Generator frequency problem and kVAR overload

[AusLee]

Hello,

Please i have these 2 questions:

1. A 2,000kVA generator, loaded at 40%, is showing a variation in frequency + or - 1 Hz. We're trying to change the fuel, but what else can be the problem?

2. On another topic, i have a transformer and i want to put a cpacitor bank of a value equal to 5% of the transformer's kVA. My question is, i read that in some cases, when there is over compensation, there might be an overvoltage, could you please tell me where would the over voltage come from?

Thank you.

 

[Yuma]

To question 1):

Is the generator paralelled to the network? if so, the cause for the frequency oscillations would normally be the network.

If the generator is islanded, then the governor (turbine regulator) is probably the problem.

 

[AusLee]

the generator is a stand alone back up source not paralleled with the network. What is the acceptable range of oscillation?

 

[jghrist]

Voltage rise comes from capacitive current flowing through an transformer inductance.

% volt rise = kvar·%X_t/kVA_t

 

[AusLee]

jghrist, so for a 400kVA transformer, what is the value of the over voltage? is it sustained or momentary and does it affect lighting fixtures and televisions for example?

 

[Snoball]

If you can test load with resistive load bank you should have stable frequency regulation.

 

[powerwagon75]

In reference to #1: Is this a newer 1500 rpm genset, or something older/slower?
Modern electronic governor systems have their own quirks and methods of adjustments, as do the older hydraulic and electro-hydraulic systems. Does is swing rhythmically between the +/- 1 Hz, or does it seem to have minor surging to it as it swings back and forth? Does it do this continuously, or randomly? What is your AC amperage doing during these movements?
A continuous and rhythmic movement would often times indicate a fixed problem, such as governor or regulator adjustments. More random or sporadic surges/hunting is often times caused by an external influence to the governing or regulating systems.
There are a number of things that can cause these conditions. Fuel supply suction leaks. Restricted fuel or air filters. Voltage regulation circuit with oscillation troubles. Governor adjustments. Fuel pump/fuel injection issues. Check all linkages between governor actuator and fuel rack(s) (if applicable) for wear/proper adjustment.
Governor parameters will sometimes need to be compensated on old or high-usage equipment. I've always found it a good idea to record any governor parameter settings on your periodic maintenance documentation forms, for historical value, and a quick check to see if someone else has altered it to "their liking."

 

[AusLee]

The generator is brand new - 1500 rpm.

The oscillations are more rythmic than sporadic, the frquency stabilizes for a couple of seconds then it goes up and down.

I know that they have set "droop" to zero because they could not get the synchro between the two identical sets when it was ON. It could be though that it is a lack of competency from the worker but in the end result now "dropp" is zero.

Again, what is the acceptable range of oscillation? i cannot test it on resistive load anymore as it has entered in service, but it is still under guarantee.

 

[ScottyUK]

Is this one set islanded from the grid, or a parallel pair or group islanded from the grid?

If it is a single set, there is no requirement for droop and the set should be in isochronous mode. If it is an islanded group, droop control is required for load sharing otherwise one set will hog the load. Droop control is essentially isochronous mode with a load-dependent input to the speed control loop to reduce the speed reference signal fractionally as the load increases.

I would look at the governor valve and the governor loop tuning as probable causes of the frequency instability. The incorrect configuration of the droop controllers should be resolved first as this could be a contributing factor.

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

Sometimes I only open my mouth to swap feet...

 

[powerwagon75]

To answer your oscillation question--there should be no movement of Hz if the load is steady-state, whether in droop or not.

Does the problem genset only have this condition when in parallel, or all the time?

It sounds as if you're eluding to the point that there may have been some less than knowledgeable adjustments made. You should probably have someone study the book and make informed adjustments as needed. Check the settings on the other genset for comparison. They should be able to be close to each other for a starting reference.

If you try governor adjustments with no improvement, I would think the OEM should still be able to come out on a guarantee (warranty) service call and make things right.

By the way, is this a new condition, are has it been behaving this way since start up?

 

[jghrist]

The voltage rise depends on the transformer impedance. See formula in my previous post. This is for % voltage rise. There may not be an overvoltage because the load may cause more voltage drop in the transformer than the capacitor causes a voltage rise.

 

[e5fornow]

As POWERWAGON75 said check the governor linkage and make sure the governor is set properly. To check for air in the fuel you might want to have the engine manufacture install a sight glass on the fuel return line and check for excess air bubbles.

 

[AusLee]

jghrist: is the possible overvoltage situation transient or temporary?

for the generators:
1. the installation is brand new
2. the "unqualified" techniccian is from the OEM
3. the problem is the same whether i have any one or the two generators functioning. I get the power from the synchro panel and i read the fluctuation on the digital meter on the ATS.

the air inlets of the generator room have been calculated for 2 gensets at full load. i do not have that many kVA in demand, so i think the air inlets of the room are enough. Could it be that because the load is around 40%? or is the generator supposed to maintain stable voltage at full range?

What i want to ask here is a little silly. I know that for some dummy loads, it is possible to have an asynchroneous generator (which has large slip values). I read synchro in all the documents of the new group, can it be that there is a memory location in the controller in which the type is defined as asynchroneous - or the controllers for asynch gensets are radically different from contollers for synchro?


we will change the type of the fuel and see what it gives.

 

[ScottyUK]

The generator should maintain stable voltage regulation over the full range of zero to base load.

