Need a 8000A-10,OOO amp magnetic field? help? 1

[drax]

Our customer wants a magnetic field around some bus bars for testing. Could anyone help me with advice on how to do this. I have some 10,000:5 C.T.s....could I push 5 amps into them and if so, at what voltage...could I use a variac to adjust the field level?

thanks

 

[davidbeach]

You aren't going to push 5A into the secondary of a 10,000:5 CT and get 10,000 out the primary. It just ain't gonna happen. Saturation will see to that.

 

[itsmoked]

Remember the magnetic field isn't amps it's amp-turns.

You can increase the field with amps (hard to come by) or turns (relatively easy).

Get a bunch of big arc-welding cable and an arc-welder. Wrap your target and go for the gusto.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

An arc-welder isn't ideal for that. A spot welder has a lot more current available. But usually for very limited time (spots take about half a second to make).

On the other hand, if you have a 20 kA spot-welder and run it at 25 percent load, it can probably be run long enough for your experiment. And, as smoked says, use as many turns as you need to produce the ampere-turns.

BTW: Is it an AC or a DC field that you need? There are both kinds of spot-welders. Some may even be switchable.

Gunnar Englund

http://www.gke.org

 

[electricpete]

The CT approach sounds reasonable to me.

Saturation is avoided as long as the impedance of the the high-current loop is small enough.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Keith's suggestion for multiple loops may or may not work depending on what aspect of the field the customer is trying to recreate. You can easily get the same field magntiude but the flux distribution will be different than for the busbar carrying 10,000A.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[drax]

Your assumption is correct on the saturation of the copper. Out of the two CT's we were able to push only 900A at 277V. The CT's are rated 400V, so we could have gotten more out of them, but not even close to the 7000A we are looking for.

The time restraint is 90 seconds.

We have gathered p additional driver CT's from around the plant to help push the power, but we believe we are constrained by the available power in the lab. We would need a new line run into the lab.

I think I would need three welders to run a three phase circuit?
How would I calculate the cable turns around the copper? I have a 3000A source.
thanks

 

[ruggedscot]

If this is a one off test and you need to do it then speak with one of your local generator suppliers - say Aggreko or the like and ask them. Using a gen set supplying a transformer you should be able to generate your current levels required for the test. Just make sure that you have adequate insurance !

Some panel builders use a genset and transformer to generate around 40-60kA short circuit withstand tests.

Rugged

 

[itsmoked]

Yo drax! What are you trying to test for? We still don't know what magnetic configuration you need.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ScottyUK]

Hi drax,

Copper doesn't saturate: the iron core does. You will need a very large CT with a high kneepoint voltage to be able to run it in reverse as a useful high current source. the principle is fine but, as ePete has correctly said, keeping the impedance of the high current path low enough could be tricky. I have an old Class X CT off a grid transformer put to one side for the day when I get round to building a heavy current injection source. The core weighs more than I do: the secret with this sort of thing just comes down to sheer volume of iron - well strictly cross-section area of iron! - to avoid core saturation.

You should able to get hold of a primary injection test set capable of delivering 2kA or maybe 4kA into a low impedance load. If you use a mains-powered PITS, there should be no problem in paralleling two of them if they are fed from the same mains supply phase. They are fairly common pieces of equipment in the utility industry.

Ruggedscot is also potentially on the right path. We have done similar tricks using a 2.5MVA 11/0.415kV transformer fed from a 415V generator connected to the 11kV winding. The 415V winding effectively becomes a continuous rated 26.5V 3500A source. You could cheerfully overload it for a short period without causing any significant problem to the transformer.

I'm a little puzzled what your customer is trying to achieve: is this for fault withstand tests, or for magnetic compatability with other equipment, or some other purpose? You are looking at some fairly high power test rigs which will be quite costly to set up - is there an easier way to achieve the same objective?


