Wye-Delta Transformer Feeding 6-pulse SCR Exciter 1

[HenryOhm]

I am investigating possible arcing of connections on the primary side of a wye-delta transformer feeding a 6-pulse, three-phase SCR-based exciter. The arcing led to a small electrical fire so there isn't a lot of evidence left of where things might have started. The transformer is 18kVA and the wye neutral is floating. There is no center-tapped delta on the secondary as the only load is the three-phase exciter. If it relates, the exciter feeds a very large motor field about 10-40 amps depending on commanded DC motor speed, etc; the motor field is about 19H. Also, if it matters, the three-phase power system feeding the primary is floating, i.e. totally ungrounded.

My understanding is that the delta secondary could operate at a lower kVA rating going open delta and feeding all three-phase loads equally. My thought is a loose connection and series arc may have led to one of the wye primary connections briefly interrupting the current flow of one primary and then perhaps re-striking an arc with the aid of some measure of voltage spike. If the transition from three-phase delta to open delta and back is smooth enough maintaing its three-phase input, perhaps the digital exciter and all it's recorded sensing and alarm points might not offer any indication that a primary phase to the transformer was in an arcing meltdown mode.

I have two basic questions. First, links like the below suggest that a three-phase balanced load shouldn't lead to that much of a voltage spike across the opened primary connection. But, because the six-pulse SCR's are commutating constantly, could these extremely brief single-phase short circuits lead to higher voltage transient on an open primary connection than would be otherwise anticipated? The below paper does mention a single-phase short circuits o Pg. 4, Special Note.

http://www.cooperindustries.com/con...esources/library/235_SurgeArresters/89010.pdf

More generally, are there any good references that discuss the types of transformers that should feed drives and what the advantages and disadvantages might be?

Thanks for any and all help!

 

[waross]

Don't forget that the primary neutral is floating.
The first thing that happens with an open primary phase is that the windings of the two healthy phases become a series, single phase connection. They are now connected line to line instead of line to neutral so the voltage across the combination rises to 1.73 of line to neutral voltage.
If we assume equal loads, the voltage across each winding drops to 1.73/2 = 87% of normal voltage.
The secondary connection now becomes two windings in series together developing 1.73% of normal voltage. The third winding is now in phase with the other two windings and is being back fed with 1.73% of normal voltage. Saturation may limit the voltage back fed to the primary winding to a lower voltage than the full 1.73%
So the winding with the open circuit will be developing a voltage on the primary winding that will approach 1.73% of rated voltage.
The actual voltage is not, however, the biggest issue. This back fed voltage will now be 90 degrees out of phase with the line voltage.
If the arc is intermittent, each make-break cycle will involve the voltage dropping to 87% on two windings with an attendant 30 degree phase shift.
At the same time the voltage on the other winding will be approaching 1.73% of normal and will be phase shifting 90 degrees.
The phase shift will result in a voltage across the open connection that will probably be over 150% of normal line to neutral voltage.
At 1.73% over voltage, saturation is to be expected. This will lead to excess current and excess heat loss in all three secondary windings and at least two primary windings.
PS the primary neutral or wye point must be connected to the source neutral in order to use an open delta connection. Without the neutral connection the open delta becomes two transformers in series. That gets interesting in its own right if the loads are not equal.

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

 

[HenryOhm]

waross, thanks for the help. I had to spend some time digesting what you said. At first I thought that the voltage across the disconnected primary would go to zero based on Figure 5.33 of this link:

https://books.google.com/books?id=q...ngrounded wye delta transformer banks&f=false

The curve shows with the single-phase to three-phase load factor K dropping to zero, the open primary voltage to ground would go to zero. But, in the previous two pages, it describes how single-phase loading would change the equation. While the single-phase load shown is static, not dynamic like these 6-pulse SCR's, it shows in Figure 5.32 how the open primary goes to the 173% as you describe. I assume with the SCR exciter, the neutral point shown shifted entirely to point C would also have a period where it might try to shift the neutral all the way towards point A(?).

Then I had to sit down and look at how the delta secondary shown would feed a typical 6-pulse SCR drive like this:

http://www.gohz.com/Content/Images/uploaded/six-pulse-bridge-rectifier.jpg

With the switching of the three-phase into cyclical single-phase current pulses, it seemed to make sense that the single-phase loads would continue to feed the drive. I do not understand at this point how the 90deg phase shift you mention or other factors would probably skew this operation away from a true three-phase feed. But, hopefully I am right that the drive would keep firing "on all six cylinders" to some extent.

I'm not sure if capacitive coupling would play a role in the equation at all? The transformer in question does have a single electrostatic shield. Not sure if that would relate to either capacitive coupling or any sort of ferroresonance? I've discounted the latter because the transformer is low voltage (480V primaries L-L stepped down to about 208V L-L on the secondary) and there are no long cables involved. And, of course, there still is the possibility of repetitive restrikes with a loose connection rather than an open switch on the primary B phase.

Is the use of a wye-delta transformer commonly used to feed six-pulse SCR's?

 

[waross]

I may be wrong on the phase shift. I will think abut it. Too late and too tired to analyse it just now.

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