Alternator Overvoltage!

[kevd]

We are commissioning a 25MVA, 4160V brushless 2-pole alternator driven by a gas turbine at 60Hz (3600rpm). Today at with the gas turbine at synchronous speed and alternator off the bars the excitation was powered up for the first time. The AVR/excitor commissioning team set the manual excitation on its lowest setting, but when the excitation was switched on the excitation power convertor fully phased forward and the terminal voltage shot up to 1.5 x rated voltage. The excitation was shut off after 5 seconds and the set ran down to stop. It appears, according to the AVR commissioning engineer, that the phase rotation from the alternator PMG into the power convertor was wrong, hence the power bridge phased forward out of control. The gas turbine engineers reckoned we drew 2.5MW out of the turbine during the incident.

My question is this?

1. What damage could we have done to the set (insulation, rotor heating, etc.)?
2. What checks should we undertake to ensure the alternator is OK to resume commissioning?
3. How can we draw 10% power from the prime-mover when nothing is connected to the stator terminals, i.e. generator open loop?

Any help appreciated.

Kevd

 

[edison123]

Is yours static excitation system ? If you were inputting the excitation transformer from your generator (as against from the external source), then the excitation power was supplied by your GT. Check your generator rotor power to know the excitation requirement. But normally for static excitation system, external power is used (called field flashing)in the initial stages before changing over to generator supply. With external supply, no power is given from your GT in except for rotational losses. Anyway, your GT is rated for 25 MVA and is designed to give the excitation requirement during regular operation.

I do not see how PMG Phase rotation could affect the AVR operation. Normally, AVR's allow voltage swings up to 10 to 20 % during initial switch in but bring the voltage within the level very fast (1 to 2 secs). If your AVR persisted with 1.5 times rated voltage, then your AVR is suspect and should be checked for proper operation and reference voltage settings.

New generators are hipot tested at twice rated voltage + 1 KV. So, you need not worry about 1.5 times. However, you could megger the stator and rotor for record purpose.

 

[boomsystems]

What was your overvoltage protection doing when voltage shot up to 1.5V.

The 10% power drawn is due to overexcitation of generator. It is common.

This is purely an AVR ( Excitation system problem). If PMG output voltage signal is reversed AVR can produce excess excitation. It has a blocking circuit and if not, fuses on thyristor circuit should blow. Under such condition the firing co-ordination is lost and unpredicted current can flow through the circuit

Nothing will happen to generator within 5 second. Let AVR commissioning engr. check his panel. Megger of generator is enough.

 

[abeltio]

1. What damage could we have done to the set (insulation, rotor heating, etc.)?

Any off-design condition takes a bite out of the parts life... e.g. an overtemperature in a GT will reduce the useful life of the hot gas path.
How much? Who knows... the point is that the incident occurred during installation so what i would do is:
1.a. Request a full test of the insulation (see answers to question 2).
1.b. Request an extension of the warranty for the insulation of the Generator.

2. What checks should we undertake to ensure the alternator is OK to resume commissioning?
Off-line
2.1. As stated in another posting: megger of course
2.2. Hi-pot testing
2.3. Polarity Index
2.4. Short circuit test

On-Line
2.5. With the unit loaded perform a flux-probe test to detect any possible damage to the coils.
2.5. Monitor vibration

3. How can we draw 10% power from the prime-mover when nothing is connected to the stator terminals, i.e. generator open loop?
3.1. How was that measured? with the control system? may be the zero is off in the MW transducer... should be contrasted with the MWh meter.
3.2. Depending how the system is setup... you may have the auxiliaries driven by the Generator so perhaps what you measured was the in-house loading. In any case if the voltage was 50% higher than nominal any small parasitic current detected by the CT's may have been scaled up to 2.5MW...
3.3. If the start-up engineer was taking high speed data with the control system all the values should be logged... it should fairly easy to correlate the output to the moment the fault occurred.

HTH
Don't believe in miracles, depend on them.


saludos.
a.

