starting 100 kW AC motor using inductance 1

[ebarba]

Hi all,

is there a way to use a 3-phase inductance to start an AC motor?

The whole picture: the motor is in reality a 110 kW asynchronous 3-phase generator (400 V, delta connected, 50 Hz), moved by an internal combustion engine. Because the generator is SO massive it tolls the engine starter way more than it's designed for and we keep burning it (not to mention constantly ruining the battery).

Our "brilliant" solution: use the generator as a motor (is it a brilliant solution???). Problem is there's no starter for this large "motor" on site.

The usual procedure to get this generating plant online is to start the engine, ramp the generator to sync speed and then connect it to mains. To avoid tripping the 250A protection on this line, there is a 3 phase inductance and a smaller contactor in parallel with the main one. First the inductance branch is closed (so we get the current peak "filtered" by the inductance) then the main contactor is closed. Finally we open the inductance branch.

Since this is an Asychronous machine, it is essentially a standard AC motor. If we had a delta-star or a soft starter we could use it to move the generator and start the engine... but we only have the inductances in place. Is there a way to use them to "soft start" the generator in motor mode?

I had the idea to close the contact in the inductance branch with the engine and motor stopped: the inductance hummed and the motor moved about 1 degree. During the test, current in each phase was 210 A and the phase-to-phase voltage in the motor side of the inductance was 50 V... too little to make it spin.

What would happen if we close the main contactor while keeping closed the inductor branch? Will this make the motor run or trip the 250A protection?

The starting power of the IC engine at near-zero rpm is less than 3 kW, plus it'll take about 10 kW to move it at 1500 rpm while it starts (about 10 seconds).

Any ideas to avoid buying a new engine or a delta-star starter?

Thanks!

 

[MatthewDB]

LH: "Gunners hatred of wye/delta starters on motors, where they can cause issues and generally should be avoided, was wrongly applied to this case where you can easily eliminate the transitional issues with a smart sequence."

What does that smart sequence look like? I really want to know. Never seen that. Didn't know that it exists.

You need to leave the motor de-energized for several seconds during the transition.

The problem with Y-D starters is that when the motor is de-energized in wye, significant magnetic flux is left in the rotor circuit. The rotor will continue to roll below synchronous frequency. When the motor is re-energized in delta quickly, the flux is still high but the phase angle has rotated well out of phase. At reapplication of power, there is a huge torque pulse and current draw because the two fluxes are too far out of phase.

In many applications, the motor can't be de-energized for several seconds because it would slow too much while coasting.

Since you have the engine running, you can wait several seconds for the flux to die down and it will be the same as energizing it after a normal start.

 

[MatthewDB]

How about throwing an auto-transformer into the mix? It can give even more torque vs. current draw since the stator current is increased above the line current.

 

[ebarba]

You need to leave the motor de-energized for several seconds during the transition.

In this particular case, we can de-energize and keep it that way even a couple of minutes, it's something we do right now to allow for even heating of the engin, so it isn't a problem.

How about throwing an auto-transformer into the mix? It can give even more torque vs. current draw since the stator current is increased above the line current.

Can you elaborate a little more, for the electrically challenged (i.e. Me)? Do you mean a 1:1 auto-transformer?

 

[waross]

AUTO TRANSFORMER STARTING.
All it takes is money.
The motor to be started is energized through an auto transformer. The taps are set at about 80% and 65%, pick one.
When the motor is energized through one of the taps, the starting current is reduced by two effects;
1/ The impedance of the auto trans reduces the starting current similar to the effect of the inductance that you are now using.
2/ The reduced voltage to the motor results in a lower line current.
The downside is reduced starting torque; 80%² = 64% when the 80% tap is used and 65%² = 42% when the 65% tap is used.
The cost of the starter is similar to the cost of a wye:delta starter plus the cost of the auto transformer.
Both methods use a starting contactor and a running contactor. The wye:delta starter also uses a shorting contactor, but the autotransformer needs a larger enclosure.

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

 

[LionelHutz]

The big advantage of the autotransformer is that it also reduces the line current by the transformer ratio which can help on limit power sources. Using the 80% tap may reduce the current to 80% and the torque to 64%, but it also reduces the line current to 64%. A soft-starter or reactor would still have the 80% line current.

 

[Skogsgurra]

MatthewDB

If THAT is what you call a "smart sequence" I must say that I know more about that than most people that don't even know about the phasing of Counter EMF when switching from Wye to Delta.

I thought that you meant a device that, taking contactor delays into account, waited for the correct phasing and then closing the Star contactor in the correct moment to minimize the transient.

I think that I shall make one so that Y/D starts can be of some use again.

BTW - still discussing starting? Wasn't the problem more on the engine/starter/battery side?

Gunnar Englund

http://www.gke.org

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

 

[LionelHutz]

Gunner. This IS NOT a simple motor driving a load. Why all the harping on about the typical wye/delta starter issues that don't apply in this application? Everyone else already understands that this application doesn't require a fast transition and in fact should not use a fast transition, specifically to eliminate any transition transients that would occur.

BTW, the thing you are going to invent to fix the wye-delta starters already exists.

 

[Skogsgurra]

Where? Are you sure you understand what I am saying?

Gunnar Englund

http://www.gke.org

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

 

[ebarba]

Just wanted to let you all know what the conclusion was: we decided to go with the lower inductance instead of the "delayed" D/Y starter, for simple economic reasons.

After discussing with the inductor manufacturer, he realized that the original one was oversized for the short period of time it works. After re-sizing, a new inductor with taps at about 66% and 50% the actual reactance has a price practical equal to that of a new starter motor, no particular hardware is needed and the labor to replace the existing one is the same as replacing the engine starter. Controls issues are minimal.

On the other hand, the cost of labour and hardware for a D/Y starter is 3 to 4 times the price for the engine starter.

Right now everything is working fine and it was decided that the stringent law of "don't fix what ain't broken" will be applied. In any case, we'll keep an eye on the battery health and have a set up a monitor of the voltage during cranking. Maybe the new starter will solve the problem, but if we burn another one we'll go for the smaller inductance.

Thank guys!

 

[7anoter4]

In my opinion, if the locked rotor current will be 8*Irat at full voltage [400 V] at 65% will be 4*Irat. MCB of 250 A will trip at 970 A in not less than 5 sec.
The starting time for 65% of rated voltage could be less than 1 second.
You have to check the motor inertia moment [for generator rotor-could be around 2-3 kg.m^2 and for DG around of 20 kg.m^2 in my opinion] and the starting torque and maximum torque[Ts=1.7-2.2*Trated and Tb=2.4-2.6*Trated].
Then a series reactor of 0.085 ohm and 650 A for 8 second could be useful, in my opinion.
However, the actually voltage drop will take into consideration the entire supply system impedance-transformer and cables.