Two emergency generators 3

[Mbrooke]

No clue where to even begin.


Do I start both generators upon loss of utility? Or let gen #1 start and if it fails to do so start gen #2 after 10 seconds? How do I prevent the ATSs from "fighting" one another?

 

[crshears]

Does it work as the gen is starting?

Hey Mbrooke, not where I was going, and I don't believe it does; to my understanding UFRO is activated when heavy / excessive loads are applied to generators running very close to their normal rated frequency. UFRO maintains the specified V/Hz ratio as the applied load bogs down the generator [the "knee frequency MikeL. described]. As the loads accelerate and draw less power, the bogging ends and the generator speed gradually returns to its normal range, returning up the same V/Hz slope it went down when the load was first applied.

If I've grasped it correctly, I believe your first question can be answered in the affirmative.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[FreddyNurk]

crshears, MBrooke, my understanding, at least from doing a significant number of step load tests with Stamford units, the UFRO means that excitation won't kick in under a certain frequency, conversely it'll drop off under load until either the unit recovers, or the threshold frequency is reached, in which case no more excitation.

It also means that you can idle an engine at 800rpm all day, and you won't get any excitation support at all. Same goes with starting a set, excitation won't kick in until somewhere between 45-47Hz (adjust as needed if you're elsewhere...).

We did a heap of overload tests (500kW on a 300kW set) a number of years back with all the protections disabled, at 45Hz (we'd adjusted the setpoint specifically) the excitation disappeared, so did the voltage, and as a result all the load. The UFRO will end up allowing the set to oscillate between no load and overload as the speed recovers and the voltage comes up until someone stops it.

EDMS Australia

 

[Mbrooke]

1) How does UFRO effect short circuits? If it cuts out to quick the breaker may not have enough time to open. Gens are short circuit limited, so breakers will take longer to open.


2) Does the excitation go away when the gen is shutting down? I know of gensets that wrecked potential transformers when shutting down.

 

[FreddyNurk]

UFRO only kicks in when the engine's rotating speed drops. There's generally (at least in anything that has an AVR) enough grunt to get at least 3 PU, usually greater at the instant of the fault. If the fault is close in and purely reactive then the engine won't slow down and the AVR will keep the excitation up, if its not then its slightly less predictable. I haven't seen too many incidents where it took 3-4 seconds to wait for the undervoltage to take a set out as compared to overcurrent on a fault.

If you're shutting down the engine, then its usually dropping speed as the fuel rack has gone back to zero fuel, and as a result, when the machine slows the UFRO still kicks in.

What I've said above is only based on my experience in high and medium speed diesel sets. Much larger sets will have different excitation arrangements, someone like ScottyUK has better coverage in that area.

EDMS Australia

 

[waross]

Short circuits.
There is a circuit that uses the output of a CT dropped across a resistor to provide a field boost to hold the excitation up during a short circuit.
I don't know how that interacts with UFRO.
I had some sets on which that feature was originally provided but had been disabled.
A switch was closed on a shorted transformer about 200 feet from the generating plant.
Nothing tripped but the 600 kW set stalled.

It's strange that a slow set would take out a PT.
Without UFRO (very old sets) an electronic AVR was often the first point of failure.
There was an excitation scheme that did not use an electronic AVR
I haven't seen one of those for over 20 years and the last new one I saw was about 40 years ago.



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

 

[catserveng]

Coming back a little late on this, i have a pretty nice slide from a recent training package I did for generator techs dealing with this issue, was a long hot day today but I will look it up and post it sometime tomorrow.

MikeL.

 

[Mbrooke]

It was when the set was shutting down. The AVR would keep the voltage up, but the frequency would would go down as the generator slowed causing the PT to saturate. This my fear with loads to should the engine stop with the ATS still in Emergency.

 

[Mbrooke]

UFRO only kicks in when the engine's rotating speed drops. There's generally (at least in anything that has an AVR) enough grunt to get at least 3 PU, usually greater at the instant of the fault. If the fault is close in and purely reactive then the engine won't slow down and the AVR will keep the excitation up, if its not then its slightly less predictable. I haven't seen too many incidents where it took 3-4 seconds to wait for the undervoltage to take a set out as compared to overcurrent on a fault.

If you're shutting down the engine, then its usually dropping speed as the fuel rack has gone back to zero fuel, and as a result, when the machine slows the UFRO still kicks in.

What I've said above is only based on my experience in high and medium speed diesel sets. Much larger sets will have different excitation arrangements, someone like ScottyUK has better coverage in that area.

Will a short trip the generator breaker before stalling out? Will a close in fault still trip the breaker? And that we are at it what is the available short circuit current of a typical 400-1,200kw genset?

