Power Factor & Regulators 2

[PAFred]

Can someone shed some light on the possible or real effect's of inductive loading of a PM generator set that is running with loads that have power factors from as low as .20 pf to .36pf on average. And whether or not this would be the most likely cause of regulators going bad every 30 to 45 days this unit currently is on the fifth regulator.

The generator is the sole electrical power for a stone processing plant, There are two separate supplies to the plant both run through either variable frequency drive systems or soft start systems.

The max gen set KW is 1500 @ 480VAC. At the time of shut down today the load was approximately 1150 amps. The monitored DC field current output maxed out at 20 ADC.

This condition is happening randomly and once started it may run all day ok , 15 minutes or an hour before shutting down on a high voltage situation. I captured a reading of 668VAC.

 

[itsmoked]

Others will have more info.

Realize that with such a bad pf the generator has to pump out way more amps that usual at any particular load. I suspect that the VR is being stressed in that it is putting out an overloaded field current essentially continuously. This is why the VRs are dying. I would run the system with the VR exposed then use a temp gun to shoot the entire board after 5 minutes of nasty pf. I would suggest that any devices demonstrating temps over 70C as being thrashed. If research shows this then either examine the parts as possibly being upgradable (power wise) or put a high velocity fan in to cool the board/components.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hi PAFred
I take it that your set is rated at 1500 Kilowatts, 480 volts. At unity power factor that would be 1805 amps.
Sets that size are commonly rated at 80% PF. that would push the rated current up to 2256 amps.
I take it that you have a PMG set. You can run this with a conventional voltage regulator in an emergency. Expect more voltage drop when starting large motors.
Your problem.
What is the power factor at the generator?
With the field current maxing out, you may have a field winding or an excitor winding failing. You may have problems with a diode, or a MOV.
If you have a high excitation current and a high output voltage, this may be a result of the regulator failing rather than the cause.
The current pulses drawn by the VFD may be causing inverse voltage spikes on the generator that are adversly affecting the voltage regulator.
One possibility is that a current pulse to a VFD causes a corresponding voltage dip at the generator. The Regulator responds by turning an scr on, but cannot turn it off when the peak passes.
The high voltage reading would be more meaningful if we knew what kind of meter it was measured with.
A low PF on a motor at a stone plant implies a light load. I would suspect the VFDs before the low power factor loads. The power factor at the genset is the important factor.
A current of 1150 amps seems conservative for your set, and this current will include the effects of the low power factor devices.
There are others on the forum who may be able to make other suggestions as to the effect of VFDs on the generator.
respectfully

 

[itsmoked]

So if a 1,500kW 480V generator is designed with .8pf in mind, then 2256A is the maximum current.

Then a reasonable maximum load at 0.2pf would be 451A.

PAFred's stated load was 1150A. Could this be a problem?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hi keith;
2256 is the maximum current at any power factor.
Multiply that times voltage to get the VA.
Multiply the maximum current by the voltage by the power factor to get the watts.
Put another way, 1150 Amps times 480 volts times 1.73 divided by 1000 equals 956 KVA.
At 0.2 PF that would be 191 KW.
I think that the low power factor is a couple of fairly small devices, not the entire load.
For instance, if a motor was running at 0.2 power factor, we would expect it to be running at about 25% load. If a motor current at 0.20 was 1150A we would expect 1150A/25% = 4600A full load current. The generator would not be able to run such a motor, let alone start it. As this is ridiculous, I must assume that the total load is at a much better power factor. In any event, I suspect field problems or VFD induced problems.
Respectfully

 

[davidbeach]

waross, your calculations only work in absence of a generator capability curve. At low power factors the generator will not be capable of producing rated kVA, in fact below the rated PF, you can no longer get the rated kVA. The generator does not have a current capability, current is what it is as long as you stay within the capability curve which is defined on the PQ plane.

 

[ScottyUK]

Hi waross,

2256A is the stator current limit, and on a capability curve it is a semicircle centred on zero MVA, zero MVAr. There other (main) limits are: prime mover MW limit (engine runs out of puff); rotor thermal limit (field winding insulation overheats); stability limit (field is too weak to maintain synch with grid under transient conditions); end winding thermal limit (operation at very low leading pf).

