IET
Decrease font size
Increase font size
Topic Title: Power quality issues on prime power generator
Topic Summary: Harrmonics in a generator supplied power system.
Created On: 13 September 2013 02:38 AM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 13 September 2013 02:38 AM
User is offline View Users Profile Print this message



Pat Lucey

Posts: 29
Joined: 18 May 2006

HI all. Take a look at the following and make some comments.
Thanks.

We have custom built 2 x 300KW sets with Stamford alternators powered by Detroit electronic governed engines. These are 600V 3 phase machines with MX341 AVRs with PMG. A Dual prime ATS to allows one machine online and the second as autostart backup. Everything's working good other than unstable voltages.......
The sets in turn power a new ( last season ) very remote lodge. .
We get unstable voltage and engine hunt in some situations. There is also a 100KW set ( for off season ) which can be made to do the same. This old set did not hunt with the old smaller lodge. We have checked and followed makers AVR stability adjustment. We have tried different droop settings on the engine ECMs for the big sets. Did not change things as far as we could tell. We have done some power quality measurements with a Fluke 43B. Voltage THD was about 2.0 %. Ran out of time to do more in depth testing. The power comes from the sets down a long cable run to Delta / Wye transformer in the lodge. ( for 208V / 120V). We are only 20 to 45% loaded. We are able to make the system unstable by turning on just a small ( 120V, 12A running ) freezer unit........which sends things off... Not really many non linear loads in the lodge .
We are thinking the AVR on the set is only sensing 1 phase and needs to be changed out for a 3 phase sense one.
The electrical contractor who built the lodge is working on the problem but seems a bit short of knowledge..( ....never heard of non linear loads..? ) .....Stamford says we can fit a MX321 with isolation transformers which will be better....but I feel there is way more to track down at from the transformer and in the lodge itself. Fluke website gives some measuring PQ tips to track stuff down ( ...transformer K factor , Neutral wiring and ground currents... ) but not really familiar with the PQ thing.

So people please give some feed back. Just hoping some of you can throw some thoughts as to what might be going on.

Thanks Pat.
 27 September 2013 05:23 AM
User is offline View Users Profile Print this message



Pat Lucey

Posts: 29
Joined: 18 May 2006

Thanks for your your help.
 27 September 2013 07:44 PM
User is offline View Users Profile Print this message


Avatar for sparkyaj.
sparkyaj

Posts: 57
Joined: 04 June 2012

Good evening. If you would like to drop me an email - I would be more than happy to give you some assistance. Kind regards, Alastair

sales@ajfinch.co.uk
 02 October 2013 06:48 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

Hallo Pat,

hunting of any system is a symptom of poor design in a servo system. All servo systems consist of a negative feedback arrangement whose purpose is to regulate whatever parameter is being measured; in your case the output voltage from the generator is being compared with a given standard to produce an error signal which is then applied to control (usually) the prime mover. However if either too much feedback is applied, or insufficient attention is given to phase problems, then negative feedback is only too willing to change into positive feedback which is a prime requirement in any system of oscillation - otherwise termed hunting.

In electronic terms the cure is to adjust the phase conditions around the feedback loop and, for myself, I've always found this all but impossible when dealing with a completed rig such as a motor-generator set. I have a small generator which is kept as a standby to keep things like refrigerators going during breaks in the power supply. It has always suffered from serious hunting and for which the only cure I have ever found is to keep a basic load on the generator - which of course cripples the poor beast in its prime function.

Surely the basic problem which you face is that the designer (s) of the set suffer from a common fault in that they think it is a simple matter to maintain a constant output by wiggling a control - not true! I believe that, unless you are prepared to make a study of servo systems and tear apart your Demon, the best cure and probably in the long run the cheapest cure is to buy a generator set which works.DEMAND OF THE MANUFACTURER THAT IT IS EITHER CORRECTED OR THE SUBJECT OF A REFUND.

Ken Green
 02 October 2013 06:56 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

second thoughts on the above posts.

