Operation of Synchronous Generator as Induction Generator.

[RAgrawal]

There is a small about 10 MW, 3000 RPM Generator coupled to a back pressure Turbine in our plant. One day, Excitation of the generator failed along with station DC power failure (supplied by 220 DC battery bank). Due to failure of DC power, no relay could the detect the loss of field / trip the generator. Even operation people could not trip the generator, as there was no DC power for that particular machine.

Since system was strong enough to supply the increased reactive power demand, there was no other disturbance in system e.g. voltage dip etc. After 8 minutes, machine copuld be isolated by tripping upstream breaker. Tripping of machine caused the over speeding of Turbine but this whole incidence caused the tremendous damage to the Generator. Its rotor was flied off towards Turbine by breaking the coupling and shaft by about a meter. Fire also taken place in the plant. Machine was operating at full load even after the loss of excitation.

Can any one comment on this incident and have any idea about the capacity of machine to operate as Induction Generator? Trends of system parameters are available, if any one of you is interested.

 

[electricuwe]

In the incidence you described the damper winding of the generator operated as a squirrel cage, but I'm not able to calcluate up to which extend this is possible.

Regarding the protection used on your system there have been significant design flaws. Circuits for tripping should always be connected fail-safe. To my mind (I'm not a specialist in power generation equipment, but I'm working in an area where similar issues have to be considered) using a no voltage coil (a coil which trips the breaker if no voltage is applied)on the circuit breaker would have been mandatory. Furthermore stopping the turbine should have been initiated by similar means.

 

[busbar]

Stationary battery sets need to be treated with highest respect and attention. "All your eggs are in one basket, so watch that basket!" Fading battery sets can really get expensive, like $30-million in one crucial spot. It's particularly embarrassing when you have to rely on a relay eighteen miles down the line to clear a circuit (and then in the mop-up you discover that you get only black tar out of the transformer oil-sample valve.)

 

[dpc]

Just to respond to electricuwe comments regarding fail-safe circuits.

In the U.S., the overwhelming majority of breaker control circuits are energize-to-trip. They are not fail-safe. In many cases,dual trip coils and redundant relaying is provided. This is why the dc control power circuits are so critical.

Undervoltage release coils are rarely used at utility substations or generating (non-nuclear) facilities, at least in my experience. For boiler protection, it is fairly common to use fail-safe systems, typically two out three voting logic, but this still is often followed by a one out of two energize-to-trip scheme.

Breakers are sometimes provided with capacitive-trip or other stored-energy devices to allow breaker tripping if dc is lost.

I'd be interested in hearing if this is consistent with the experience of others.

 

[vandax]

I would suggest to apply breaker relay, in case the generator breaker woudln't operate, the time delay and relay sensing also used to operate the next level breaker.

Any comment?

 

[electrageek]

Coming from a Field Engineering background I have seen some pretty nasty failures. Yours is right up there with the best. All the Generator Circuit Breakers I have worked on have been energize to close and energize to trip. I believe the reason “energize to trip” schemes are preferred is that a full load trip due to a loss of tripping voltage is also considered poor reliability. This especially true in the Utility business where a nuisance trip could cause a cascading Generation failure or at least a power disturbance. Because the trip circuit is so important, manufactures have implemented many ways to help insure proper operation. Normally the Red and Green indicating lights are connected to the breaker in a very specific way. The Green light (which indicates “OPEN”) is wired to a normally closed contact on the breaker and gets its power from the trip fuses. The Red light (which indicates “CLOSED”) is wired in series with the trip coil in the breaker and gets its power from the trip fuses. This design gives the following protection. If the Green light is not “ON” do not close the breaker it might means that there is no tripping power. With the Breaker closed the Red light being “ON” indicates that there is tripping power, and that there is a complete circuit through the trip coil of the Breaker. This same scheme can also be used for lockout relays, (wire the lockout trip coil in series with a white indicating light). Unfortunately this design does not help insure that the Batteries and charging system are in good condition. The Batteries are so important that a good maintenance program will check the voltage drop, specific gravity, and connections on each battery once a month. A load test of the Battery Bank once every six months is also recommended. This was a very tough lesson but one you will never forget

 

[busbar]

Words well spoken, electrageek. Just make sure that a neat row of red lights may mean the battery set is doing OK, but no one noticed that all those red lights were slowly [but equally] dimming from a malfunctioning battery charger. All that "mop-up" overtime may or may not be welcome.

 

[electricuwe]

To my mind its important to distinguish between fail-safe and redunant circuits.

For Fail-safe operation of the breaker tripping circuit a no voltage coil supplied from the the battery voltage and with NC-contacts of the protecive devices would have been fail safe (providing the protecive devices are fail safe themself). In case of a bad battery the power-station would fail to start operating which is less severe than the failure RAgrawal experienced.

Redunant operation is further step, but its without value if the circuit which has to trip the breaker is not fail safe.

 

[peterb]

Guys, fail safe as defined above is definitely not normally used for breaker tripping in the type of power systems that we are discussing here. Electrageek has defined the purpose of those red lights, but there are some additional considerations - such as, the red lights should have a properly sized series resistor in circuit (so that if the bulb fails short circuited you don't trip the breaker).