The following links might be useful - the operation of the generator is well explained in the first link. The capability curve on page 49 of the second link is quite good too. Note that these curves are purely for the generator: no account of the prime mover mechancial limit is shown, hence the curve does not have the usual flat top shown on a combined prime mover / generator unit.

Warning to dialup users: these are quite big files!

http://canteach.candu.org/library/20030801.pdf

http://canteach.candu.org/library/20030804.pdf

I e-mailed a copy of a marvellous old CEGB document giving a really detailed method for constructing capabilty curves to some of the members a while ago - perhaps ItSmoked could forward me your e-mail address and I can send a copy by return?

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

I don't suffer from insanity. I enjoy it...

 

[itsmoked]

I stand ready to serve..

Interesting generator links Scotty!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hello davidbeach, ScottyUK, and itsmoked.
Thank you sincerely for your kind contributions to my continuing education. I now have reading material for quite awhile into the future.

PAFred;
Can you clarify please, is the low power factor measured at the generator, or is it the power factor of some of the loads? Do you know what power factor the generator is running at?
Can you tell us something about the voltmeter that you used to measure the 688 volts?

davidbeach and ScottyUK;
If the 20% power factor at the generator is going down to 20%, would it be safe for PAFred to add a couple of hundred KVAR of capacitors to the system.
My suggestion is based on the following points.
a> The condition is either borderline or intermitent. Most of the time the system works.
If enough capacitors are added for 100% correction, PAFred will probably have problems with leading power factor at light loads and at start up and shutdown.
I propose enough capacity addition to move the system into less of a borderline situation.
b> With a current of 1150 amps and a worst case power factor of 0.2, the KVA will be 956 KVA. Let's round it to 1000 KVA. power will be 200 KW. KVAR will be 980 KVAR.
If we add 200 KVAR the power factor goes from 20% up to 25.5%
If we add 300 KVAR the power factor goes up to 28%.
If we add 400 KVAR the power factor goes up to 32.6%.
This is speculative until we hear from PAFred as to where the low power factor is measured.
Thanks again
Respectfully

 

[RalphChristie]

Nice links, Scotty
PLS

 

[ScottyUK]

Thanks Ralph. I accept it on behalf of the author.


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

I don't suffer from insanity. I enjoy it...

 

[waross]

Hi Scotty;
Another lps.
As I understand the curve on page 49, It shows the reduction in generator capacity with a reduction in the pressure of the cooling hydrogen.
With a relatively small air cooled set I assume that we could use the outermost curve, coresponding to full hydrogen pressure, or normal cooling. On page 48, there is a good generic capability curve. It shows quite clearly the reduction in capacity due to field heating.
However, PAFreds information seems to indicate that the generator is operating at about 51% of KVA capacity. This seems to be well within the reduced KVA capability shown on the curves on page 48 and 49.
Am I missing something here?
respectfully

 

[PAFred]

Thank's (Waross) for the time to reply. I went back to the site today and found out that the gen set is feeding two different types of soft starts. One is a transformer type and is in the circuit all the time the second is an SCR type that is supposed to disconnect from the SCR's once up to speed and go into by-pass mode.

In clarifying the gen set for you it is a Caterpillar unit 1000KW (not 1500 as I'd thought), 480 VAC , 12 lead , 60 Hz , 1506 AMPS @ .8 pF and the generator is a PMG type.

The power factor that I'm seeing was measure at the generator output breaker. As for the brand or quality of meter I used to record the output was from a Summit Technology (Powersight PS4000) which if you aren't familiar with is an extremely good meter for the money. If you would like to know more about this meter or possibly purchase one use this link to go to the page that has information, you can also scroll down to the bottom and click on the PDF file link for more detailed spec's about it.

LINK:

http://www.powersight.com/ps4000.shtml

I made up my mind today that I was going to take every single wire apart and or disconnect each wire and physically clean the dirt off and while doing so check for and abrasion's, nicks', cut's or possible grounding and while at it disconnect all generator coil output wires and megger both between each coil to chassis as well as each other to see if there was any problem in that area.

