Excessive reactive power on a generator/transformer set 5

[PMACP]

Hi,

What could be the reason that leads a synchronous generator connected to a 15 kV network through a 0,4 kV/15 kV transformer to produce excessive reactive power?

This effect tends to occur during periods of reduced load on the MV network.

The generator must operate in a situation such that Q = 0.4 P. In fact, most of the time, the generator works on such regime. However, at certain periods, the reactive power generated goes far beyond the active power generated.

Some help?

 

[racobb]

Bill,

Please re-read my posts. I specified in each that the unit must be equipped with and must be set for var/pf control for this to work. Not pure voltage control.

I believe that you missed that part....
Thanks

Alan

Democracy is two wolves and a sheep deciding what to have for dinner. Liberty is a well armed sheep!
Ben Franklin

 

[waross]

I apologize. I did miss that part.
Yours
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[racobb]

Thanks Bill....just wanted to be sure I was not missing something important. It has appeared that most of the posts are essentially saying the same basic stuff...Alan

The major point that I wanted to make is that the AVR will have a sliding window of about 10% that must be adjusted at a level that will allow for the nromal utility voltage swings or it will run out of regulation range. This can be especially true if they set it up on a "normal" non-peak day and then is run during a peak time when the utility has reduced the voltage for load control. I have seen units that can't even parrallel when that happens. The same can happen when the load on a feeder drops and voltage rises during non-peak times of the day or month.

BTW...Love the handle for the member ivegotgas....that's a hoot!

Alan

Democracy is two wolves and a sheep deciding what to have for dinner. Liberty is a well armed sheep!
Ben Franklin

 

[catserveng]

It looks like you have gotten very good answers to most of your questions from those above.

About your question, a) Is that possible that the excitation system allows the existence of unbalanced active and reactive power generation between the phases?

As far as I know, the answer is NO. The excitation system acts to increase the magnetic field of the rotor, which increases the voltage (very simple explanation), the balance is mostly affected buy the connected loads. Some newer design generators, with what is called "square laminations" by one manufacturer, has an inherent votlage imbalance, and in networked systems canhave some strange readings at times, but not usually a problem, except if you're unfamiliar with it it may cause some head scratching.

As well explained above, increase excitation from the AVR to the pilot excitor (I'm assuming this is what you have due to size and age you have described)and the terminal voltage will rise if operating in a standalone condition. If you are networked, or paralleled, the network pretty much defines the system voltage, unless your generator has enough capacity to change the system. So your AVR can't work very well for controlling voltage. Most power factor controllers for your size unit work by becoming active when told the unit is in parallel, then biasing the voltage adjustment reference setpoint. If the unit is under-excited and consuming VAR's, the controller raises the reference setpoint, causing the AVR to increase excitation, in turn causing the generator to consume less VAR's, or produce VAR's into the system. If the unit is over-excited and producing excess VAR's, then the controller reduces the setpoint, the AVR lowers the excitation level, and so on.

Along with this is the dynamic adjustments, the AVR has stabililty settings to determine how well it responds to transients in the system, and most power factor controllers also have some dynamic adjustments to determine their response capabilities. In my experience most VAR/PF control systems get setup during periods of stable operation, then when system disruptions or transients occur, the VAR/PF controller may get into a dynamic conflict with the AVR, causing some instability, usually seen as large VAR swings.

So I guess if you want to be sure your system is functioning correctly, I'd try monitoring the field output of the AVR along with the other parameters you have. Then you can see the actual system response to some of the apparent disturbances your data showed.

Hope that helps, Mike

 

[racobb]

A man after my own heart! Good description.

Alan

Democracy is two wolves and a sheep deciding what to have for dinner. Liberty is a well armed sheep!
Ben Franklin

 

[BSEE1980]

You guys are right on: regulator problems.

My guess is your AVR does not have a PF/KVAR regulator.
If not, setting the droop higher might help.

We paralleled for years without PF/VAR regulators: we just set the droop higher. That is, more current, less voltage. As long as you stay in the capability curve of the generator, you will be fine. But those were big gensets.

You might have a stiff, moving, grid vs. a smaller genset. In that case, you will need to add a PF/VAR regulator or have someone manually adjust the AVR while watching a PF meter. And maybe change some components or your AVR. Some AVR's cannot accept a voltage reference, so you may need to add a motor-operated pot, for example.

Good luck
Steve

 

[nawao]

Fully agreed with what BSE1980 said:
"My guess is your AVR does not have a PF/KVAR regulator.
If not, setting the droop higher might help.
We paralleled for years without PF/VAR regulators: we just set the droop higher. That is, more current, less voltage. As long as you stay in the capability curve of the generator, you will be fine"
Set the droop higer and your problem will be solved.

 

[catserveng]

I have worked lots of sites where running in voltage droop between 3 and 6% worked fine while paralleled to grid, and a lot of sites it didn't work worth a darn.