It is possible that there is a spring loaded link to the throttle of your prime mover. Sometimes the link is a solid strip of metalat one end of which you will find several small holes - these enable you to vary the tension spring which effectively is a variable-damping arrangement. You will find this kind of link on almost every power grass cutter; if you're lucky the best of luck; if you're not lucky then you can tear out your hair!

Ken Green
 02 October 2013 06:56 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

second thoughts on the above posts.

It is possible that there is a spring loaded link to the throttle of your prime mover. Sometimes the link is a solid strip of metalat one end of which you will find several small holes - these enable you to vary the tension spring which effectively is a variable-damping arrangement. You will find this kind of link on almost every power grass cutter; if you're lucky the best of luck; if you're not lucky then you can tear out your hair!

Ken Green
 08 November 2013 01:29 AM
User is offline View Users Profile Print this message



Pat Lucey

Posts: 29
Joined: 18 May 2006

Thanks Ken for your thoughts.
The issue is still not resolved. We have put the makers Ultimate AVR system on one of the sets. ( Three phase sensing transformer to MX321 with original PMG exciter. ) This set like the others still hunts. We did resistive load bank testing on site which told us nothing new. The set performed perfectly as they did at the shop before installation. There will be more PQ testing at the lodge this coming week. Time is moving on and there is snow at the lodge.. Heli ski starts at this lodge in December. The sets ran last season with this problem.
The electronic engine speed is governed straight from the engine computer so there is no linkage to adjust. There is the correct industrial generator program in the engine computer.and there are not any engine or fuel issues. The AVRs are adjusted to makers settings.

Would be nice if more people who know PQ stuff replied to this topic.
I guess we all get too busy....

Thanks Pat.
 08 November 2013 06:11 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

hi Pat,

The more I think of this the more convinced I become that you are involved in an eternal problem which besets anything which operates within a closed loop; indeed diagnosing faults in such setups involves a knowledge of how to open the loop without destroying it?

What is not generally realised is that hunting is a prerequisite of any feedback system; i.e. even if your sets are operating correctly they are, of necessity, still "hunting"! The phenomenon is a result of system inertia wherein an action cannot cease at the instant of removing the control and neither can it commence at the instant when the control is applied. As a consequence the control system (the servo system) constantly overshoots and undershoots in a never ending sequence.

To put that in non-mathematical terms unless there is an error then there cannot be an error signal; without an error signal you cannot derive a correcting signal; without a correcting signal you are up the creek without a paddle. If you're servo system operates as ideally required then there will not be an error - without an error you have no control?. As a consequence any closed-loop solar system always oscillates about the required position even though it may not be detectable either by your body senses nor by your instrumentation. The amplitude of this oscillation is governed by the "loop-gain"; i.e. the gain around the feedback loop from the point of error measurement to the point at which the control is impressed.

The instant reply to this of course is increase the loop-gain? Unfortunately we have to work with practical components and this introduces "back-lash" otherwise known as "time delays" or in electrical engineering terms "phase problems". In other words you just can't win!

Again, in electronic terms, you must turn your system into one that is critically damped which is techno-speak for a bandpass arrangement; but beware - such is bedevilled by phase-shifts at the edges of your passband. Hence you must arrange that your pass-band is greater than the operating (mechanical) range of your machine. That is where it becomes difficult, if not impossible, to correct a design fault except by viciously damping the mechanical system; i.e. by keeping a constant resistive load on the generator unit.

I hope this is making sense - unfortunately I have to rely for writing with a computer application called "the Dragon". This little marvel occupies something around 2.5 Gb of my hard drive but is still cannot escape its roots all the way from the other side of the pond - yes, it does use an English dictionary (under protest) but it is firmly dedicated to American syntax and punctuation and also is given, when puzzled, to taking a running jump which invariably lands it neck deep in the brown stuff! I hope I have removed at least most of its hysterical guesses?

Ken Green
 08 November 2013 06:56 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

hi that, me again,

Have re-read that last post and I Spy boo-boo!

