Help a mechanical engineer who is playing an electrical engineer

[Girswald]

It is hard to believe UL specifies full current i.e 4000A should actually be passed through the bus bar joint to test it.


You should wait for temperature equilibrium and then take measurement of the temperature.


Check no load output voltages of the transformers: they should be same.

UL specifies that the entire rated current be applied, in this case the device is rated at 3200 amps, so 3200 amps must be fed into the system, it is then divided up so that 800 amps go into the vertical bus system.

UL specifies that temperature readings are taken at 15 minute time periods and only when there is less than 1 degree C change for 3 successive readings are the temperature measurements counted.

The voltages are different, and that is what I think is wrong with our system.

 

[Girswald]

Well, I finally got the test running. It took a whole lot of cable, bus bar and steel.

I will attach some photos of the set-up after it finishes running. This is interesting stuff to see actually working.

Right now we are sitting at a few degrees between samples, so it will be a while until it stabilizes.

Oh, and it is 97 db in the test lab! Steel makes a lot of noise vibrating at 59.97 hz!

 

[Girswald]

150911-1354 EDT

Girswald:

That your not getting equal current on each phase means we need to know more about your source.

First, is your source supposed to be an actual constant current source, but adjustable? At this current level I doubt that it is. More likely it is an adjustable voltage source.

If it is an adjustable voltage source, then are each of the three output voltages the same? Are voltages measured line-to-line, or line-to-neutral?

What are the values of these voltages when you produce your output of 4000 A or so?

Is the system a wye or delta?

.

It is a constant current transformer that we vary the voltage into.
As far as I can tell, the incoming voltages are fine. I measured line to ground (common neutral) but have not measured line to line.

The transformer nameplate states 240/5 or 480/10 and we are getting roughly 24 or 25 volts. That is the part that keeps throwing me. While this test is running, I looked at the 3 transformers with a FLIR camera and the two that seem to be working well are showing heat from top to bottom, but the one phase that isn't working well is only hot on the top half. I think it is a mechanical issue...

 

[Girswald]

A couple of pics of the set-up

 

[Girswald]

A couple of more pictures

First, the horizontal short, 10 @ 10' 500 MCM with 6 pcs 3 x 1/4 bus. Also, you can see the thermocouple wires running into the device to measure the temperature rise.

 

[Girswald]

Another of the vertical short, 3 per phase of 10' 500 MCM. Also note the 12 gauge steel chokes to trim the resistance in each phase.

 

[Girswald]

The feed bus, note again, the 12 gauge steel used to choke the incoming current, the steel reached over 150 degrees C, that's as high as the FLIR can measure. The silver is aluminum thermocouple tape to hold the steel from falling off the bus, even though it took a lot of force to break the magnetic bond, the 60 Hz vibrations would eventually make the steel fall off. That color is only from heat!

 

[Girswald]

By the way...I want to thank all of you for the help and your suggestions. The new connection was well under the maximum temperature rise and this could be a $30k per year cost savings

I hate going on a forum and helping someone out and they never respond to let the forum know how things turn out...it is just common courtesy to let someone know if what they added helped.

 

[gar]

150915-2100 EDT

Girswald:

A tightly coupled transformer, mean a very high permeability magnetic core material with no shunt magnetic paths, is simultaneously both a voltage and current transformer.

I believe you might better understand why you had differences in your current readings between the phases if you considered the transformers as voltage transformers, and measured the output voltages of the transformers. I suspect your three different phases had different impedances. This might mean some high joint resistances in one phase vs another, and in turn greater maximum hot spot temperatures somewhere.

.

 

[Sahib]

150915-2100 EDT

I believe you might better understand why you had differences in your current readings between the phases if you considered the transformers as voltage transformers, and measured the output voltages of the transformers. I suspect your three different phases had different impedances. This might mean some high joint resistances in one phase vs another, and in turn greater maximum hot spot temperatures somewhere.

.

It looks like the set up bus bar and steel wrap together acting as a transformer themselves: the steel wrap as shorted secondary. So adjusting the steel wrap may bring about the uniformity of currents.

 

[Girswald]

150915-2100 EDT

Girswald:

A tightly coupled transformer, mean a very high permeability magnetic core material with no shunt magnetic paths, is simultaneously both a voltage and current transformer.

