Generator Protection systems 1

[ruggedscot]

Have one to ask....

Setting up for a generator panel I take it requires some specialist theory, Im looking at a situation that happened and want to try and work out if this would have been an expected result or if there was an issue with protection grading.

Scenario multiple generators set up to a paralleling board. This linked as a neutral seperated system. ie the first generator to start and connect to bus the neutral breaker closes. All other sets the neutral breaker remains open - reduces circulating currents.

The generator board feeds out to other boards that have utility / generator automatic selection fitted. No utility availiable then the source changes to generator. A signal is sent to start the genertators. If already running supply will connect when the generator breaker closes at panel. Anyway fault happens at remote board. This causes shut down on the generators as an overcurrent fault.

I think that this indicates a problem with the set up - should I have not seen the breaker on the generator board feeding the faulted board open and the generators keep running ? Lost all supplied boards as the breaker did not clear the fault allowing the fault to affect multiple users.

Generic thinking of this leads me to think that these must be some differentiation on setting up of ACB's as generators can not supply what would be typically supplied from a utility source under a faulted condition. Is there a specific routine that is followed up for setting acbs up that are supplied from generator and not utility sources?

All advice and thoughts would be greatly appreciated.

 

[waross]

This sounds like a co-ordination issue. This can be particularly challenging when you have a breaker (at the remote panel) fed at times from the utility, and at other times by a varying number of paralleled generators.
I would be trying to co-ordinate on a time delay basis as well as a current basis. This will need a lot of calculations to assure that the amp-seconds resulting from a fault at any location are not excessive under any condition of utility power or any number of generators online.
You may be able to use two protection relays to trip the problem breaker. One relay would be configured to co-ordinate with the utility and the second would be configured to co-ordinate with the generator protection. Only one relay would be active at a time.
You may consider air core reactors to limit some fault currents.
Without being onsite, we can't help too much with specifics, just offer suggestions.
respectfully

 

[ruggedscot]

Thanks

I dont think I worded it too clearly - what Im trying to work out is the generator panel board breakers. I think these are set up wrong. Forgetting the supplied loads for a bit, If there was say a cable fault on the outgoing cable from the generator board the ACB feeding that cable should have tripped and not taken down the generators. To have the generators go out on over current would mean that there was some serious descrimination issues with the breakers.

Looking at the inst settings on the ACB's these Id imagine should be lower than what they would be if the ACB was utility fed ? Am I right in saying that the faulted current from a generator is much much less than what is supplied from the utility company network even a bank of them the expected current would be much less.

 

[davidbeach]

Full selective coordination may not be possible using low voltage breakers with instantaneous trip units; multiple trip units may simultaneously trip, not just the closest to the fault. There are a variety of ways of providing coordination, but straight time and current alone probably won't do it. The disparity in fault currents between utility and generator sources further complicates the situation.

That isn't to say it can't be solved, but it will take more than the trip units on the breakers, possibly some interlocks between relays. Perhaps differential relays can isolate things close to the generator and block tripping for out of zone faults until sufficient time has elapsed for devices further downstream to trip. It's hard to say without seeing the details of the system.

 

[ScottyUK]

Is the generator overcurrent relay a plain O/C relay or something more sophisticated like a 51V voltage restrained overcurrent? The latter modifies the overcurrent setting as a function of terminal voltage so as terminal voltage collapses under a heavy fault the O/C setting is reduced.

Thinking aloud having the first coffee of the day (so apologies if my brain slips into neutral occasionally):

Do you have the option to either separate the loads into a larger number of smaller circuits, or to run a pilot wire for a diff scheme to protect the large circuits?

Some ACBs will have the facility for a residual current relay which you can set to (say) 50A which will provide instantaneous trip in the event of a fault on that circuit. This might be troublesome if you have multiple system neutral earth connections.

The generator feeder ACB may require the STD function enabling to maintain coordination, but be aware of the significantly increased arc energy let-through and the withstand rating of the generator and downstream cable.

You will need a fair bit of data from the generator vendor to model how the combination of the generator and AVR behave under fault conditions. The OEM should be able to provide this. Once you have the data it should be a fairly simple task for a consultant to model the scheme for a grading study.


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

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

 

[ruggedscot]

This information is greatly appreciated - It is something that we are now looking into very closely as we cant allow us to go through something like this again. The impact of the fault was pretty large and it brought down a hell of a lot of equipment. So it is something that is going to have to be looked into. I knew that generators didnt supply anything near to utility power to a fault so this would have an impact on the time grading and selectivity of the system. Looking at a way round this and seeing if we can find out how to manage this and ensure that we dont have to face a similar position again.

We had a utility outage and our generators went to bus the generator switchboard fed out to individual remote package substations and we had a breaker failure within one of those substations. It caused the generators to trip out and we lost all the supplied loads.

