Generators as load banks 2

[tomad]

For facilities served by 2 or more emergency generators, these generators could be load-tested in pairs, one as "generator" and the other one of the pair as "motor".
This way the requirement of providing a load bank could be eliminated.
Is there any code stipulation that prohibits the use of this testing method ?

 

[tomad]

pebee

In our case, the driven "generator" is also a synchronous machine, its speed is dependent only on the supply frequency.
In the case mentioned by you, it is probably an induction (asynchronous) machine, which, by being forced to rotate above its synchronous speed, becomes a generator (if it is connected to the power grid).
But, with regard to the fact that that generator cannot accept over 0.2 of its rated load, if it is a regular 3-phase motor, its characteristics I-n (current - rotating speed) should be similar to its motor one, so it should be able to accept up to its rated load, with a slip equal to its full load slip, but in the opposite direction (with minus); thus causing an overspeed.

Hoxton

Yes I was talking about generating sets, but the driving machine is a Diesel engine, the generator is a synchronous generator.


rmw

The question arose as a general idea, there is no specific Diesel engine involved.
But by the way, how do you increase the speed of a Diesel engine, do you admit more fuel ?

 

[davidbeach]

Its not the generator that will have the problem. The generator could function as a motor up to its full nameplate rating and it will rotate in the same direction. No slip for synchronous motors or generators, both rotate at exactly synchronous speed. It is the diesel engine that can't accept more than 20% of rated power as motoring torque on its shaft. Load bank or grid tie are the best way to do what you want.

There is one other option that might work if you can connect more load to the pair of generators than one of them could provide but less than the maximum output of both together. Run the generator to be tested in base load mode at what ever output level you desire and then let the other generator operate in load following mode to make up the difference. Generator system controls will not have been installed with that option so it will take some custom control work but at least it won't ruin your diesel engines.

 

[rmw]

Tomad,

To answer your last question, if your diesel/gen set is tied to the grid, no, adding fuel only adds torque to the generator shaft. The grid controls the speed. If the generator is separated from the grid (or driving another gen set as a load unsynchronized,) then yes, you will spin the other gen set up to the frequency determined by the capability (governor, power limits, etc) of the driving gen set's diesel.

rmw

 

[itsmoked]

But! While the alternator can motor at its rated power there will be no mechanical load to use that power.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[rmw]

I would change the verb in DavidBeach's sentence from "...can't accept more than 20% of rated power..." to "can't dissipate more than 20%..."

The work of compression is being returned to the shaft by expansion during what should have been the power stroke so the only power being dissipated is the purely friction losses of spinning the crank and dragging the auxiliaries.

rmw

 

[waross]

Compression stroke; hot.
Water jacket cools the compressed air charge, removing energy.
There is less energy to return to the shaft on the expansion stroke due to the cooling of the charge.
The amount of energy loss varies greatly depending on several factors.
David has an excellent suggestion. It may not be usable in all circumstances. The loading may be done manually by adjusting throttle settings if you are running in droop.
If you don't have enough load available to load one set to 100% then then bite the bullet and find a load bank.
Grid loading is technically feasible but may not be possible because of utility policy or may be very expensive.
With electric motor driven sets the loading drops to the windage loading which will be much less than with a diesel engine.
Beware of tinkerers. It is possible to play with governor and exciter settings and get circulating currents equal to full load current. Despite what your ammeter is telling you, the engine is not being loaded. A power meter (watt meter) will confirm this. This condition is not a valid load test.
respectfully

 

[catserveng]

David Beach wrote,

"Its not the generator that will have the problem. The generator could function as a motor up to its full nameplate rating and it will rotate in the same direction."

When I first started as a service technician working on "smaller" generator sets between 100 and 3000 ekW, on several occasions we would go to a customer site where an engine had failed and "motored" all night due to some kind of control or component problem. The usual cleanup of cylinders, exhaust, oil mess, etc as discussed followed these events. Rarely we found the surge suppresor on the rotating rectifier assembly blown, and once in a great while a diode failed, but in general the generators did make pretty good motors.

In recent times, as generators have gotten "better engineered", meaning smaller, lighter, higher temp rises and more power/pound, we are seeing failures to gen ends after motoring events, primarily in exciter components and in rotors. When we do report failures in the sizes mentioned above to the manufacturers we use, one of the first questions is always "was the unit motored?".

So one of the current challenges we face today is a rather large population of very experienced engineers, who are not recognizing the changes made by manufacturers and their operational effects. And we are having to re-evaluate many of the protection features we use and the settings applied.

