Black Start Diesel Generation 4

[HVIke]

In modeling a system with three 2MW diesel generators, feeding a 21KV ring bus through a step up transformer, and thus serving loads off the ring bus through step down transformers--how should the generators and their exciters be sized in relation to load?

The inrush to the transformers can be sizeable, depending on the risidual magnetism of each.

 

[dpc]

What's the largest motor you have to start and what will the running load be when you have to start it?

 

[HVIke]

Largest motor is 100 HP. Running system load will be 6MW's.

 

[davidbeach]

you aren't going to have a lot of success feeding 6MW of load with three 2MW generators. Where is the power to supply all the losses? Where is the reserve capacity to handle motor starting? Where is the reserve capacity to handle the fact that your calculations were slightly off and you actually have 6.01MW? Don't count on getting more than 2.000MW out of the generators. That is cuttings things way too close.

 

[waross]

It depends a lot on the nature of the load.
One standard I have been exposed to is that there shall always be enough generators online that the system can sustain the loss of any one.
For a 6MW load that would require 4 2MW sets running. One or possibly two 2MW sets on standby.
If you're crushing gravel, it isn't as critical.
Engine sizing may depend somewhat on your projected operating parameters.
If your policy will be to keep the loads moderate, 80% or less, you may get by with 110% engine size or less. If you're going to be pushing the loading as close to 100% as possible, you may consider sizing the engines at 125%. Spend some time browsing Caterpilars Marathon web site. Compare the set ratings and engine sizes for various applications. Prime power, Standby, Emergency.
Re the transformers; will you be using one main station transformer or will each generator have a close coupled transformer?
What size step-down transformers will you use and how many? Will you have to energise the ring with full load connected after an outage or can you stagger the energization of the step down transformers when you are black starting?
yours

 

[itsmoked]

Reiterating waross's last question:

If it is a ring structure running near 100% how do you deal with having all the load drop due to a fault then trying to start the whole network with only 100% available in 1/3 chunks. Any one generator will always be overloaded when trying to put them back on line.

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[ScottyUK]

Regarding inrush: if the transformer is close coupled to the generator with no interposing LV breaker, effectively forming a 21kV generator, there is no inrush and the generator AVR will ramp up the voltage gracefully. This is a very common design for large installations because it avoids use of very high current LV breakers and reduces the possibility that loss of one large shared transformer will take out a number associated generators.

I agree with the others - your system appears to be marginally sized for the load - I hope you don't have any large drives on this system because the starting current might tip it over the edge. It would be nice to think this wasn't a life critical system considering the lack of redundancy in the available plant.

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

I don't suffer from insanity. I enjoy it...

 

[cflatters]

I concurr with the above - your calculations are too close.

Is your 6MW load calculated or measured - ie have you taken diversity into account ?

If each of the generators and their associated transformers are 2.5 MVA (ie 2MW @ 0.8PF), then each generator transformer will have copper losses of approximately 25kW and perhaps 2kW of iron losses, thats another approx 80kW of load losses in total.

Taking your LV cable losses into account too you may find you have lost 100kW before you start.

The motor starting conditions need to be looked at very carefully, you should get the characteristic curves from the alternator manufacturer and check these carefully - I suspect you have a problem with this !

 

[waross]

Is this an actual project or a school assignment?
yours

 

[HVIke]

Thanks for the responses. Have been out this morning.

This system is hypothetical at this time. The client (a casino) presently uses the three backup generators to feed load at the 480V level, in the event of a loss of utility service to the ring bus.

They are considering feeding the MV ring bus through a single transformer with an automatic transfer-make before break to keep motors on line. There are 6 MW's of inertial storage for ride through.

Therefore, not only would the main GSU have to be energized, but also all the downstream step-down transformers feeding LV load.

Years ago, I ran into a problem where energizing a main step up transformer (40MVA) at a GE Frame 6 facility after an outage would cause the system voltage to dip sufficiently to trip a nearby sister facility. We scratched our heads for some time, but ultimately attributed the cause to a weak end of the line system, combined with the risidual magnetism--or lack of it.

The trips were intermitant. thus the inrush to the transformer was dependent upon what the voltage and current had been when the plant went black.

I am thinking a major problem with the casino's proposal might be transformer inrush while maintaining load. The inrush is 100% reactive power that must be supplied.