 

[waross]

The AVR would keep the voltage up

It has been many years since I saw an electronic AVR that didn't have UFRO.
That said, There is an excitation scheme that uses a constant voltage transformer and a CT derived boost to compensate for load changes.
It is so many years since I have installed one of those that I don't remember what it did on shutdown.
As installed, the load was always removed before the set was stopped so under-frequency would be an issue only as a secondary problem.
Your control scheme should always remove the load before shutting down.
In the event of an unexpected stop, such a fuel starvation, the AVR should protect you.

Will a short trip the generator breaker before stalling out? Will a close in fault still trip the breaker?

It depends.

And that we are at it what is the available short circuit current of a typical 400-1,200kw genset?

Compared to a transformer, not much after the initial transients.
If the breaker does not trip during the transients, it may not trip at all.
It depends.
Google transient reactance and sub-transient reactance.
Close in faults:
Work done is I²R
On a close in fault, there may not be enough "R" in the circuit to overload the set mechanically and stall it.
On faults further out, the lines will introduce enough resistance that the I²R is great enough to stall the set.

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

 

[FreddyNurk]

Thinking about this a bit more, based on the decrement curves I've seen for small sets, the UFRO and AVR isn't really relevant, at least for the first few cycles.
If we look at something like the UMC274 from Stamford, at the point of the short circuit, the fault is about 2kA (from memory) as opposed to FLA somewhere around 180A. After 2-3 cycles the fault current drops through the floor, and if a PMG AVR is used, it starts to pick up again, and ends up at a maximum of about 600A.

As per waross' notes on work done in I²R, for a far fault, the engine will slow, but not appreciably within the first few cycles. If the protection can detect it within that duration it'll trip, well before UFRO kicks in. If not, it'll very quickly drop below the short circuit threshold, and then it might go out on overcurrent, or the machine might stall.

Bigger machines will have a longer duration for that first asymmetrical current that occurs, the behaviour as I understand it is still the same.

Off the top of my head, a 1MVA LV machine will probably have about 25kA subtransient, and may not have any more than about 6kA once the excitation kicks back in, although I might be off as its been a while since I looked at that size of machine.

EDMS Australia

 

[Mbrooke]

Can you explain transient reactance and sub-transient reactance in your own words. Also how close in are we talking?

 

[FreddyNurk]

Transient and subtransient are different reactance figures as an approximation in order to indicate the likely fault current for certain time periods after an initiation of a fault.

Using basic figures (e.g. x''d by itself) will give an approximate value of the expected fault current at the initiation of a fault. For smaller systems, that's generally enough to size equipment and be confident that the equipment is rated correctly. For much bigger systems then more detail and additional data is needed to predict the system under normal and abnormal conditions.

I don't really have a feel on how close in and when it makes a difference, I didn't have much coverage of the feeder issues on the sites that I was involved in as that was done by a different team. They also fed HV networks (MV if you're not down under...) which complicates any estimate on distance. It wasn't uncommon to trip the (single) generator feeding the network for a feeder fault and the whole station would go black.

I would expect though, that within 10-20m at the same voltage level the fault would be mostly reactive and wouldn't slow the machine down, beyond that I'd be guessing.

EDMS Australia

 

[Mbrooke]

I see the NEC mandates that they be listed on the genset.


But, what do I do with the numbers? What equations do I use?

 

[waross]

Another issue that is often overlooked with standby generators is maintenance of the starting batteries.
The owners of the residential standby sets that I installed never gave a thought to the starting battery.
Eventually the battery would fail to start the set and I would be called to replace the battery.
Then the sets started coming out with trickle chargers.
The chargers would keep a battery in service for maybe a year longer before it failed.
The problem was that then the failure was almost always a battery explosion.
I quit connecting the chargers.
Then when a battery failed I just had to replace the battery.
I didn't have to clean up the acid everywhere.
Something to be aware of when you prepare the maintenance schedule.

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

 

[Mbrooke]

Doesn't NFPA 110 mandate a battery charger? Could be wrong on this one.

 

[FreddyNurk]

MBrooke, the base equation to use is the same as for evaluation of fault current at a transformer, i.e.
Ik = (Base Power / (Base Voltage * √3 * Impedance in PU)). There are means to evaluate the duration of each stage based on datasheet values, although I've never had to look into it to any great extent as the decrement curve provides an approximation, and the machines I work with are small enough that the subtransient values decay within 3-4 cycles.

Where the time to first current zero is longer than say, 10 cycles, it becomes more important to consider the duration of the fault and withstand ratings of equipment. ScottyUK has previously posted here about the ratings of larger generator breakers and their application as an example.

EDMS Australia

 

[itsmoked]

!! Really Bill! Exploding batteries.. We just had a guy in here with exploded batteries at all his generator sites.