The way that concerns about battery supply voltage is usually handled is to have an undervoltage monitor on the battery charger (which fires up an urgent annunciation point in case the battery voltage goes low. The annunciation would be in a control room, where hopefully someone is awake to contact the service personnel. Of course, the alarm level has to be selected high enough to allow time for corrective action and low enough to not be a nuisance. Breaker trip circuit and protection supply monitoring are other options that would give early warning of impending trouble.
Dual trip coils and even dual batteries are used where there is an ultra-high security requirement - the cost does add up, though.

Back to RAgrawal's original concerns, conventional wisdom says that the generator itself should be able to withstand the loss of field condition for a fairly long time - the concern is usually the effect on the rest of the system, as the excitation for the generator will have to be supplied from the external system, with accompanying low voltage conditions and the risk of instability. The excitation supply isn't normaly taken from the station battery - this supply is usually only used for field flashing. Could this have been the root cause of the battery failure?

 

[electricpete]

Wow, an outstanding discusssion. One more brief mention of advantage of traditional energize-to-trip scheme: the coils don't have to be designed to carry the trip current continuously (whereas a deenergize to trip coil does)... results in more force margine available for the same coil (or ability to use a cheaper coil). Also continuously energized coils sometimes bind up due to baking of the insulation varnish which later deposits on the plunger and gums up the works... not very fail safe if it fails to reposition when deenergized.

 

[ksv]

Reverse power flow relay could be used to avoid such incidences

 

[RAgrawal]

freinds
thanks to all for your input but i think every one has mised the important question that for how much time a generator can withstand to oprate as induction generator. perterb has said that it can withstand for fairly long time whats exact time, because as far as i know it depends on saize n cooling media of generator. and time ranges from few seconds to 3-4 minutes, not more than this.

 

[peterb]

RAgrawal-
Tell us some more about the excitation system of this generator. Is it brushless or brush type?
Sorry that I can/t give you a definitive answer re the permissible time for operation without excitation, that will have to come from the manufacturer. There are recorded incidents of large generators operating for extended periods with the excitation off, the concern is more for system conditions.

On re-reading your original post, it sounds to me as though the damage to the machine is more consistent with unstable operation where the generator has fallen out of step with the external system. Out of step swings would place severe stress on the shaft, possibly leading to failure. These swings would be accompanied by very noticeable voltage swings, which your post suggests were not very apparent. With the loss of excitation, unstable operation is a distinct possibility, depending on system conditions and generator loading.

One more thought - the turbine overspeed trip occurred when the generator was isolated by operation of the upstream breakers. As the unit was still supplying real power up to this time, this indicates that the shaft was intact up to then. Is it possible that the shaft damage occurred due to the overspeed condition following the full load rejection? Given the lack of voltage swings, it is a distinct possibility that this was the root cause of the shaft & coupling failure.

 

[electricpete]

peterb - you're last thought makes a good point that the damage occurred only after the generator was electrically. Apparently the overspeed trip did not operate soon enough and centrifigual force of a slightly unbalanced or resonant generator rotor while overspeeding caused the damage. It seems like some logical questions are:
1 - Why didn't the overspeed trip act soon enough to prevent this?
2 - What is different about this trip compared to a normal trip from full load that would prevent the overspeed trip from acting in time to prevent damage?

 

[peterb]

electricpete -
You've started a trend of thought here -
- Does the turbine overspeed trip provide direct fail-safe operation of the stop valve, or does it require the DC supply to operate?
- If there is a trip solenoid that operates the stop valve, and if this solenoid didn't operate because of the lack of DC supply, then the governor valve would try to control the speed. The governor valve is not a stop valve and even when fully closed could possibly pass enough steam to overspeed the turbine.

 

[dpc]

We recently completed installation of a new steam turbine-generator to replace an older one that had been wrecked by overspeed condition.

The limit switches on the main stop valve were home-brewed at this industrial facility. Apparently there was something blocking the stop valve from completely seating, but the limit switch indicated it was closed. The limit switch contact was used to trip the generator breaker.

The control system thought the turbine was shut down, but there was enough steam leakage through the stop valve and "closed" governor valve to run the turbine up to 10,000 rpm or so. At some point, everything came apart, including the generator. I'm glad I wasn't there.

 

[peterb]

dpc, I've never actually seen this happen but that's the classic reason why breaker tripping is interlocked via a reverse power or low forward power relay, for turbine mechanical or electrical non-emergency trips.

 

[dpc]

peterb,

That would have been better, no doubt.

Do you supervise the generator breaker control switch with the reverse power contact as well, or do you trust the operator?

dpc

 

[peterb]

Good point, dpc; I haven't come across this usage, but it could be considered for a high-risk situation. I guess the answer is that we just have to trust those pesky humans once in a while.
Normally, you try to design the control board ergonomics so that the operator has to really think about it before executing the action. One way of doing this is to make the breaker operation a two-step process, where one switch selects the breaker to be operated and the second is the traditional close/trip switch. Different shaped /coloured handles are another approach.

 

[RAgrawal]

perterb

the excitation system of generator is brush type. AVR gets power through excitation transformer and feeds the field of exciter which intern feds to main fiels o generator.

As far as broken shaft is concern, so you are right, generator was supplying real power and that too rated untill explosion occurred.

solenoid trip of turbine could not be possible due to absence of DC power, it tripped mechanically. The setting of over speed trip was about 15 %.