And please Mr Beach, don't take this as any sort of negative feedback, as in you former job I was a participant in many of the seminars and training sessions you were present at, and miss the your expertise in your former company.

 

[slavag]

Dear Catserveng.
Thank you for your infromation ( news from site).
We always used 32 function as mover protection and not as generator protection.
Now, I see it's important also for generator.
OK, I take this in account and also we will request additional information from manuf.
BTW, I think it's big problem for excit system with rotated diodes and not for static excit. system.

Additional Q: in rotary UPS you used motor/generator option.
Why it's not problem?
Regards
Slava

 

[catserveng]

Slava,

If you are talking about systems that use a lineup with a diesel engine, a generator, and an electric motor in a single drive line, then the systems I am famliar with use some type of clutch at each coupling. So when the electric motor is driving the generator the engine is uncoupled and when the engine is driving the motor is uncoupled. I know there may be some other types out there, but this is what I am familiar with.

As a general comment on this subject

Load testing of standby generators is becoming, at least in our area, an increasingly discussed topic. The usual options have been to either install a permanent load bank, from partial to full load rating, or the regularly contract a load bank test with a portable unit. Some facilites are designed where there is opportunity to use facility loads to load test without disrupting operations, but these are pretty rare.

The single biggest factor driving these changes are emissions reduction requirments for the prme movers. To meet the upcoming tier levels for allowable emissions there have been significant changes to the diesel engines. And one of the primary operational "paybacks" for these changes is that newer engines from most all manufacturers don't like running lightly loaded. Add aftertreatment for PM and NOx reduction to this, and yes, even on standby units it will be required, and the ability to run unloaded or lightly loaded for any length of time is dramatically reduced.

So customers are looking for options, and load banks are expensive and hard to finacially justify in a lot of cases. And portable load bank testing usually introduces system outages and interference with operations as when installed no one hardly ever makes accomadations for future testing.

One recent development in our area, and several California utilites are doing similar programs, is Demand Response programs where the utility is providing incentives for customers to make their standby units parallel capable so that in power emergencies the utility can dispatch these standby units to parallel with the grid and export power, or the customer can choose to seperate and remove his load from the system in a closed transition. A great side benefit of these programs is that the required regular testing by the utility allows the customer to load test his unit against the grid.

I personally think these types of programs are of great benefit to both customers and utilities, the downside is that it does reduce some of our service business, but it is hard to watch thousands of BTU's go do nothing when the power produced could have some beneift somewhere. And the installed generators in these programs actually have fewer problems because there are being run loaded and regularly. It also better tests the generator ends and the controls as we have the opportunity to go into a lagging power factor and also supply VARs into the system.

Great topic, thanks for starting it.

Mike LeClair
Service Engineer
Hawthorne Power Systems

 

[slavag]

Dear Mike.
Thanks a lot.
I'm also recommend to our customers (my side is protection
and station control syatem)close some agreement with utilities about this type of test.
Best Regards.
Slava.
PS.
Maybe you have some form of this program? it's help

 

[slavag]

Hi.
Mike, sorry, I mean form of agreement with utilities, not of test procedure.
Regards.
Slava

 

[davidbeach]

catserveng, my guess it that the excitation controls are not responding properly to the motoring events you describe. For a synchronous machine, rotating at synchronous speed (starting issues ignored), the only difference between a generator and a motor is whether the shaft is delivering or receiving torque from whatever is connect. The typical generator with brushless excitation might have a difficult time starting, but once it is going it could certainly operate as a motor.

What is probably happening is that while motoring the terminal voltage is lower than the AVR set point and the excitation system goes wild trying to bring the voltage up. Obviously there are missing protective elements, or a reverse power relay would have kicked things off early on, so an excitation system blowing itself out might not be too surprising.

In general, the multiple levels of safety factors have been worked out of equipment compared to many years ago and there is a lot less room for error before something goes seriously wrong.

All in all, it is best to not allow a generator to be motored unless the whole system was designed to use the generator as a starting motor as is done with some gas turbine system.

 

[catserveng]

David Beach, Thank you for the explanation. What is making this a little hard is that with new digital AVR's and supposedly improved protection from the on-board controls, you would have thought we would see fewer, not more failures related to motoring events. Although I would have to admit we are seeing more failures in general with generators in the last few years than in previous years, at least in our typical size ranges.

