Suggestions on Improving Efficiency (Motor-Generator ) 1

[jimmynora]

(Drawing of the setup attached)

Background:

We conduct electrical tests in our lab on electrical components (circuit breakers, switchgears,etc). In the lab we have a short-circuit Generator that generates the electricity for the tests. The generator is powered by an asynchronous motor with a slipring rotor. We are trying to improving the efficiency of this process by measuring our electrical consumption and comparing it with how much we actually need. A suggestion is to feed-in the excess electricity back into the grid.

Specifications of Motor/Generator:

6KV
2500KW
280A
Pf= 0.9
f= 50Hz (R= 2975/min)

I need suggestion on how we could improve the electrical efficiency of the testing procedure.

Q) Any other feasible methods on how to save the excess electricity apart from feeding-in the excess electricity into the grid ?

Q) How would we need to modify the circuit (best method) and what additional components would we need to install to allow us to feed-in the excess electricity ?

Q) During shut-down of the generator or braking of the generator to slow it down (eg. 1000/min) can we use this mechanical energy to provide electricity to feed-back into grid or circuit. How can we do that and what components will we need to install (In this case the frequency will drop below 50Hz required frequency of the grid.)

 

[LionelHutz]

You are chasing your tail. Turn this unit off when it's not being used if you want to save energy. Replace the motor with a more efficient one if you want to save energy. Investigate removing or reducing the motor fan if you want to save energy. If you make this motor-generator set easier to keep spinning you will save energy.

You still seem "stuck" on this idea that the generator is always producing its rated output power while running, ie 2500kW, meaning you need to find some means to "save" this energy. Forget about it because it's not.

Q1 - Suggesting powering a motor coupled to a genset connected back to the source power is a way to save power is supporting an over unity scheme.

Q3 - it's not worth the expense considering the small amount of energy you will recover.

What really surprizes me is that a company which is making/testing circuit breakers and contactors is asking or considering such things.

 

[Skogsgurra]

This setup does not consume much power above what the device under test consumes.

Is it the difference between rated power and actual power that is your concern? Not much to be done about that.

When decelerating, the rotation energy in the motor and test specimen will automatically be fed back to the grid. That is how a genset works. Genuine four-quadrant drive.

Has some bean counter told to improve the efficiency of this test rig? If so, you can safely tell him that efficiency is good and not much that can be done to improve it further.

If the tests are running continuously and if the existing test rig is very big for the tested device, it may be a good idea to buy a test rig more suited for low-power applications.


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

You bet me to it, Lionel. I think this guy is a victim of some romantic ideas.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jimmynora]

The motor/generator runs even at time when no electrical tests are taking place or in between tests. The actual tests take only 5-10mins individually. That's why saving energy is a big challenge for us.

Starting up the generator takes time and energy that's why turning it off is not always the automatic reaction. Part of this project is to find out when is it more feasible to turn off the generator and in which cases is it better to slow it down... or feed-back the electricity.

The issue of replacing the motor is possible but very unlikely. At first we want to see what we can do with what we have. I am monitoring how much electricity the motor draws from the 6kV supply and it is huge (MWh range).

What about components like Regenerative variable-speed drives for feedback ?

If we remove the fan the motor will over-heat and burn-out during the braking process.

How can the generator be automatically feeding back electricity into the grid while decelerating. None of my current measurements show anything coming into the grid. Plus we have no frequency converting equipment to accommodate the electricity at a different frequency than the standard 50Hz.

 

[Skogsgurra]

Every motor is a generator when driven. That is how asynchronous generators in wind turbines work. If you cannot see any energy fed back when decelerating, it is probably because your losses are bigger than the probably small amount of energy available during deceleration.

If stopping the motor isn't an option, then you either have to accept the losses or find another piece of equipment for your tests.

What tests are you performing? Could it be done using a simple transformer? I understand that you do not need a variable frequency - just a variable voltage or current. In that case, a Variac may suffice? Or a thyristor controlled transformer? Or soomething else?

Please give us some details. Voltage, current, duration of test etcetera.

Your location can also be of interest - there are different possibilities in different countries.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jimmynora]

Im working in the testing department of a huge company in Germany.

There are many test being conducting on a whole range of electrical devices to test them to see if they adhere to IEC/ANSI/IEEE specifications.

Tests like breaking and making current on circuit breakers (capacitive and inductive currents), Lightning tests, short-circuit current tests,etc...

I know every motor can act as a generator but so far I do not see any energy being fed back in the grid.

The generator is a two pole short-circuit generator.

 

[electricpete]

Our of curiosity, can you give us nameplate info for the motor and generator? Is there a flywheel? How often do you run it and how long? You mentioned MWh range.... over what period of time (1MWhr over one hour or some longer period?).

=====================================
(2B)+(2B)' ?