Having done this I didn't find anything that I was looking for. So I'm stumped as to why this thing is doing what it is. I cant think of anything else to look for.

All I can say is that tonight I'm popping the top on another stress reliever.

 

[waross]

I am impressed by your meter.
A couple of other things to check.
Inside your machine, behind the main field and the main windings, you will have the rotating exciter field.
You will also have a rotating diode plate with six diodes and one or two Metalic Oxide Varistors or other type of surge suppressor.
You may want to disconnect or short the diodes and megger the exciter winding to ground and the main field winding to ground. While you're at it you may want to check the diodes and surge suppresors.

I have a suggestion, but I prefer some comments from other members of the forum as to whether it may be a worthwhile test.
I think that your instrument will handle DC. Even if it won't do DC, it may well be able to pick up transients on a DC line.
Try monitoring the DC output current of the Automatic Voltage Regulator to see if is doing anything strange.
Look at the generator output as well for transients and strong harmonics.
You may get a better harmonic reading on current than on voltage. A large current peak may cause a corresponding small voltage dip. The AVR may respond to voltage dips with increased current output. Maybe not. It depends on how the AVR is damped. If you can see anything strange coming out of the AVR, try small adjustments on the stability control.
These comments are subject to comments from our friends.

How many pieces of gold do you need for a meter like that?

You can probably, safely add enough capacitance to the transformer to supply the magnetizing current.
Measure the no load amps of the transformer and select a capacitor bank that will draw the same current or less. Remember that capacitor current increases as the square of the voltage and do your calculations at the actual voltage, not the nominal voltage.
You may want to check the voltage coming from the PMG and going to the AVR. It will probably be about 220 volts three phase. This is the power supply for the AVR. The connections to the generator windings are for sensing the voltage.
Respectfully

 

[PAFred]

Hello again and once again your input is appreciated. As to the DC amperage output from the regulator it was running at between 4.6 to 5.7 respectively based on the loading. The generator regulator exciter rated max DCA is 8.6 amps.

Oh, regarding the meter I spoke about, if you go to the main web page you can find different meters, the PS4000 is the top of line I think (or at least it was) the whole kit which includes calibrated flexible CT's , Deluxe test leads , Software to monitor and or control the metering while testing and capture wave forms and readings including the case was just about $7000.00. A far cry from anything else on the market that even comes close to what can be done with this meter. I happen to know the owner / engineer that designed it. He has added some things over the years based on imput from me on testing requirments regarding generator set's. A very fine gentelman and he is very receptive to input regarding the addition to his meter that will help to make it serve the service industry further. The owners name is "Ken Kios" the company is in California.

Anyway back to my problem, the exciter is three phase and is putting out 240 VAC to the regulator. I cant rememder the voltage readings while under load regarding the field circuit but they were staying within spec's.

I noted that while running the loads would jump up to as high as 1150 some amps but only for small durations of time. I did note that on at least two occassions that the generator went into high voltage shut down when this load hit it. I also noted that it seemed that the power factor improved when more load was added. Of note this generator is a sole source power for a remote stone / concrete crusher plant. The generator is being used as a primary source rather than a stand-by which is what it was built for. As a try-it yesterday I put a resistive load bank in the system and added up to 100 KW while running under load to see what the affect would be (bring the pf back in line or worsen it) and I found that although the power factor was improved from an average .20 / .29 to a .67 it still didn't stop the random and imeadiate rise in ACV to a point that tripped the high voltage relay thus shutting down the gen set.

I did forget to megger the filed circuit of the exciter and exciter stator today while there in the dirt bowle which now that you've reminded me really tic's me off so now it's time to pop another top!

I'm beginning to think that the problem may be a situation of the regulator being hit with large amp loads and reacting too quickly causing voltage overshoot before it can recover because the load was removed as fast as it appeared.

 

[buzz41]

Waross:"Remember that capacitor current increases as the square of the voltage..." Typo?
I think you meant VoltAmpsReactive increases by the square of the voltage.

Good comments on this prob though. And, yes, nice meter. $$?

I may be bugging you power side guys soon. Re: VSD's and soft starts for a little water co.