I believe you might better understand why you had differences in your current readings between the phases if you considered the transformers as voltage transformers, and measured the output voltages of the transformers. I suspect your three different phases had different impedances. This might mean some high joint resistances in one phase vs another, and in turn greater maximum hot spot temperatures somewhere.

.

The problem with measuring voltages is that this transformer system is never used unless it is shorted, meaning the voltage line to ground can be measured but line to line is always going to be zero.

Even with the 3 phases shorted at the transformer with a solid copper bar still shows the same imbalance of current.

 

[Girswald]

It looks like the set up bus bar and steel wrap together acting as a transformer themselves: the steel wrap as shorted secondary. So adjusting the steel wrap may bring about the uniformity of currents.

Actually, they act as very inefficient space heaters and very loud noise generators

But, you are correct, simply adding or removing steel will adjust the current flow in that phase. The bad part is that each phase also adjusts the other phases, so its a real game to try to keep everything balanced.

One other issue, when things heat up, resistance changes and the steel needs to be constantly adjusted. If only I could turn the air conditioner to "Absolute Zero" everything would be just fine!

 

[Sahib]

The bad part is that each phase also adjusts the other phases, so its a real game to try to keep everything balanced.

It looks like you are not using the neutral from the transformers. You need to connect the phase leads R Y and B of the transformers to one end of the three bus bars. Short all the other ends of the bus bars together and connect to ground wire and ground wire to the transformers neutral.

 

[gar]

160915-2328 EDT

Girswald:

I don't believe you ever defined your 3 phase bus. Is it a wye or delta for the bus distribution? In other words are there 3 or 4 bus bars.

What do the current transformers as you refer to them and I might call them voltage transformers look like and how do these connect to your bus bar system?

To simplify a concept consider two of your bus bars as one closed circuit, shorted at the far end, and connected to the secondary of the source transformer. Adjust the input voltage to the transformer to obtain a desired and known current in the bus, then there will be a voltage difference between the two bus bars at the source end. This source voltage and the total series impedance of the bus bar loop determines the current in the bus bar loop. Because the end of the bus bars is shorted does not mean that the source input voltage is zero as you stated.

From one of your comments I do not believe that each of your bus bars have the same impedance.

By measuring two bus bars at a time you can determine if there is a substantial impedance difference between them.

.

 

[Sahib]

gar:
The OP already stated that one end of bus bars are shorted out and the other ends are given power supply. But it looks like he has not connected the shorted end to the transformers neutral and so there is unbalanced current distribution due to unbalances voltages!

 

[gar]

150916-0653 EDT

Sahib:

Girswald has not provided an accurate description of the circuit, and it appears that the shorted bus bar load looks like a delta load, and the delta load is unbalanced. This unbalance may be a result of either or both the bus bars or the shorted loads.

.

 

[gar]

150916-0723 EDT

Girswald:

Consider a delta load with the following resistance values:
Between A and B 1.0 ohms.
Between A and C 1.0 ohms.
Between B and C 1.1 ohms.

With this delta closed the calculated measured resistance between any two terminals are:
Rab = 0.6774 ohms
Rac = 0.6774 ohms.
Rbc = 0.7097 ohms.

Change the resistances to:
Between A and B 1.00 ohms.
Between A and C 1.05 ohms.
Between B and C 1.10 ohms.

Then the calculated values are:
Rab = 0.6825 ohms
Rac = 0.7000 ohms.
Rbc = 0.7159 ohms.

Instead of closing the delta just short one pair of bus bars at a time, then for:
Between A and B 1.0 ohms.
Between A and C 1.0 ohms.
Between B and C 1.1 ohms.

The results are:
Rab = 2.1 ohms
Rac = 2.1 ohms.
Rbc = 2.0 ohms.

For the other combination of resistances:
Between A and B 1.00 ohms.
Between A and C 1.05 ohms.
Between B and C 1.10 ohms.

The results are:
Rab = 2.15 ohms
Rac = 2.10 ohms.
Rbc = 2.05 ohms.

Sahib has suggested connecting the load as a wye. This will make adjustments easier, but requires your bus distribution to consist of 4 bars.