 

[rasevskii]

for jimmynora:

For what possible reason do you want to recover energy? At your "Huge Company in Germany" I would imagine that this test room and MG set has been in operation for many years if not decades.

These types of MG sets, usually with a flywheel and/or massive construction of the generator were a standard item for SC labs.
It is possible in fact to brake the induction motor to a stop by injecting Low Voltage DC into the stator winding and dissipating the energy in the already existing resistor setup in the rotor circuit, but the energy is wasted as heat. This method was/is used to brake large synchronous condensers to a stop by using the starting motor (aka pony motor, a wound rotor machine)in this manner. Stopping time reduced from more than 2 to 3 hours to around 10 minutes. Reduces possibility of bearing damage in case the HP oil pumps fail during a stop.

Extra switchgear and rectifier required. Complicated.

Why now is this needed, if in fact it has been running for years?

I am sure that the "above company" could in fact engineer some kind of VFD system that would do a smooth start up and energy-recovery stop of such a machine fully electronically. In fact they have done this all before for hydro units in pumped storage plants. As well as thir competitors...

rasevskii

 

[Skogsgurra]

That huge company ought to have expertise on the kind of equipment you are working with available somewhere in-house.

Wouldn't it be smarter to ask them for help?

I thought you were working in an area with limited electric supply. But I don't think that is the case in Germany.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jraef]

Starting up the generator takes time and energy that's why turning it off is not always the automatic reaction

A common misconception is that starting takes significantly "more" energy than leaving it running for long periods of time. Although the starting current looks high, it is at very low power factor at first, so the actual kW consumed is not as high as people think. Once running and unloaded, the PF is again very low, same issue. So as was mentioned, you are not consuming that much energy when the motor is running idle, all it is consuming is the kW necessary to overcome it's own inertia and not slow down. You'd probably be surprised how little that is.

In addition, your prime mover is a wound rotor induction motor I take it. if so it will not inherently regenerate unless being over-driven by the generator, which is unlikely to happen. That would explain why you are not seeing any regen. There are slip recovery systems for the wound rotor that will recoup some of the no-load losses but as others have said, in the use profile you have the savings will be difficult to justify the expense.

About the only suggestion I would have is to see if, when you don't need as much test energy, there is a benefit in running the prime mover at a lower speed. If so, you may consider a line regenerative VFD for the prime mover. That will also allow you to almost fully regenerate and recover the acceleration energy, then you can leave the unit powered off in between tests. Very expensive though if it's a large motor. If you work for Siemens though, your internal cost for a fully regenerative VFD will be a lot lower than it is for the rest of us!


"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[ScottyUK]

Would this large facility have the initials I, P, and H by any chance?


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

If we learn from our mistakes I'm getting a great education!

 

[LionelHutz]

Is there any requirement to stop the unit between tests?

Obviously, you're not wrong stating the motor is consuming power in the MWh range because it will do so if measured for enough time. But, that's certainly not an hourly number and you're leaving off the important detail. A 100W light bulb will consume 1MWh of power if powered on long enough. Tell us how many hours it takes to consume 1MWh of power. Tell us what percentage of the time you are testing vs time the machine is idling. Then, we can have an idea of the losses of the machine.

When it is just spinning idle I'd expect the losses to be something like 6% or 150kW for a 2500kW motor. This is 150kWh for each hour it runs idle. At $0.2/kWh this would cost $30 an hour to leave it idling during the work day.

To regenerative brake the motor you will have to install a 2500kW, 6kV VFD. This could run something like $300k by the time you get it installed. That would take well over 5 years to pay off if you worked 250 days a year and 8 hour a day and shut the motor down every time you were not testing. It would also use more power while the motor is running, which would further hurt your payback time.

You will have to do some calculations like the above to figure out how much energy you could save so you can confirm if the solution is in fact valid. You measure the kW (not kWh) of the motor when running idle and calculate your losses due to leaving it running idle. Then, decide how much of the running time can be changed to off time. Your savings will be due to changing idle running to off time. This really isn't that hard a project which, once again, is why I'm stunned that this can't be figured out internally.

Also to repeat myself again, forget trying to "feed back" electricity while the machine is left running between tests. It's nothing but a fool's errand, that is unless you are a firm believer in over unity or perpertual motion.

 

[jimmynora]

@electricpete
The info on the nameplate I wrote in my opening post. The Generator runs as long as the tests are being done. Usually a test procedure goes for 1 hour. When there is a long break between procedure the generator is either turned off... or slowed down to like 1000 rotations/min.

I am taking measurement from the 6kv feed-in for the motor that runs the generator.From my measurement today I can give you an example of what I saw:
2 tests procedures each of about 1 hour were conducted.
Total electricity used overall 2.3MWh
Total electricity consumed during test procedures: 1.6MWh

Clearly shows that even ignoring the start-up electricity consumed to get the generator up too full speed twice today there is a potential to somehow save electricity in some way.

