Large Pump Starting

[Rodmcm]

Ive posted this on the motor starting forum but perhaps someone mechanical can help. We are looking how to start multiple centrifigal water pumps each at 2MW, The problem we have is that the electrical system is weak and the simulation shows voltage drops too big to contemplate. One option thought about was donkey motor starting, ie using a small motor to accelerate the motor and pump and then switching on the starter, the pump manufacturer has come back with a 1MW motor starter for each pump!. How does he derive this value against closed head? Is delivery bypass an option. Any 'sensible' solutions out there?

 

[StoneCold]

I am in no way a large pump expert but you may ask the pump vender if you can start the pump up on just minimum flow spill back and then open the discharge once the pump is up and running.

 

[saxon]

Rodmcm, Sounds like a similiar problem at a plant I once worked in (polyplant). We had ethylene Hyper-compressors that had Hi-HP Synchronous motors that were tied directly to the power grid. Every time we started and stopped these things, we had to notify the power company 24 hrs in advance so that they could gear up or gear down the system so that brown outs or power surges didn't occur in the grid region. Also, if the unit didn't start up correctly, the whole unit had to be shut down for 48 hrs. to allow the windings too cool off before start up was tried again.

Check with the power company. They will give you start up and shut down requirements to minimze loads on the grid.

Hope this helps.
saxon

 

[smckennz]

Hi rodmcm

you have a few options to pursue:

Have you looked at autotransformer start or liquid resistance start. The latter requires the motor to have a wound rotor.

Another possibility is installing a mechanical clutch between motor and pump. This would enable motor starting and pump runup issues to be separated. Eaton Airflex and Witchita make suitable clutches although wear can be a issue.Liable to be expensive.

One other possibility is to start the pump with an empty casing and then slowly introduce water when up to speed. There would be quite a few issues to work through with surge, but it is routinely used on hydro turbines so it should be possible to adapt to some pumping operations.

If your voltage drop is caused by line loss, then sometimes replacing the conductors is an option. The power supply company should be prepared to chip in as you will be paying for their power so they will get something back. Alternatively if the lines are overhead and there are not too many spurs, increasing line voltage and replacing transformers and HV switchgear may be an option.

The 1MW sizing of the donkey motor sounds believable as a first guess but would need to be worked out from the pump curve and inertias. A lot of centrifugal pumps have a shutoff power less than half the best efficiency power. But then again, some high specific speed pumps use more power at shutoff. It needs to be worked out. Liable to be expensive though.

Cheers

Steve

 

[Rodmcm]

Thanks Steve/Saxon, this is a new installation under design. We have been through nearly all the 'electical' options ie soft start etc but we have 8 off pumps at 2MW. You can understand the financial attraction of one small soft starter or variable speed drive switched sequencially to 8 donkey motors as opposed to 8 sets of special items. The main problem is the voltage dips in starting and the effect on the local community. We can get away with a 'dirty' bus but others can't. I suppose my question is how did the pump suppliers get to 1MW. They are reluctant to tell us.

 

[rmw]

rodmcm,

Is the 2MW figure you are giving the rated HP of the motor, or the power required to start up?

If it is the rated HP of the pump, then that would be the power draw at rated flow, while start up would be assumed to be against a closed shut off valve, at which point most centrifugal pumps (exceptions noted above by smckennz above) which would make a lower HP donkey motor make sense. The donkey motor has enough HP to spin the pump, but not enough power to operate it.

I can relate to your problem. I once had a client who was at the end of the line at the outer fringe of his local utilities coverage area, and if his boiler ID fan tripped off, he could not start it again until the next shift change, or it would drop out all the other motor starters in his plant. He wasn't worried about his neighbors lights flickering, he was worried about his plants production.

rmw

 

[MikeHalloran]

Given a disconnect clutch, a fluid coupling, and the right gearbox, a 2000HP Diesel should be able to get a pump like that up to nearly 100 pct speed. Diesels that size are not made in large quantities and tend to be pricey.

However, a commodity 500HP truck Diesel should be able to get a pump going fast enough so that the starting surge won't take out the neighborhood. The truck engines may be cheap enough to buy one for each pump instead of installing a lineshaft and multiple gearboxes and clutches. Actually, I'd compare the lineshaft + 8 clutches + _two_ engines, to 8 engines with clutches. No sense designing in a single point of failure.

-Mike-

 

[Rodmcm]

2MW is the norminal rating of the pump at nominal flow

 

[MikeHalloran]

I wonder just how fast the donkey motor needs to spin the pump to get the electrical starting surge to a manageable level?

Will you have any pressurized fluids available? E.g., fairly serious vane- type air motors are sometimes used to start Diesels. No, I'm not particularly enamored of Diesels; I just work with them.

If you didn't mind the mess, you could fit a Pelton wheel to the motor- pump coupling and run it up to speed with a jet of water.

Could the pump manufacturer be persuaded to build an impulse turbine into the pump?

Wait a sec; half is already there. It would be inefficient, but a nozzle directed at the pump's impeller should be able to make it spin up some.

If the 2MW pumps run at a fairly high head, then the only remaining problem is starting the first one. Or you put in a small high head pump, and power it with ... a Diesel.

;-)

-Mike-

 

[notnats]

Mike,
Is it possible that the first one won't cause as big a voltage drop as the last one, because the last one already has 14MW connected to the grid? In that case a donkey may not be necessary.

Of course the crossover piping and valves will be big and expensive.

Jeff