Don't assume that the OEM tech knows everything! If the droop configuration is as you described earlier then you have an error in the configuration when the two sets run in parallel. However if the problem exists when only one set is running then this is probably a secondary concern.

We installed a new Cummins set a few years ago which had what I can best describe as a 'sticky' governor valve which caused a definite frequency fluctuation. It calmed down once it had bedded in and the frequency stability is now very good. That engine (an NT-855) had a lot of teething problems which the factory commissioning tech missed or ignored. The local service depot and our engineering team resolved the problems. During the hours of test running accrued while fault-finding the frequency instability settled right down. When I queried this behaviour, the tech from the local depot said it "wasn't uncommon" and that it usually settled down in the first hundred hours or so like ours did. The governor valve is a highly integrated block with no external linkage. I remain unconvinced whether the problem was really due to the valve or whether the whole engine was tight when new and needed to run under load to bed in. In my experience sick control valves usually one head in one direction.

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

Sometimes I only open my mouth to swap feet...

 

[AusLee]

The engines atre from MTU, the generators have 400 and 300 hours each, the problem is general, whether i have 1, 2 or both gensets.

The droop was disabled because:
in the early testing stages, one geneset kept generating more and more kVARs and the other kept receiving, it was progressive until a certain relay tripped in protection. when droop was disabled the synchro was very fast, like 1 or 2 seconds. this was done by the local supplier.

later, a guy from the manufacturer came and found that the CTs on the circuit breakers were short circuited (they supply them this way for conenction on site, the local guy made the connection but did not remove the strap). we were reading 0 current on the meters in the synchro panel and that solved it. he left the droop disabled, to the best of my knowledge.

i know i can read droop settings on the generator control panel, i will go a verify, if it is still disabled, what will i gain from having it back?

 

[ScottyUK]

Hi AusLee,

Are we talking AVR droop or governor droop? The former is intended to even out reactive load sharing, while the latter is designed to balance up real load. The CT's would most likely form part of the AVR quadrature current compensation. From your description of the symptoms the AVR's are fighting each other - perhaps a QCC CT reversed? - or the droop settings are not of the same polarity effectively causing one set to shed VAr's as the reactive load increases.

One set isn't running in power factor control and the other in voltage control by any chance? PF control can cause some really weird behaviour on an islanded system where the load determines the PF but the generator tries to maintain a constant PF seen at its terminals. If connected to a larger network which can absorb or deliver VAr's this problem is less significant, although it can still lead to voltage regulation problems for the utility if you have a high reactive import or export.

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

Sometimes I only open my mouth to swap feet...

 

[waross]

It sounds like at least two problems.
Problem one. Load sharing;
It sounds like your quadrature circuit was not connected originally, and then when it was connected, it was disabled.
This circuit is used when sets are operated in parallel to share the real power and the reactive power.
If the quadrature circuit is connected in reverse, I would expect strange results. You can try shorting the quadrature CTs. If the problem gets better they may be connected backwards. From your description, it sounds like a reversed connection on the quadrature circuit.
If you wish, I can try to explain the operation of the quadrature circuit, but I don't want to waste the time if you don't need the information. It's basically a CT and a resistor.
In a correctly wired system, there are quadrature shorting switches on the control panel to disable the quadrature circuit when only one set is running. The switches may be labeled "Single" and "Parallel". There will be one for each generator.
I have seen operators forget to turn on the quadrature circuit when adding a second generator and not have problems.

Problem two. Frequency variation.
I had an ongoing horror story with a Cummins 855. I did the original install and it ran fine for over a year. Then the dealer called the owner and said that their records indicated that the set was due for a dealer inspection. The dealer techs destroyed the electronic governor and it had to be replaced.
A year later, the dealer called and said that the fuel valve needed service. (This was a standby set with very low hours.) After the "Rebuild" the valve was sticky and would not close. As the load was applied, the governor would increase the signal to the valve as much as required to overcome the stickyness, and pick up the load. When the load reduced, the governor would reduce the signal to the valve, but the return spring was not strong enough to close the fuel valve and the engine would overspeed until the high speed safety shut it down.
I have seen numerous governor problems on other sets after a boat mechanic rebuilt the engine on a gen set.
I saw a 60KW set with a badly adjusted governor that wouldn't carry any load. It would slow down with only a 5 KW load. The engine was completely rebuilt. Same problem. Several turns of the deadrack adjustment in the governor fixed it.
The point;
A lot of good diesel mechanics and technicions are completely clueless when it comes to the governor on a gen set.

I know that for some dummy loads, it is possible to have an asynchroneous generator (which has large slip values).

I think that you are confused with induction generators.
You have synchronous generators.
The governor may be configured for "Droop" mode and will have a frequency variation from no load to full load of from 1% to 3%. Typically the set will run at 51.5 hz. at no load and 50 hz. at full load. (3%)
The governor may be configured for isochronous mode, and the frequency will change slightly as the load changes, but will quickly compensate and return to 50 hz.
The set may be in isochronous mode and responding to large load changes.
The governor or the governor linkage may be sticky.
Single sets may be run in isochronous mode or droop mode. Parallel sets should be in droop mode in a basic system.
If you have a computer controlled load sharing panel, the panel will make its own decisions.
respectfully

 

[jghrist]

On the voltage rise with capacitors on a transformer secondary (perhaps this should be a separate thread because it does not relate to most of the posts):

is the possible overvoltage situation transient or temporary?

The %rise equation I gave was for steady state quantities. It does not address possible capacitor inrush when energizing.