 

[jbartos]

Suggestion to kevd (Electrical) Oct 10, 2003 marked ///\\We are commissioning a 25MVA, 4160V brushless 2-pole alternator driven by a gas turbine at 60Hz (3600rpm). Today at with the gas turbine at synchronous speed and alternator off the bars the excitation was powered up for the first time. The AVR/excitor commissioning team set the manual excitation on its lowest setting,
///Has any start up procedure been followed?\\ but when the excitation was switched on the excitation power convertor fully phased forward and the terminal voltage shot up to 1.5 x rated voltage. The excitation was shut off after 5 seconds and the set ran down to stop. It appears, according to the AVR commissioning engineer, that the phase rotation from the alternator PMG into the power convertor was wrong, hence the power bridge phased forward out of control.
///Has any wiring been done or modified in the field?\\ The gas turbine engineers reckoned we drew 2.5MW out of the turbine during the incident.
///It appears to be on higher side, since the efficiency of generator is much higher than [1 - 2.5MW/(.8 x 25MVA)]x100%=87.5%\\My question is this?

1. What damage could we have done to the set (insulation, rotor heating, etc.)?
///Check the tech specs of affected equipment for 1.5 p.u. x rated voltage with upper margin.\\2. What checks should we undertake to ensure the alternator is OK to resume commissioning?
///Check the wiring of AVR, and exciter for correctness. measure the insulation resistances/impedances of 1.5PU x Vrated affected hardware.\\3. How can we draw 10% power from the prime-mover when nothing is connected to the stator terminals, i.e. generator open loop?
///2.5MW or so is a transient. It would probably level off to some smaller MW.\\

 

[dpc]

I doubt any significant damage was done to the generator for such a short time period.

However, I'm not quite buying the story of how it happened. With VR in manual, it should be attempting to provide a fixed current to the exciter, regardless of polarities of PMG or any voltage sensing. There should not have been any automatic control action taken by the regulator. It sounds more like the AVR was in automatic and had a bad PT input or something similar.

By starting in manual and gradual increasing excitation, the exciter characteristics can be determined and the VR limits set to avoid this kind of excursion when the automatic control goes gunnybags.

And I agree wit boomsystems, a generator of this size should have volts/hz protection and this should have tripped you off pretty quickly at 1.5 pu V/Hz.

 

[boomsystems]

A phase sequence change at the power input terminals could cause a problem.But the result will be erratic firing. It is hard to believe that a wrong phase sequence could take the excitation to ceiling conditions. It could simply be that 0% & 100% could be due to reversal in potentiometerconnections if one was employed.

REMEDIES

a) It is common practice to provide a phase sequence indicator /relay to avoid this situation.

b) The overvoltage timing setting of AVR needs to be brought down to 1-2 Sec

2) Even under saturation conditions the excitation power can not exceed the full load excitation power. For a general size of 25 MVA the full load power could be only in the range of about 250 to 400 kW. A 2.5MW power recorded seems tobe high (attributable to wrong readings caused by
CT,PT polarties etc.
3) Nothing would have happened to alternator because of the above conditions.

4) Let the customer disconnect AVR and ensure that by giving separate excitation (by a variac, rectifier & Aux AC 60HZ,1Ph Power supply) to the exciter field & do OC test. The result should match with the factory readings.

If SC could be conducted, nothing like it. Here also the excitation should be as per factory readings

The sum of the No laod excitation squared & SC excitation squared under root should be slightly less than the full load excitation figure marked in the name plate.

Generally generators are designed to withstand such abuse conditions like ceiling excitation on no load, wrong synchronising etc etc.

 

[busbar]

That a modern generator-excitation system would not be restrained from full-tilt output during AC-phase reversal seems like a very poor design.

 

[hsj]

1.1.5 times voltage for 5 sec should not cause any damage to alternator because even during an insulation resistance test 5 kV is applied for a minute. The probable damage could be to AVR because during an insulation resistanc test the AVR is disconnected to avoid over voltage.

2. On no load 2.5 MW power transfer is unlikely. The power consumed will be no load losses and excitation circuit.

3. Recommend to test AVR