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

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

 

[waross]

Hello drax
Let's do the numbers on your CT.
Ratio 10,000:5 or 2000:1
At 277 volts, your open circuit secondary voltage will be 0.1385 volts.
The impedance of the secondary path includes buth the bus bar impedance and the impedance of your connecting cables.
If the impedance of your cable is three times the impedance of your bus bar the voltage drop across the bus bar will be only 1/4 of the total voltage, or 0.1385/4 = 0.034625
Current:
With 900 amps the primary current would be 900/2000= 0.45 amps.
Suggestions;
If possible mount the CTs on the bus bars and use bus bar jumpers to short the ends of the bus bars. If you can mount two CTs in series on each bus bar you will get more voltage to work with.
Saturation:
Saturation is exceeding the allowable flux densiy of the iron. Flux density is dependant on amp turns. Depending on the class of the CT you may be able to go to 10 amps or more without saturating. The limiting factor will be the insulation strength of the CT winding. It is probably good for 10 amps or more without saturating.
Fun with numbers: this is to give you a feel for what you can change and what the results will be.
Impedance. Use the bus bars to form the return path for each other. If you can reduce the impedance by a factor of 4, your current should go up to 4 x 700 = 2800 amps. (Factor 4)
Voltage: At 400 volts you current should increase from 900 to 400/277 x 700 = 1300 Amps. (Factor 400/277=1.44) At 1000 volts (If the insulation doesn't fail) the current will be 900/277 x 1000 or 3250 Amps. (Factor 1000/277=3.6)
2 CTs in series = 1800 amps (Factor 2)
If you reduce your impedance by a factor of 4, increase your voltage by a factor of 3.6, use 2 CTs per phase for a factor of 2, Your total current increase will be:
900 Amps X 4 X 3.6 x 2 = 26,000 amps.
Your project may be possible if you can reduce the secondary circuit impedance and increase the secondary voltage.
You are nowhere near saturation, your secondary voltage is too low and your secondary impedance is too high.
There have been several suggestions to use a generator. The reason is two-fold. First to avoid drawing too much current from the utility. Second to avoid inrush currents.
If you get this setup to deliver 10,000 amps to the bus bars under steady state conditions, the energisation inrush may be over 50,000 amps and that could quite easily convert your bus bars into ballistic misiles.
Use a generator or a variac to bring the voltage up slowly.
respectfully

 

[drax]

Thanks for all of your help, it looks like a generator is the way to go. This test is for magnetic reaction around equipment. One engineer stated this about looping welding cable around the bus bar, do you guys agree?

"It doesn't make sense to me based upon my college days as well as a few years of practice. There is very poor coupling a low frequencies, which is why iron cores are used. If the welder output was a RF, then this might work, but inductive reactance becomes a huge factor. There is an impedance which has to be overcome. The generator has to be able to generate sufficient voltage to drive the desired current through the impedance. The AC version of Ohm's law V=IZ (Z= R +JX =R=J2PiL) applies. Z is mostly inductance. If a capacitor bank of sufficient size was available to counteract the inductive reactance then we would only have to overcome the resistive voltage drop in the system. That's still doesn't say we can accomplish what we need with what we have. I still think you will need to go somewhere that can generate 12kA of current for 1 minute."

 

[davidbeach]

The use of a generator is probably your best bet. You'd need a big one to handle the 10,000A, say a 3750kVA transformer with a 208V secondary, and a 12kV (anything from 10 - 15kV) primary. Then, feed that primary with a 480V generator. If the transformer had a 12470V primary, the primary current would need to be 1/60 of your 10,000A or 167A. The secondary, the 208V side, would be essentially short circuited, and you would control the generator output to produce the voltage necessary to drive 167A into the primary of the transformer. All will be well, the transformer will be vastly over rated voltage wise, but running right at its ratings current wise, the generator will not need to exceed 480V to produce the 167A. In theory you might be able to get that out of a 150kW generator, but because that short circuit will be essentially all reactive, you may need to go to a 250kW or even 300kW generator to keep the reactive current within the generator capability curve. Then again, it may take a 500 or 750kW generator just to energize that 3750kVA transformer. If you could put a 10,000A breaker on the secondary of the transformer, you could have the transformer connected to the generator as the generator voltage builds and then not have any inrush. You'll get your current on the 208V side of the transformer, but no voltage to speak of; but 10,000A is nothing to be trifled with, even at close to zero volts.