Each site will have different dynamics, including voltage sag and swell, transients, and imbalance. Voltage droop is a simple method, but isn't a cure all. Neither is a VAR/PF controller, have seen sites where they caused their own problems. Some sites do well with active power factor control, mostly "stiff systems" in my experience. Some softer systems have done better in VAR control, some sites did fine with droop even though a controller was installed, we just didn't activate it.

While I'm sure operating in droop has worked for many people, based on the OP's information, just increasing the droop and calling it good may not solve his issues.

Best bet is to find someone experienced in DG systems to take a look at your system, see what you do have and how it's installed, check it tuning/adjustments, and get some experienced eyeballs on your system to make sure their ins't additional contributors to the issues you're seeing.

Hope that helps, Mike

 

[PMACP]

Hi,

The control cards that are installed on the generator are presented on attach. There is a A V R card and a cosfi/Var card regulated to cos fi control.

I don't know until this moment the set points of the cards. Do someone recognize those control cards? How do they the control?

Assume that the voltage on the grid increases. On that circunstances, the AVR card try to decrease the excitation of the generator in order to bring the voltage level to the adjusted value (as waross said before). However, decreasing the excitation imply lower generation of reactive power and, as a result, the cos fi tends to change unlesse the active power is also reduced. The active power is controlled by the power governator and, as long I know, it remains constant. So, on such circusntances, what may happen?

Thanks

 

[waross]

Can you tell us the manufacture of the controllers? Then those of us who are not familiar with those particular controllers may search for instructions on-line.
Thanks.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[racobb]

"Assume that the voltage on the grid increases. On that circunstances, the AVR card try to decrease the excitation of the generator in order to bring the voltage level to the adjusted value"

If the voltage on the grid increases, proper var/pf control would raise the excitation on the gen set to compensate, unless it is at its upper limit.

Alan

Democracy is two wolves and a sheep deciding what to have for dinner. Liberty is a well armed sheep!
Ben Franklin

 

[catserveng]

Is this an FKI AVR and Controller?

I have only worked on one of these on a shipboard system to parallel the ships's service generator to shore power. My experience with this controller was not good. Maybe my ignorance or unfamiliarity.

But we installed a Basler AVR and an SCP250 PF/VAR controller and the system worked fine after that, no reported problems after about 3 years.

You may want to see if someone from the forum from Europe may have more experience or information on your controller.

 

[nawao]

PMACP,
Find in the attachment a complete description of the reactive power/power factor regulator M50FA400A. I think, this document will answer all your questions

 

[PMACP]

Hi,

I have now some more data. My measuremnts show me that the reactive power that is supplied by the generator to the transformer are different on each phase. Moreover, the reactive power, at some periods, are capacitive on two phases and inductive on the other one.

Between the generator and the transformer there is some ancillary services (ie, the generator supplies some loads corresponding to ancillary services). This means that only a part of the generated power (the one that is not absorved by ancillary services) is injected on the transformer and then goes to the main grid. I checked that the ancillary services are reasonably balanced and so, the unbalanced reactive powers do not result from those loads.

The transformer has a voltage of 400 V on the side that is connected to generator. However, most the time it is working on higher voltage levels (420-440 V). Is that possible that the transformer may produce the referred effect on the reactive power due, for example, to satutation effects? The transformer has a wye with neutral - delta connection. The Wye winding is connected to the generator (with neutral connected).

Thanks

 

[waross]

Try floating the neutral. Leave the grounding connections on the generator. Normally a wye primary (power in) must have the neutral connected to the source. The exception is the wye primary:delta secondary. The delta secondary holds the phase relations ships of the wye and keeps the neutral point stable. Connecting the wye neutral and supplying balanced voltages forces the secondary (high side in this case) to have equal voltages and equal phase angles. However, if the high side voltages are not equal, you will have phase shifts on the high side. Now when you connect that to a transformer that is constrained to equal voltages and angles, you get circulating current transferring energy from one phase to another. There are a lot of ways to explain what the phase angle shifts do to the kW flow and the KVAR flow.
For instance, power flow depends on phase angle. When you open the governor to generate more kW, the relative angle of the rotor in relation to the utility phase angles advances until the power flow matches the new governor/throttle setting. When you have phase angle errors on the parallel system, you get different power transfer on different phases depending on the phase angle errors. The net power will still be governed by the throttle setting on the prime mover but one phase may be exporting more or less energy than the other phases. Even though the line currents may be moderate, the circulating current may be high.
If the voltage unbalance exceeds the %Z voltage of the transformer or bank by very much you may have overloaded transformers with little or no actual load. On a three transformer bank you may measure the circulating current in the connecting jumpers. In a sealed, three phase, high voltage transformer it is virtually impossible to measure the circulating current but disproportionate transformer heating is a symtom.
As I suggested, lift the neutral connection. Measure the voltage between the generator neutral and the transformer neutral. If this voltage is significant when compared to the %Z voltage of the transformer, unbalanced high side voltage and circulating currents are suspect.
Or, check the current in the neutral. If there is little current, move on to other solutions. If there is a heavy current on the neutral, suspect unbalanced high side volktages and proceed to disconnect the neutral and check further.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[PMACP]