By "disconnecting the chargers" did this leave charging to an engine driven generator?

I suspect the battery explosions were due to a lack of water level maintenance and not charging per-say.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[catserveng]

On battery chargers required by NFPA 110, Table 5.6.4.2 Starting Equipment Requirements, for both Level 1 and Level 2 systems a float type charger with a DC volt and DC ammeter is required. The charger shall also have alarm contacts. More details on chargers in section 5.6.4.6

Battery explosions have been an issue I have dealt with for years, attached is a good article you may find helpful. (I had more but seems you can only add one attachment at a time?)

In standby systems, most battery failures that result on "explosions" are,

Low water level in battery, level well below tops of plates. Most common occurance is just poor maintenance, have to maintain proper level in the battery.
Also in many cases "wrong" battery used for application, a "maintenance free" battery really isn't, just doesn't need to be maintained that often, in my opinion a poor choice for standby generator applications. Also see once in a while where someone has installed a deep cycle battery for cranking, unless the battery is rated for cranking service, it will likely fail when called to crank the engine.
Could also be battery and charger not compatible, find a lot of lead acid batteries on a charger rated for NiCad's, result is that it overcharges or causes the battery to overheat, resulting in low water levels.
Batteries installed right under generator cooling air discharge or right next to engine block, under load the batteries get nice and hot, water level drops, no one thinks to check levels more often.
Battery chargers not properly setup/adjusted, amazing to me how many times a year I still see this, not all that hard, make sure right battery type selected, and float and equalize settings match what the battery manufacturer calls for.

Loose or corroded connections, during crank a bad connection may spark, in my experience a leading cause of battery explosions, usual culprit is a cheap clamp on style terminal adaptor that as fine strands of the battery cable relax, the joint loosens.

Internal battery connection faults, usually the result of a cheap battery not designed to sit next to a running engine exposed to heat and vibration, if you want to use a cheaper battery, move it away from the unit where it doesn't shake, rattle and bake. Sometimes failures are just due to a manufacturing fault internal to the battery.

A frozen battery (not real common where I live but have seen it when working in cold climates) that someone tries to fast charge or jump start, usually results in the top blowing off.

A battery beyond its service life, corroded plates, sulfation, metal fatigue can all lead to internal arching that can lead to an explosion. Many genset manufacturers recommend changing standby unit batteries at least every two years for lead acid batteries. Actual life depends on a lot of things, heat and vibration kills batteries, units with remote mounted batteries and properly sized cable tend to give better service life. Well maintained batteries live longer than batteries that get ignored. Batteries live longer with the proper charger setup correctly compared to ones using a cheap charger not properly adjusted.

Hope that helps, MikeL.

 

[itsmoked]

Thanks Mike. Interesting stuff.

Maintenance Free are a disaster in that kind of service since a charger misadjustment is magnified over time.

.
.
.
.
.
.
.
.

a charger - all done
wrong! check water regularly
a human failure

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Hi Keith.
These were package sets. The battery charger and battery were installed on the set by the builder.
No maintenance. Zilch. Nada.
On the other hand, the sets that did get maintenance, I would rather that they didn't.
Typically the power was out every Sunday and every Monday, the conscientious yard boy would top up the radiator, to the brim.
Next Sunday the set would run and the coolant would expand and some would be lost out the overflow.
After about a year the coolant would be 100% water. No anti corrosion properties left.
On one site the handy-man knew that the set should have anti freeze and topped up with 100% anti freeze every week.
The floor was green and slippery with anti freeze and the rad had 100% anti freeze.
Can you imagine if the battery was filled to the brim every week and lost electrolyte by heat expansion?
The engine driven alternator was more than adequate to keep the battery charged.

I believe that the failures were a combination of two of the reasons Mike stated.

Low water level in battery, level well below tops of plates.
....
A battery beyond its service life, corroded plates, sulfation, metal fatigue can all lead to internal arching that can lead to an explosion.

With no charger other than the set driven alternator, eventually the battery fails and the set doesn't start.
With a charger, eventually the battery fails and explodes and the set doesn't start.
While cleaning up the acid and debris I would often say a prayer of thanks that no-one was in the room, or worse, starting the set manually when the bomb went off.
Also thanks that it wasn't me trying to start the set.
If the engine driven alternator won't keep the battery in shape to start the set, it's time for a new battery anyway.
A charger is not needed with a battery in good condition and is not needed if a battery is properly maintained and changed out regularly.
If the battery is not maintained a charger will make the difference between a failed battery and an exploded, failed battery.
A charger will allow an unmaintained battery to remain in service for too long after it should have been replaced due to poor condition.
Does the NEC dictate the charging rate?
If you have to have a charger, consider setting it to as low a charge level as possible. Like zero!

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