but there's no doubt the frequency is oscillating even with the gensets offline, my other sites that do not have any issues the frequency barely moves (59.98-60.02) and usually only changes from night to day operations, not a rapid fluctuation. From a laymen perspective this hard not to look into this as a part of the equation.
Rodney

 

[jim hardy]

Thanks @anorlunda

my reason for suggesting look at the Voltage Regulator is they're all too often considered a mysterious black box set up initially by black robed wizards who speak in tongues , never again to be touched by mortals.

I'd guess from this snip of his voltage regulator sheet

that there's no reactive compensation at all dialed in, droop or rise

which with his generators paralleled on the low side of the transformer i'd expect to cause var sharing troubles between them
one machine way in lag and other way in lead, difference going out to the grid

but to make that call one needs to be there and watch the ammeters as he adjusts excitation.

I have exhausted all my tricks with tuning the switchgear VAR/load PID loops along with engine governor system and generator voltage regulator gain settings, and unfortunately the local CAT techs have less technical knowledge in this department then I do.lol

Rodney needs to find that black robed wizard. i once had the great good fortune to watch over the shoulder of one .

 

[anorlunda]

@jim hardy in #1 he said that they have problems even with the genset offline.

@rodney1981 , I should have also said that if the data logger shows that utility voltage is within the legal limits, you have no grounds for a power quality complaint (I remain skeptical of the frequency part). Then you would have to look internally as to why your process is so sensitive.

Gather evidence on both voltage and frequency and let the evidence guide you. Make sure the logger samples often enough to record any oscillations.

 

[jim hardy]

@jim hardy in #1 he said that they have problems even with the genset offline.

i confuse easily

We also recently load banked our gensets to see the gensets stability and no anomalies were found. Because the gensets are unstable at load we are experiencing cylinder detonation due to unusual loading and unloading of engine.

i took 'at load" to mean on grid.

Gather evidence on both voltage and frequency and let the evidence guide you. Make sure the logger samples often enough to record any oscillations.

@rodney1981
A stroboscope on the shaft you can do very quickly and it's nonintrusive. Your mechanical types will really enjoy seeing the shaft advance its angle as you advance the throttles. Synch strobe to machine terminal volts - either from a PT or a wall outlet from same bus generators are on.
Watch it with Voltage Regulators both in auto and in manual. Much difference? Does angle hunt more in auto? How about with one VR on auto and other on manual ?