Someone ckeck my math.

If the bus bar resistances are not close to equal, and about what they should be, then this implies a bad joint somewhere in at least one location. If this occurs, then there will be excessive heating in any location with a poor joint.

.

 

[Sahib]

gar:
I guess his is star connected from this:

I am testing thermal rise in a copper bus system by loading current and shorting out the opposite end of the bus system, it is 3 phase bus and I have a 3 phase current generator.

 

[winnie]

gar:

From the pictures in post 25 and 26, the 'shorts' are wye connected. In post 25 you see three sets of wires (10x500MCM bundled by cable ties) going to a set of two hole lugs all bolted to a short, massive bit of bus bar. In post 26 you just see the wires, but not the short.

Sahib:

I think you are right about the steel acting as a shorted turn.

Girswald:

Even though the transformer secondary is shorted, there still has to be some (very small) voltage between the output terminals. This is necessary because the bus bars have to have _some_ resistance. This resistance means you have to have _some_ voltage drop, and some voltage that you can measure. The measurement would not be trivial; you would need to measure a small voltage with lots of magnetic field interfering with your measuring equipment...but the voltage will be there.

Could you post the dimensions of the bus bars?

As an example, a 1/2"x6" bus bar has a DC resistance of 2.74 _micro_ ohms per foot, is large enough that 'skin effect' will be an issue, and will also have an inductance that is significant at 3000A. Imagine a 3 phase bus assembly 100 feet long and shorted at the far end, with 3000A flowing on each phase. The phase to phase voltage at the supply end will be on the order of 1.5V. (Approximated using DC resistance only, real value should be higher.)

As a separate point: you describe the steel being used to 'choke' the current. The current in the wire induces a magnetic field in the steel, and that magnetic field induces a voltage that opposes the current flow in the wire. This 'choke effect' depends upon the magnetic properties of the steel.

As Sahib notes, current is also going to flow in the steel, heating it up. Not only will the heat change the resistance of the copper, but the heat will change the magnetic properties of the steel, and could cause damage to surrounding materials (it looks like some of the bus bar insulation has been melted away).

If you can, try to get steel 'chokes' made of transformer steel that is laminated into thin sheets. This sort of steel has high resistance, and the thin lamination material breaks up the induced current flow. I would suggest split cylinders made of stacks of 'washers' made from transformer steel laminations.

-Jon

 

[Girswald]

To answer some of the questions posed, I will refer to the current generator wiring diagram:

Incoming to our lab is 460VAC, 3P 60Hz, 200A, that runs into a dry-type transformer that is delta to wye with lighting circuit (unused) 150kVA 460/230, the neutral is not used and simply goes to a ground bar at the final current generator. Next step is that the 3 phases go to the variable transformers which are rated at 240 to 0-240VAC. The 3 phases then connect to the final large current transformers, T1 and T2 go to one, T2 and T3 go to the second and T3 and T1 go to the third. No neutral is available. The output of this current generator is a stack of 3 per phase 6 x 1/4 copper bus with shrink tube covers. Again, no neutral bus is available, but we do have a 250MCM lug wired back to the original dry type transformer that is used as a ground for the metering and indicating lights on this system.

This is all taken off the wiring diagram and I think it is pretty accurate, although for some reason I think there may be 4 bars per phase on the output.

Winnie...You are correct that better steel would work better, but we are a sheet metal fabricator and commercial grade 1020 we have. We go through 25 tons of it per day! We measured one set of steel chokes at 390 degrees F today. We can keep the current remarkably stable with this method, between 790 and 810 amps for about 5 hours today.

Gar...I am a mechanical guy, those numbers mean very little to me being that they are so small. If I short the current generator directly on its output point, what and how would I measure those values? Would a Fluke meter be good enough? We do have a device to measure resistance of ground bonds and it uses a 10 or 100 volt probe that you place on opposite sides of the bond and it measures very small resistance values.

Sahib...Would I need to size the ground/neutral bus and its connections per the amperage that we are developing, i.e. do I need 3 @6 x 1/4 bus bars for it? I seem to remember using a single cable at another lab I worked for regardless of the induced current, but I could be wrong...