 

[PHILEC]

I agree with them that this is a grading problem on your system, you need to review again all documentations on relay coordination study. Check also the short circuit study, I mean the basis on how they set the protective relay because in your system you can have different scenario like all generators are feeding the system or only the utility is supplying power to the loads or only one or two generators are running or generators are paralled to utility. All these scenario will result in different fault current level.

Just want to add and clarifies the aplication of 51V relay. In practice, we used 51V relay on generator overcurrent protection as a backup protection for fault current which is not cleared by downstream protective equipments. It detects the sustained short circuit which is lower than the full load current of generator by 40%. The simple 51 relay cannot be used to protect generator because it cannot be set to operate on sustained fault current. Hence a voltage restraint overcurrent relay (51V) is used.

 

[ruggedscot]

Hi

Mode of operation never includes generators running in parallel with the mains.

Generators remote to package substations - when utility drops at any package substation signal sent to start generators, generators then start and connect to Generator board. Generator board has mulitple outputs and one of the outputs feeds the package substation that has lost power. Two ACB's set up as an interlocked source selector either the transformer or the generator infeed.

The actual incident that we had resulted in all connected gensets going out on overcurrent so the genset protection worked as I would have expected. But as the fault itself was quite a distance from the generator board and on the downstream side of an outgoing ACB fitted to the generator board Id have expected that ACB should have opened and removed the fault instead of the generators themselves seeing the fault and tripping on overcurrent.

 

[PHILEC]

Ok, you need to check the downstream protective device and equipment near the fault point better inspect for misoperation or check the generator protective device settings. The problem maybe lies on malfunction on protective devices near the fault point or the grading/ coordination between generator and downstream protective devices.

 

[ScottyUK]

What was the nature of the fault? Earth fault or inter-phase, high impedance arcing or bolted short, etc? This may give you a clue where to look for the problem with the grading.


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

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

 

[ruggedscot]

Type of fault - probably a mix off all !

Best described as a complete breaker failure.....

phase to phase fault with possible phase to earth and mesh with utility power....

 

[PHILEC]

I agree with you mate youve got to withdraw that breaker and have to inspect it, better if you can invite pecialist from the manufacturer of switchgear.

at least the upstream didnt fail!

 

[DanDel]

ruggedscot, typically, a system like this is designed so that an overcurrent or fault tripped C/B will not reclose without a reset of a lockout or bell aram switch, so that the energencyb generators do not feed into the same fault.

That being said, it is assumed that the coordination must be correct so that the device closest to the fault (downstream from the Utility/Generator transfer) clears first, because tripped C/B upstream will allow the generator(s) to feed the fault.

PHILEC, I think you meant to use "full (or instantaneous) fault current" instead of "full load current " in your statement: "sustained short circuit which is lower than the full load current of generator by 40%"

 

[ScottyUK]

I think PHILEC is describing the setting of a voltage controlled overcurrent relay rather than a voltage restrained overcurrent relay. The two are similar but distinct: the voltage restrained O/C relay typically has a pickup setting greater than rated current, while the voltage-controlled O/C is set below rated current. In that case his statement is valid.



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

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

 

[ruggedscot]

This is great information guys....

The situation resulted in a loss of two breakers - the genny breaker and the incommer breaker. These were replaced with new devices as they could not be reused.

My concerns were with the generator switchboard. I had thought that the circuit breaker on the generator switchboard would have tripped disconnecting the fault and the generators would have not shut down. This is the part that I was trying to work through and seek some input - There must be a co-ordination issue with the generator protection and the circuit breakers fitted to this panle board. We should not have lost the generators each set shut down on overcurrent when we had a problem that was downstream of the circuitbreaker on the generator panel board.

Rugged

 

[davidbeach]

You may find that for your generator breakers you will be better off with the trip unit turned off and the breaker tripped by a generator relay with differential protection. make the differential the primary protection along with negative sequence overcurrent and the voltage functions. Leave overcurrent, both phase and ground, as backup protection and set it deliberately slower than anything going out of the generator distribution board. It could also be helpful to put bus differential on the generator distribution board. With differential on the generators and the generator distribution, you can instantaneously trip for faults in those zones and let the feeder breakers take care of everything beyond the board.