Bill,

Hi. thank you for your very good tips. I think some of the problems I have are in fact related to what you said. However i need some more details if you can give them to me.

a) Have you some written information on that subject (some book, or a paper for example)?

b)You said that I should "floating the neutral". I have some current on the neutral conductor (50A while the phases are near 2000 A). This means that the currents supplied by the generator to the transformer are quite balanced! However, the ancillary services have some single phase loads, so I need to keep the neutral connected (or to create an alternative). Besides that discussion, at this moment I need to understand those "strange" unbalanced powers. That is why I need some more information.


c) you said. "However, if the high side voltages are not equal, you will have phase shifts on the high side. Now when you connect that to a transformer that is constrained to equal voltages and angles, you get circulating current transferring energy from one phase to another."

Well, because we are talking about a delta winding (MV side), those circulating currents only appear on that side. Is that true? Is there some way to catch some evidence of that phenomena on the LV side of the transformer?

d) You said "There are a lot of ways to explain what the phase angle shifts do to the kW flow and the KVAR flow.
For instance, power flow depends on phase angle. When you open the governor to generate more kW, the relative angle of the rotor in relation to the utility phase angles advances until the power flow matches the new governor/throttle setting. When you have phase angle errors on the parallel system, you get different power transfer on different phases depending on the phase angle errors. The net power will still be governed by the throttle setting on the prime mover but one phase may be exporting more or less energy than the other phases."

Well, in fact I have some unbalences on the values of active power. However, the more important unbalances takes place on reactive power. In some periods it occurs that one phase is absorving reactive power while the others are injecting it on the transformer. Is that possible? And, why are the unbalances more significant on reactive power? Also it is important to stress that in some periods the total reactive power increases significantly and it becomes greater than the active power.

e) You said: "On a three transformer bank you may measure the circulating current in the connecting jumpers. In a sealed, three phase, high voltage transformer it is virtually impossible to measure the circulating current but disproportionate transformer heating is a symtom."

I have no equipment to measure the currents on the HV side (due to voltage level). So I have a problem. However the transformer is, in fact, heating more than what is supposed to. You suggest me to "lift the neutral connection and measure the voltage between the generator neutral and the transformer neutral." Which could be the risks of doing that? Also you suggest me to check the current in the neutral. If there is little current, I should move to other solutions.

Which do you believe to be a significant current in terms of phase currents for instance?

f) Finally, what about a possible situation of transformer saturation?

Thanks

 

[waross]

Can you measure the primary voltages?
Can you disconnect the generator and measure the voltages at the 400 volt side of the transformer? (Use the transformer as as a potential transformer to determine if the primary circuit is unbalanced.)
With the neutral disconnected from the generator the delta winding should keep the neutral in place.
Your neutral current looks like about 3.6% of full load current.
Circulating current is limited by the transformer impedance so let's hope that the 50 A represents unbalanced loads and not circulating current.
Circulating currents caused by voltage unbalance are a classic case of actions and reactions. They tend to lessen the effect that causes them. Once the neutral is disconnected treat it as energized. There should not be much voltage between the generator neutral and the floating transformer neutral, but in the event of a ground fault on your 400V system the voltage on the floating neural may rise to lethal levels.
Floating the neutral allows the phase angles to align with any phase errors on the utility side.

Yes, saturation may be an issue. Is it possible to change the taps on the transformer? 440V is 10% high for a 400V transformer. +5% would be right on for 420V and more forgiving of 440V. (or maybe -5% depending on how you look at tap ratings.)

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[nawao]

In order to avoid unbalanced voltage on the primary side of the transformer, could you try feeding the auxiliary services from a separate transformer 15kV/0.4 kV delta-wye of adequate kVA. In this case you can leve floating the neutral on the primary as waross sugested

 

[PMACP]

Hi

"Can you measure the primary voltages?"

Yes, I have some values. I'm sending you an attach with some violtage, current and power values. As well I made some comments about the regime of operation of the generator-transformer set.


"Can you disconnect the generator and measure the voltages at the 400 volt side of the transformer? (Use the transformer as as a potential transformer to determine if the primary circuit is unbalanced.)"

Well, in fact I did it but during times when the generator is not deliberately disconnected from the transformer. at those periods, the transformer remaisn conncted to the grid in order to supply some ancillary services.


"Your neutral current looks like about 3.6% of full load current."

You can see some more detailed values on attached file.



"Yes, saturation may be an issue. Is it possible to change the taps on the transformer? 440V is 10% high for a 400V transformer. +5% would be right on for 420V and more forgiving of 440V. (or maybe -5% depending on how you look at tap ratings.)"

yes, it is possible, but I need to understand what happened.

Do you thing that primary overvoltages greater than 5% will create a saturation situation?

 

[PMACP]

nawao:

in fact I can't do it. I don't have an available transformer to do that. Note that the ancillary services current may achieve 300 A per phase.

Thanks