Slava, I don't get directly involved with these programs except when the customer or in some cases a third party get's involved and we do the system modifications to allow for the peak demand response to function per the utility. I know all the California utilities are actively doing this, and there is a third party handling a number of the projects we have been associated with called ENERNOC. The Utilities in Hawaii are also starting to look at this option as well. In most cases I believe the utility is contacting likely candidates and then looking at the ablility of the installed equipment to be made parallel and air permit compliant.

 

[EEtheEE]

New York State also allows facilities to participate in EDRPs (energy demand response programs). Many state facilities participate in these programs and receive a kick back dependant upon the amount of power that can be "removed" from the grid. Enernoc and Energy Analytics are among a few third party companies that assist facilities with these programs.

 

[peebee]

I still don't see why it wouldn't be possible use one generator to load the other to 10-20% of nameplate.

Bear in mind there's a difference between what's a good idea, and what is possible.

If I can use engine braking in my car to coast down to a stop sign, then why can't I use one generator to load another, even if it's only to a very small percentage of its nameplate rating?

And if I only need a 10-20% load on my generator during excercise, and it's possible to use one gen to load another to 10-20%, then why wouldn't it be possible to avoid the expense of a load bank?

I personally would not recommend it. But there's lots of things that are possible but not recommended. Some of the posts above seem to be confusing the two. Will an input power of 10-20% nameplate damage a generator? Or would it just be difficult to control?

 

[dpc]

If the voltage regulator worked properly in motoring configuration, the generator should be able to deal with the motoring. The problems will come from the engine. Probably 10% to 20% load for short periods of time might be theoretically possible.

 

[rbulsara]

peebee:

10-20% load is not a issue for the "motoring" unit, but its virtually useless as the "load test" for the unit under test. For a diesel unit you need at least 30% of rated load to avoid "wet stacking" or "slobbering" that is to burn the fuel completely. Therefore, it is not worth the effort. Bad idea is a bad idea, whether or not it is possible.

 

[cirtcele]

I have once discussed this issue. I guess it depends how much you rely on this emergency generator when the time comes. In other words, you may want to test the generator(s) with the real load as a complete package (transfer switch, regulator, etc)

 

[peebee]

Hmm... so maybe it would work if you stuck some Jake Brakes on your gens - or if you had 4 gens and used 3 as a load for the other 4....

Neither would be very advisable....

I just realized that we all got very far off the original question, which was specific to code issues.
I agree with itsmoked that there should be no code issue -- the only possible issue I could imagine is that using a generator as a load bank might be inconsistent with it's UL listing.

PS -- the cheapest load bank I've ever heard of is a drum of salt water with two electrodes, just keep dumping salt in until resistance is low enough (and load is high enough) for your purposes. I've never seen that done, just read about it here on engtips. Motoring the generator sounds safer to me than this somehow.....

 

[waross]

As far as load testing and codes;
If a load test is specified by any codes, it will most probably specify a test loading. To be realistic, such loading should be related to the full capacity of the generator or related to the normally supplied load. A realistic load test may be from 75% to 110% of the set rating, depending on the duty classification and/or the normally connected load.
You are not going to get that by loading a similar gen-set.
Re; the salt water trick.
Instance #1; This was a new set of about 15 KW, 120/240V. The set normally ran lightly loaded and had a severe oil pumping problem from day one. We used a 60 US gallon plastic barrel for 2 or three days and were able to seat the rings. After the loading the set ran happily at light loads with no further problems. (Why not use a smaller set? Motor starting, specifically A/Cs. After a few years the set was changed out for a 28 KW, which ran at a slightly higher load because it was able to start a few more A/Cs. The load as a percentage of set capacity was even lower than the load on the small set.)
Instance #2; This was a new set that did not have any problems but the vendor was threatening to void the warranty on the grounds of light loading. Vendor agreed to honor tyhe warranty if we did a 5 day load test. The set was 280 KW 120/208 V. We used six 60 US gallon barrels. I was on site, up on the roof (for safety concerns) filling the barrels with water, four times a day. We used a lot of fuel and boiled a lot of water. The set still worked well and the warranty was not voided.
This is cheap and quick for a one time test, but if regular load tests are a code requirement the safety concerns and the monthly labor costs would probably rule this method out. And remember the water vapor produced. It doesn't take even a moderately sized set long to put several hundred pounds of water vapor into the air.
Speaking as one with hands on experience with the salt water method, I would may use the salt water for a one time acceptance test but not for monthly load tests.
An exception may be a purpose designed liquid rheostat where the make-up water, water vapor venting, control and safety issues are properly engineered.
respectfully