@rasevskii
Yes the generator is very old and uses mechanical braking to slow it down. A fan keep the system cool. One suggestion is to use the mechanical braking energy to feed-in energy if it is feasible.
We have a drive system used to maintain a constant speed and control the frequency of the generator but clearly not very high-tech. New VFD or regenerative VFDs seem like an interesting solution.

@jraef
The generator has to rotate at 3000rpm because we need a 50Hz AC frequency from our 2pole generator. Can you please explain this slip-recovery system that we can use to regenerate electricity ?

So a fully regenerative VFD to power the generator instead of the asynchronous motor (slipring rotor) is the best bet ?

 

[wolf39]

I'm surprised that six of the most competent members are responding to a post of this calibre.

By the way: I've calculated the rotational energy of one of the world's largest hydro generators. The rotor of a 825 MVA Itaipu unit is weighing about 2000 tons but the energy stored at rated speed is equivalent to only 1000 kilowatthours.

Wolf

http://www.hydropower-consult.com

 

[jimmynora]

Ok so far there have been 2 concrete recommendation. Correct me if I am wrong

i) Slip Ring Power Recovery Sytem (SPRS):

A variable speed drive for slip ring induction motors. It recovers and delivers the slip* dependent rotor power from the motor to the grid.

This sounds more promising since it can be integrated in the normal circuit during the operation of the motor/generator.

ii) Regenerative Variable Frequency Drive:

I guess this may be a solution for generation of electricity during the braking process. Apart from that I do not see how we can use this. Any suggestion ?

 

[Skogsgurra]

I think I am getting even more confused than jimmynora now.

There is a 'drawing'attached to the 31 May 11 8:37 post. It shows a three-phase motor fed from a 6 kV grid. No indication as to what kind of motor it is but the text says slipring motor and the specification says 2975 RPM and PF .9, which is plausible. There is no easy way of controlling the speed of a standard asynchronous motor (if it doesn't have a cascade arrangement, which has not been indicated).

In your 16:20 post, you say that "the generator is either turned off... or slowed down to like 1000 rotations/min" and that confuses me a lot. How do you reduce speed to 1000 RPM when running an insduction motor with rated speed 12975 RPM? I think that we all need to understand what you mean by that statement.

There is an obvious need for improvement - your numbers show that.

"2 tests procedures each of about 1 hour were conducted.Total electricity used overall 2.3MWh. Total electricity consumed during test procedures: 1.6MWh"

Does that mean that each test consumes 800 kWh and that (assuming a fully utilized generator) each test has a 20 minutes duration. Would that imply that the 700 kWh is consumed by the idling machine during two 40 minute periods? Then, the losses are huge. More than 500 kW loss power. That is twenty percent losses. No wonder your company wants to do something about it.

But, these numbers are not very realistic. They mean that your efficiency is a low low 80 percent when idling! And probably a lot less when running fully loaded.


See why we think that you should check both your numbers and your thinking? And - how the heck can you run the machine at 1000 RPM? That's the biggest mystery.


And what about cooling? It must be getting pretty hot in that room. I guess that cooling costs are even higher than cost for electricity for the motor.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Yes wolf - it surprises me too.

But you know how it is; you think that you have all the facts and that the OP understands basic electricity.

By the time you discover that he doesn't, you sometimes leave him alone or you try to understand what the problem is.

I have now reached confusion state. I hope to get some facts soon.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

OK, but the Itaipu machines are not spinning at 3000 RPM. Speed squared, you know.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jimmynora]

@LionelHutz
The breaks between testing procedure are usually because of rest breaks for employees. There have been cases that the Generator runs 4 hours and the actual time for testing is only 2 hours but spread over this interval. This is why we are looking at if and when we should turn off or keep it running and but back the energy in the grid (either at normal speed or slowed down).

I am a bit sceptical about the braking regenerative theory because I doubt it will be feasible. I am more favouring a slipring recovery system kind of solution. Allowing us to feedback the excess/unused electric energy when the generator is running and no test is occurring.

@Skogsgurra

It is a normal asynchronous (induction) motor having a slipring rotor. It is being used as a pony-motor to power up the generator.
The drive system controls the speed of rotations, hence the frequency of AC.

When there is a pause of about 30minutes or more between tests sometimes the generator is turned off. Or in some circumstances the speed is just lowered down to 1000 rotations/min so decreasing energy taken from the grid. Our drive control displays the rotational speed and when can run it at any speed we want. Keep in mind we have a mechanical braking system apart from ways to increase the resistance of the circuit to slow down the rotor.

The statistic i wanted to share is that about 400KWh of energy we wasted today (after subtracting startup energy and other losses)during which the generator was running idly. Theoretically this is the energy we want to save some-way... perhaps by feeding back in the grid.