 

[waross]

Generally, the GSU is close coupled to the generator and rides the voltage curve as the voltage builds up at the generator terminals, that's the point of them. It gets a large transformer energised without inrush, but it's not a nice thing to do to motors and other loads. I know you said this ScottyUK. Good advice bears repeating, sometimes in different words.

Here's a suggestion to throw around.
Transfer switch location.
At the junction of the ring bus, the GSU and the utility feed.
On utility failure a start signal is issued to all three generators. ONE generator is close coupled to the GSU and brings it up gracefully.
The generators syncronize as soon as possible. With no load, you may be able to use settings with more latitude than if loaded machines were being synchronized.
Now, you have three generators running and the main transformer energized.
You can close the switch to the ring bus, or better sectionalize the ring bus and energise in two or three steps. If sectionalizing is not feasable, bring the lighting on asap and stagger the restart of the motors over 15 or 20 or more seconds.
As for loading; If this is an operating system I will take a chance and assume that the losses of the 480 volt distribution system are being replaced with the transformer losses for a net gain.
At this point you can use make before break or closed transition to return to utility power when it is available.

Another issue with a large closed transition generator installation is fault currents from both systems simultaneously or even an out of phase connection between the systems. This is a possibility in the event of an issue with the transfer switch synchronizing circuit or a transfer switch failure. One of our other posters mentioned this possibility recently. I hope he will reprise his good advice.

Comments welcome gentlemen. This is presented for discussion.
respectfully

 

[MikeHalloran]

If you can't divide the load into single- generator sized pieces, then you logically need to size every generator to handle the entire load, at least for long enough to bring up an additional generator.

[ Okay, I know zip about casinos and almost as much about power generation. I've been marginally involved in power generation on large yachts, where there is usually one smallish generator for black starts, and the remainder are sized so that the primary generators acting together can handle the entire load, with at least one primary generator assumed out of commission for repairs. ]





Mike Halloran
Pembroke Pines, FL, USA

 

[waross]

Hi Mike
The technique is to synchronize and parallel the generators before applying the load. In effect making one big generator out of three smaller machines. From your mechanical background you will realise that the governor settings and characteristics have to be closely matched.
It works. I saw a mill on diesel power that needed both diesel generators online together to start the hammer hog motor. It was a problem on cool mornings. Never a cold morning in the tropics, but cool.
One generator had an electronic governor and the other had a hydraulic governor. First thing in the morning, when the oil was cold, the hydraulic governor couldn't respond as fast as the electronic governor. The EG machine would hog the load and trip its breaker before the other set could catch up. Once the oil warmed up, the sets worked and played together well. Shuting down and restarting at mid-day was no problem.
respectfully

 

[windie]

Or can you keep the GSU energized at all times;
normally using utility power, and then during generator start up and synchronisation with your 6 MW of "intertial storage".

This could depend on whether the system has enough grunt and storage to keep you 21 kV system alive while you start up and sync the generators

 

[ScottyUK]

Hi windie,

Check the title of the thread. 'Black start' is a term usually related to recovery from a total grid failure, so there would be no utility. In the event of a total grid collapse, the black-start-capable power plants will be the first stations up and capable of providing power for other plants to run their own auxiliary services and re-establish themselves. Typically a black start scheme will start a small aero-derivative turbine from a diesel generator, and the aero-derivative drives a generator large enough to run limited plant auxiliaries to start up a larger machine.

The casino application just sounds like a normal standby power application - nothing to do with black start.

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

I don't suffer from insanity. I enjoy it...

 

[waross]

Hello HVIke;
Would you please elaborate on
" There are 6 MW's of inertial storage for ride through."
I'm wondering if you have 6 MW of motor load and want to make a closed transition back to utility power, or if you have a fly-wheel based Uninteruptible Power Source, or something else alltogether.
respectfully

 

[HVIke]

Hi Waross,

I do not have the details of these inertial storage devices. However, the client has told me they are vertical, with the mass storage below the motor/generator on a vertical axis.

I have concluded that they would only be useful in, as you say, paralleling back into the grid.

It seems as thought the generators, since they are not operating prior to separation from the grid, can only be brought up one at a time, with the first unit operating as the Isochronous unit.

Therefore, inrush into the transformer is not much of an issue. However, the loads would have to be powered up incrementally as the generation became available.

Any other thoughts?