 

[waross]

Am I correct in surmising that both breakers may have closed simultaneously, or one closed before the other had interrupted the arc?
If there was a beaker protecting the feeder to the breaker that failed, it may be the breaker that the co-ordination efforts should be directed to.
If the breaker that failed was the first breaker downsteam of the gennerator bus, then the only protection left would be the individual generator breakers. I think there may be two issues here.
1> The reason for the failure.
Was the fault current capability of the failed breakers exceeded?
Was there an overlap of the breakers.
Did one breaker "hang up" mechanically and not clear before the other one closed?
All of the above?
The complete destruction of both a genny breaker and a mains breaker suggests a combination of the above.
2> If there is a second breaker ahead of and supplying the feeder to the failed breaker, then there is a co-ordination problem between this breaker and the generator breakers.
Davidbeach's last post may be the answer.
respectfully

 

[rcw retired EE]

ruggedscot- Philec & others brought up what I call the generator decrement curve that shows how the generator's output during short circuit drops from its initial higher value (2-5 X generator full load - still much less than the utility) to a value about 60% of normal full load current. A 51VC or 51VR relay is designed to sense this condition for backup overcurrent protection.

Most larger generators have this protection or have the ACB's set to trip somewhere along this curve. (Their trip characteristic intersects the generator decrement curve on the time-current coordination graph.)

But a typical molded case circuit breaker or a low-voltage air circuit breaker's trip unit will never trip because the fault current drops below normal full load current before the trip unit picks up.

Is it possible that each of the generators have this protection but the breakers in the generator distribution panel that carry the sum of the generator currents do not?

Each generator's output would drop according to its decrement curve. The total current through your generator distribution board would also drop. Maybe it dropped off the overcurrent element's trip curve, leaving it up to the backup protection to trip each generator.

The problem is magnified as each generator trips off, since the total fault current is reduced, extending the breakers' trip times.

It is very difficult to coordinate the breakers with standard trip units, especially if the number of generators on line varies, causing a variable fault level.

Try zone interlocking the breakers?

 

[ruggedscot]

Hi

The set up is as follows -

We have multiple generators that connect to a switchboard. This switchboard has incoming and outgoing ACB's. The generators connecting through the incomming ACB's and the packaged substations connecting through the outgoing ACB's so there is one bus in this switchboard.

The package substations that are fed from this Generator switchboard contain two ACB's a generator incommer and a transformer incommer. These are interlocked and switch automatically. Utility fails the generator breaker is pulled in.

In this package substation there was a failure on one of the breakers. We reckon it was a failure on either the breaker or the bus connections behind the breaker. This resulted in a fault developing. The end result being that the generators tripped out and we lost all the packaged substations.

What I was aiming at was that we had an ACB on each feed into the generator board from the gen sets. On each out going feeder there was an ACB installed. The cable then ran for quite a distance and then entered the package substation. It was within the package substation that the fault occurred. So you can see now where Im coming from. The board feeder ACB installed in the generator board should have seen the fault and dropped but it didnt. The end result was that the generators shut down on over current and power to the generator board was lost. This had an effect of killing all the other pakcaged substations as they were fed from the generator board. Utility supply was dead.

I had an inclination that under a fault condition the generator current drops back a fair bit. And it is no where near as large as what the utility can supply. So I started to look at the ACB settings and think that this had some how impacted on the system. The ACB did not pick up on the fault but the generator picked up and shut down. Trying to work through that and start to appreciate what has happened and what leassons can be learned from such a situation. Maybe put in some different protection and look at ways of ensuring that the supply continuity is maintained under such a scenario. If we have the system configured as is we should never have lost the whole installation from what was a breaker fault. If some how the interlocking had failed and the utility company had restored the power then maybe we had a fault that comprised of an arcing fault and also partial bad syncronisation to the network. Still this gives me concern as I dont see why the generator panel ACB did not see and clear the fault before the generators picked up the fault and shut down.

Going back through and thinking on each and every eventuality on this one and trying to see why this happened and how we could work round it and prevent it from happening again.

The zone protection seems to be a good idea but it is trying to get this linked to the generators - we can have all the best switchgear out there but if there is an issue with the generator protection then no matter how good we make it the generator is going to see the fault and shut down before any downstream protection can operate. Trying to work out just how a zonal style would improve things here - The Protection would have to be graded in such a manner as it could react and trip the breaker but yet remain imune to start up currents and any other transients that exist through the loaded network.

 

[waross]

Something to consider;
There is a fair possibility that the breaker failure fed utility power back to the generators.
When reviewing the co-ordination curves of the generator breakers and the feeder breakers, you may want to extend the curves to a level approaching the fault current of the utility source. The characteristics of the breakers may be such that co-ordination is good for faults fed from the generators, but not suitable for utility level faults.
You may wish to consider the possibility of utility level fault currents backfeeding the generators when setting up the new co-ordination scheme.
Yes, I know that it is not normal to plan co-ordination for a generator fed system based partly on the current available from the seperate utility source, but given the difficulty of destroying a breaker with the fault current available from a generator, and the fact that both the utility breaker and the generator breaker were destroyed in the same event, it appears that in this case utility power may have backfed the generators.
I think that this should certainly be a subject for discussion with your team.
respectfully