Ron

 

[waross]

Hi Fred;
A couple of terms that we use differently.
Where you use the term exciter, I think that I would call that the PMG, or Permanent Magnet Generator. The tip off is the 240 V AC Three phase. The voltage will not change much with loading. Useually if you have over 200 volts, three phase it's OK.
This supplies the power for the voltage regulator. The voltage regulator rectifies and controls this and sends a DC Current to the Field of the Brushless Exciter. The PMG will be on the back end of the main shaft outside the generator housing. The brushless Exciter will be on the main shaft inside the generator housing. The brushless exciter supplies three phase AC to the rotating diode plate which rectifies it and sends it to the main field.

Power factor: Yes you are right. The power factor of a motor improves with loading. The addition of the resistive load would improve the combined power factor.

I suspect that you have a low slip motor on the crusher. A rock crusher should have a high slip motor.
Can you tell us the rated speed on the motors? More information is great, but just the rated speed will give an indication of the type of motor.
It sounds like your generator is set pretty "firm" also.

What can we do?
First, if one of the motors is on a VFD, it can be set to current limit the motor and help mitigate the current peaks.
If you have a VFD on one of the motors, give us the specs. There are a couple of VFD experts who can help you with the settings so as to reduce current surges.

AVR. I would try dialing in more stability.

UFRO Under Frequency Roll Off. Make sure the jumper is set for 60 Hz. This is the feature which reduces the voltage so as to reduce the loading when the engine on a set is overloaded. Typically when the frequency falls below 57hz. the AVR will start to ramp the voltage down proportionally to the frequency. This is a form of load shedding to allow the engine to recover. However, I don't think this will be a factor. You are drawing current but not loading the engine.

Governor. You may have 3% droop or even isochronous operation. Try setting the governor to droop control and if it is on droop control, increase the droop. This allows the generator to slow down when hit with a heavy load.
I understand that the load consists of large motors in a stone breaking operation.
If you can accept voltage dips in the office, start at 5% droop and possibly add some more. Your gut feel will be a pretty good indication of how much you can get away with.
This will allow the generator to slow down as the load hits. Then more of the energy surge in the crusher will be supplied by the flywheel on the crusher. The load on the generator will be less severe, but longer lasting as it accelerates the flywheel back up to speed. I would start at 5% droop and increase 1% at a time. The speed drop will be proportional to the KW, not the amps. However, your observation that the power factor improves when the load hits indicates that there is a fair amount of KW involved in the surges.
Power factor correction. Most of the current surge is real power current. Improving the power factor will reduce the peak current, but it will reduce the no-load current even more. If the problem is being caused by the AVR's reaction to sudden current surges the difference between normal current and surge current may be greater with corrected power factor.

The next step is series reactors to reduce the current surge. I hope we won't have to go there. It will be much more expensive than the other remedies.
I hope adjustments to the AVR and governor will correct the problem.
I have some more thoughts and suggestions, but it will be better to get some feedback from you and other forum members before I get too creative.
Re; The PS4000. Can you display the wave-form of the AVR output current? If so, you can compare the wave form with the characteristic wave forms of open and shorted diodes in the rotating diode plate.
respectfully

 

[itsmoked]

I still want to know why PAFred has fried 4, (four), IV, voltage regulators.. The discussion is on better(functional) control but is missing the failing VRs aspect.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hi Keith;
I haven't forgotten the voltage regulators. We have determined that the current surges are creating voltage surges. The unit has been shutting down on over voltage. That implies that the voltage regulator is over controlling, or being confused. A current surge creates a voltage dip on a generator. The AVR overreacts and you have a voltage surge.
Overvoltage shutdowns caused by transient events are a good candidate to fry AVRs.
I hope that the AVR settings will help mitigate the AVR output swings.
I hope that the governor tweaks will reduce the current swings that may be causing the voltage dip/surges.
I hope that some of the VFD gurus will help us further reduce the current surges when PAFred posts the VFD information.
I guess I have been sort of jumping to the end without explaining the in between steps. And that's one AVR, and anudder one, and annuder one, and anudder one.
respectfully

 

[itsmoked]

Ahhh the special VR fuse units.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com