Each reactor was fused, and they did blow, just not before the reactors and wiring burned up.

It is my opinion that the design must be wrong. Millions of circuits all over the world blow fuses and pop circuit breakers without burning to the ground. If the fuses did not keep the wires from burning, the test rig is designed wrong.

 

Okay, so it seems I have an engineer at work arguing that the problem described, should not have caused the reactors to burn up, because the taps that were shorted were 277v and 346v while running at 277v. Had it been the other way around, running 346v it would make sense. This would certainly be the case in a resistive load, but it's not a resistive load.

Wire one up with amp clamps on all three tie in points and then start applying power to it though a variable source. Odds are the high amp readings of the circulating current between the 277 and 346 taps in a short circuit condition ought to convince him otherwise being as we all here agree a single winding multi-tapped inductor operating on AC is just another name for common autotransformer.

Just because someone has a piece of paper hanging on their wall doesn't mean they know Jack Shyte about the actual applied reality of what the paper says they are supposed to know.

 

Wire one up with amp clamps on all three tie in points and then start applying power to it though a variable source. Odds are the high amp readings of the circulating current between the 277 and 346 taps in a short circuit condition ought to convince him otherwise being as we all here agree a single winding multi-tapped inductor operating on AC is just another name for common autotransformer.

Just because someone has a piece of paper hanging on their wall doesn't mean they know Jack Shyte about the actual applied reality of what the paper says they are supposed to know.

Yeah, i agree. I planned on setting up a simulated scenario to prove my theory. It's difficult sometimes, to convince someone that they are wrong because of the paper they hang on the wall of their office. Even if my test proves it, I'd really like to show the math as well.

 

The math seems simple to me. Shorted turns being powered with the magnetic field of the core resembles a short circuit with a voltage applied. I = E/ (nearly zero)

 

And other than the math, I'd suggest you build a diagram of the entire assembly by physically tracing every wire to every component. Just to make sure that things were actually set up the way they were supposed to, and that no mistake was done when it was put together.

 

And other than the math, I'd suggest you build a diagram of the entire assembly by physically tracing every wire to every component. Just to make sure that things were actually set up the way they were supposed to, and that no mistake was done when it was put together.

I agree. In addition, I suspect there is something wrong with the fuse ratings or their wiring.

 

Thank you everybody for your input. I know there are alot of details i did not go into, and it leaves the situation open to speculation. The load banks, or test rigs as they've been called here, were built professionally by a company that specializes in them, from a standard design. We have dozens of similar ones in our facility, and this particular one has been operating without issue for 2 years. The problem that caused 2 taps to be shorted, was a programming issue caused by our own guys. But the wiring is original. As far as the math goes, perhaps i dont fully understand, because the shorted taps are connected equally to in-phase voltage, which would be 0 potential difference. The only difference would have to be a result of phase rotation in the inductor, and I dont think that can be accurately explained without the math.

 

Current is flowing round the short by induction (magnetic coupling) not by the direct connection of line conductors across the short. What matters is the voltage from the neutral to the line being applied to the non-shorted section, creating a magnetic field that controls how much power is stored or consumed by the inductor due to the shorted taps that would normally have a fair amount of voltage across them in a open condition..

 

Who signed off on the QC portion of production? We're the checks they were supposed to perform actually done?
What parameters do you check before hitting it with full power?

The taking of an employee scalp in full view and hearing of other people often has a very sobering effect on the quality of production



Unless maybe it is your scalp they want???

 

Ran a controlled test today, set up with 120v, but the same idea. Turned on power with the 120 tap connected to 120V, with a clamp meter I measured about 1A. Then I closed a contactor and connected the next tap down the line, 139V,in parallel with the 120V, and it smoked, burned the wire and the reactor. The clamp meter read 60A and slowly fell to around 45A when I turned it off. Circuit was protected with a standard 20A panel breaker, but it did not trip. I did another test using a tap further down the line away from neutral, this time, the breaker tripped instantly, I measured an inrush of 201A.

 

Congratulations! You proved your company engineer wrong with nothing more than a basic test so for your reward don't be surprised if you get let go from the company for some undefined reason.

(Been there done that more than once.)

 

and it smoked, burned the wire and the reactor.

You used a second reactor for your test?

 

Ran a controlled test today, set up with 120v, but the same idea. Turned on power with the 120 tap connected to 120V, with a clamp meter I measured about 1A. Then I closed a contactor and connected the next tap down the line, 139V,in parallel with the 120V, and it smoked, burned the wire and the reactor. The clamp meter read 60A and slowly fell to around 45A when I turned it off. Circuit was protected with a standard 20A panel breaker, but it did not trip. I did another test using a tap further down the line away from neutral, this time, the breaker tripped instantly, I measured an inrush of 201A.

Great, all the theory in the world is worthless if a correct experiment proves it wrong. Normal thermal magnetic panel line breakers can have a long time delay for 2-3X over-current but not fault currents. Trip times can up to tens of seconds.
http://myelectrical.com/notes/entryid/159/electromechanical-relays

For your testing application you might want something more sophisticated and adjustable.

http://electrical-engineering-portal.com/the-basics-of-circuit-breaker-tripping-units

 

You used a second reactor for your test?

Certainly. How you gonna prove an experimental result if you start with broken parts?

all the theory in the world is worthless if a correct experiment proves it wrong.

I been working on that one for a while. Right now it's at, "One experimental result is worth a dozen theories."

 

Certainly. How you gonna prove an experimental result if you start with broken parts?

Good point... but he did say that the original malfunction cost around $10G of damages ... I just wonder how much more dough he baked for this experiment, that's all...

 

Good point... but he did say that the original malfunction cost around $10G of damages ... I just wonder how much more dough he baked for this experiment, that's all...

I would hope that company engineer signed off for the test saying "nothing will happen'.

 

I would hope that company engineer signed off for the test saying "nothing will happen'.

... and put Homer Simpson supervising the safety of the experiment...

 

Okay, so it seems I have an engineer at work arguing that the problem described, should not have caused the reactors to burn up, because the taps that were shorted were 277v and 346v while running at 277v. Had it been the other way around, running 346v it would make sense. This would certainly be the case in a resistive load, but it's not a resistive load.

Give the guy this.