Excitation Loss protection reliably provided by other means in sub-MW generator? 2

[electrickiwi]

For a 600 kVA, low-voltage grid-connected brushless synchronous generator, can loss-of-excitation be picked up reliably enough via a combination of under/over voltage and Q import limits?

My understanding is that a sudden loss of excitation (AVR fault, open-circuit field winding, etc) would result in the grid supplying reactive power to magnetise the rotor and the machine running as an induction generator, potentially open to pole-slip, stator overheating, etc.

However there must be a size of generator for which dedicated ANSI 40 protection isn't economical... Would a kVAr import limit (set just inside the generator capability curve) and suitable u/o voltage protection be sufficient cover for excitation loss?

Thanks

 

[waross]

Remember that an induction generator must run above synchronous speed. Exported power may depend on the governor settings and configuration. The exported power running as an induction generator may be much less than expected. Two factors, the reduced input power due to governor action and the slip frequency may mitigate against pole slip issues as seen in a synchronous machine.

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

 

[dpc]

You're probably OK with over/under voltage and power factor limit (plus overcurrent) at this size. Although, once you get to that point, the relay is likely to have LOF included. The new generator protection relays pack a lot of functions into one device. Back in the era of electromechanical relays, it would have been rare to see a 40 device on a generator this small. Now it is pretty common.

 

[electrickiwi]

Thanks both,

Unfortunately the multi-function relay in question doesn't include LOE protection and it'd be hard to justify the economics of adding it on a machine this size, especially given the likelihood, as far as I can tell, of an event tripping one of the other protections (excessive kVAr, u/o voltage, overcurrent, loss of comms to AVR) before damage results.

 

[waross]

I have never seen loss of field protection on a machine under 1500 KVA. I have seen quite a few generators.
Catserveng??

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

 

[dpc]

The only time I've seen it for a machine this small is when the function was available in the multi-function relay anyway. Also, newer digital voltage regulator may have some type of loss-of-field protection capability as well.

 

[catserveng]

When we had a pretty good cogen business, 300 to 1000 kW units fueled with natural gas and tied to the grid, we found in many cases a loss of field relay to be a benefit. We used relays from both Crompton and Wilmar, they were just simple reverse VAR relays but worked quite well for our needs. On larger units we mainly used Basler BE1-40Q relays. Starting in the later 90's we moved pretty much across the board to multifunction relays, for smaller units we used relays like the Basler GPR and for larger machines, usually 800 kW and up or anything medium voltage we used either GE MultiLin SR489's or SEL 300G's.

Early in the game we found loss of field to not really be much of a problem, however as tail ends changed (got lighter, smaller, cheaper) and AVR's went from pretty robust analog designs to smaller and less expensive designs, we saw a pretty noticeable increase in failures. We also went thru similar exercises in cost reduction with two of our primary tailend suppliers in regards to rotating rectifier assemblies. A number of failures caused us to more closely look at loss of field protections and how to better protect our machines.

A couple of things I found over the years, this is based on salient pole generators, from 300 to 4500 kW driven by recip engine prime movers, mostly 480 VAC, some 600 VAC machines, a fairly large number of 4160 VAC and some 12kV units.

On smaller simple baseload systems parallel to grid, we used a DC voltage relay set to about 10% less than the no load excitation level of the machine and a 10 second time delay, enabled by breaker close. If the field volts dropped more than 10% of the no load level, we tripped the unit off line. It was inexpensive and never personally had a case where it tripped off without good cause. Used this mostly in system with 32 VDC nominal field voltages, smaller units on up to about 450 kW.

The smaller relays by Crompton and Wilmar we used on units from about 400 to 800 kW on 480 VAC systems, they were small, cheap and pretty reliable, so talking the customer into them wasn't that hard. We also used them quite a bit in diesel power modules up to 2MW in rental service for grid support and large paralleled systems.

On larger units or more complex systems we mostly used the two relays mentioned above. Over time, as we got more operational time and dealt with differing failures, we found that while the Mho offset was the primary/preferred loss of field detection mechanism in relay, that "partial field loss" situations, mainly due to rotating rectifier problems or a rotor problem didn't cause a trip, but ended up causing problems or damage. In a total loss of field, like a PM failure, power fuse blowing or an AVR failure, the Mho offset worked as expected, but in many cases it didn't, and so we moved to always include a reverse VAR protection in all the systems we had a say in.

For a unit the size you talking about, frankly if I was putting it together it would probably have a multifunction generator controller like a Woodward EasyGen or ComAp, and both (and most others) have a pretty nice generator protective package built in. Or if you prefer a smaller multifunction relay, like a Basler GPS100 or something similar is also pretty popular in that size range and includes loss of field in it's standard offering.

Lastly, the DPC comment above, when digital relays came out, we had a number of plants install then on mostly larger 4160 volt systems. Because the digital relay had an alarm feature it was decided to save cost we could eliminate the 40 device, thinking being that if the relay sensed anything wrong it would alert the control system and we could trip the breaker and shut down. Initially it worked fine, however as the AVR's and systems got older, a complete failure of the AVR or it's power supply became more common, a dead AVR did NOT advise the system of a problem, and so we had larger more expensive machines fail, in a few cases requiring major removal/repair. While newer digital regulators with auxiliary power supplies reduce those kinds of possible problems, I'm not sure based on what I've seen over the last several years that I want to rely on it, my personal opinion.

Hope that helps, Mike L.

 

[dpc]

Mike,

Thanks for the info and your experiences.

Dave

 

[waross]

Thanks Mike.

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