 

[waross]

Hello davidbeach
A question. The op stated a time constraint of 90 seconds. If we went to 5 minutes for a safety factor, could he safely use smaller transformers for a 5 minute test?
Would a variac and a bridge rectifier be adequate for manual field control of the generator output?
Thanks
respectfully

 

[ScottyUK]

Hi David,

Do you have transformers rated 3750kVA / 208V in the US? That seems a huge rating for an LV transformer. What is the typical impedance? We rarely see much over 2.5MVA at 415V because switchgear becomes expensive and fault level gets very high. Our typical impedances are about 6% or so at 2.5MVA which keeps things to a reasonable level.


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

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

 

[davidbeach]

Scotty, I didn't look to see if a 3750 with 208V secondary was actually available. I would imagine that the 3750 would be something well above base rating to get there. The rating isn't the real problem, the problem is the ability to carry 10,000A and there is very little designed for those kinds of numbers. It might well be more practical to parallel several smaller transformers, say five 750s. His 90 seconds is way beyond fault testing setups that use capacitor discharge to get very high currents/voltages; no stored energy for a 90 second test.

 

[ScottyUK]

Agreed - the interconnections will be difficult to arrange at 10kA on anything not designed for the purpose. That was the reason for my suggestion of using paralleled primary injection sets - probably easier to arrange than a gang of utility transformers.

The Programma range of test equipment - now swallowed up into GE - has had a couple of additions since I last looked. The Oden primary injection set can now deliver up to 22kA at 4.3V, which I find pretty amazing. There are numerous smaller variants. Page 8 of 10 in the following link shows the big 'uns.

http://www.gepower.com/prod_serv/pr...ds_new/down_en/test_inj_testing/odenat_en.pdf

As commercially available equipment appears to be out there, I suggest using it. High energy tests using temporary rigs are best avoided if possible - too much scope for things going wrong in the hands of the inexperienced.

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

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

 

[davidbeach]

Scotty, interesting. He'd have to use the 3 minute rating, for which that unit is rated 7700A, close but not quite the lower limit of his stated range (but maybe close enough). I shudder to think what a unit like that might cost though.

 

[ScottyUK]

I guess you could assume the current rating for a time of 90 seconds is greater than that for 3 minutes, so it might scrape by at 10kA for 90s. The applications guys would know. It is by far the most powerful commercially produced PITS I have seen, and I bet it costs an absolute fortune - their smaller sets aren't exactly cheap!

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

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

 

[waross]

Respectfully fellas. He got 900 amps out of the CTs and he only wants 7000. Consider reducing the impedance and increasing the voltage on the CTs. posibly two CT's in series. (With 10,000 amp CT's he needs more voltage not more current.)
he needs 7000/900 more than he has now. That's only a factor of less than 8.
He can probably get a lot of that by eliminating the secondary cables and using bolted bus bar connections. He will not be overloading the current transformer's KVA capability as long as the primary current is below 5 amps.
The insulation on the CT primary should easily withstand close to 2000 volts.
Why does a transformer over heat and fail on over voltage? The rated current has been exceeded. As long as the primary current is below 5 amps, we are not saturated and we are not overheating the transformer.
The huge transformers that have been discussed will be putting out about 1 or 2 volts instead of 480 volts. That's an under utilization factor of 240:1 or more. That's OK if you have the transformer in the yard and can borrow it.
Let's look at what he does have. Reduce the secondary impedance and increase the primary voltage and the 8:1 factor we need is probably within reach with the transformers on hand.
respectfully