A critically damp circuit is one which is not able to exhibit resonances but which behaves over its designed band as though it were a resistor. Described in mathematical terms such as circuit does not contain a term which involves omega. This is another poorly understood circuit arrangement but this is not the correct place in which to expand.

If you would like to take it further send me a PM.

Ken Green
 09 November 2013 12:43 PM
User is offline View Users Profile Print this message



dlane

Posts: 685
Joined: 28 September 2007

Hi,

You described your system as having a long cable run from the generator output to the lodge. How long is this run and what is the cable type, I am wondering if this is creating a capacitive loading on the generator and pushing it over into a leading power factor which may be the cause of your voltage stability issues. Is it possible to carry out measurements on the generator output safely?

Your THD of 2% is acceptable, IEEE limits are 5% for THD and 3% for individual harmonics. Was this measured on the generator output or on the transformer secondary? I also don't believe that power quality measurements will be of much use to you, all THD is telling you is how much distortion there is in the sinusoidal waveform.

I am not familiar with your engine type but I believe your AVR can be set up for manual control. You could try having the AVR in manual and run through the tests to see if the engine starts to hunt and if possible then set your engine speed manually and put the AVR in auto and run the same tests. That way you may be able to establish if the AVR is the primary fault that causes the engine to hunt or if the governor is causing the AVR to adjust voltage output.

By your desription you are running one set at a time and you do not have sets operating in parallel which could give you circulating currents under some connection arrangements. You would also want power factor control enabled in your AVRs for parallel running, this should be disabled for running sets on their own.

Kind regards

Donald Lane
 20 December 2013 05:16 AM
User is offline View Users Profile Print this message



Pat Lucey

Posts: 29
Joined: 18 May 2006

Hi Don thanks for your thoughts.

The cable run from the generator building is just over 1000 ft. There are two cables in parallel. They are aliuminuim 3 core Tec cable. ( BC, Canada ). We have thought about the capacitance loop thing between the two cables but it was felt that the cables should be close to the same length. The installer did not measure out to make the cables exactly the same but the cable runs do not have many turns in them. I think we would be able to safely run just one of these cables just for testing. Easy to disconnect at the either end and run with just one cable. ( Generator is lightly loaded and generally see 50 % load....due to the builders insisting on 250 kW PRIME when specs drawn up for the generators. ).

The THD was measured at the generator end by us and later at the Secondary of the distribution transformer by the lodge builder / Installer.
Both seemed to be OK.

I think it would be possible to do a manual excitation of the AVR. However would not be able to do this with guests in the lodge. They spend way too much money for their Heli skiing to have me put the lights out...

There is only one set running at any time. The second one is autostarted and will then power the lodge. ( Dual Prime.)

Thanks to people who have thought and replied about my problem.

Merry Christmas and all that.

Pat.
 20 December 2013 09:44 AM
User is offline View Users Profile Print this message



christait

Posts: 24
Joined: 08 February 2013

Hi Pat, have you looked at (or do you have) resitance grounding at the generator, direct coupling to ground at the generator end can cause voltage instability? It might be worth exploring HRG or LRG.
Merry Christmas.
 02 January 2014 04:06 PM
User is offline View Users Profile Print this message



grahamstyles

Posts: 9
Joined: 07 June 2003

I have experienced a similar problem with temporary camp generators in the desert .Like you we went through every electronic control setting we could and still the same problem of hunting. Our gennies were powered by Detroit Diesel (supercharged 2 stroke V12 800 Kva) that were rarely loaded above 50%, the problem was found to be the bores of the prime mover (piston bores/cylinders) had glazed as they were never worked hard enough to keep the bores clean as such power output of the prime mover was greatly reduced and the governor was constantly trying to match engine power to the generators load requirements and hunting. A strip down of the engines and 5 minutes with a honing machine on each bore returned them to new and no hunting. After that we kept the load as near to 80% as we could and had no more problems. One generator in particular had not run for more than 6 Months before it glazed its bores.
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.