Regenerated voltages

[killerwatt]

There have been a series of "on-site" loss of phase situations at a large nursing home in Vermont. The nursing home operates on the load side of a utility owned primary meter. The configuration consists of a 3 phase system (12.5KV) that radially branchs off the main utility line, pass thru a primary meter to a transfer switch/generator arrangement. The transfer switch will operate upon loss of utility power.
Their load consists of several single and 3 phase transformers that feed various facilities on their campus.
All of these loads are individually fused with overhead line cutouts.

The Problem: When a phase is lost (1 fuse opens) on the campus, the transfer control senses loss of phase and transfers to generator power. The generator will start but will quickly transfer back to utility almost immediately, causing an oscillating effect between sensing a loss of power and incorrectly viewing "regenerative voltage" as a restoration of utility power (a phantom voltage?).

The entire system, both utility and customer are wired grd-wye/grd-wye, 12.5KV.

Generator reps and transfer switch techs have been on-site and upgraded the t-switch control in an attempt to more properly sense and correctly react to the available voltage under normal and single phasing conditions. After recent changes, the problem still exists.

As a utility engineer that has worked extensively on grounded wye systems, I have not experienced regenerated voltage upon loss of one phase. I have, tho, seen a wye delta connection act as a grounding bank upon loss of 1 phase IF the high side was tied to ground and not floated. This is not the case here.

Any ideas are appreciated. Thanks.





cluster of transformer locations that are all individually protected by fuses.

 

[davidbeach]

Look closely at the transformers. They may actually be YDY with the tertiary delta not brought out; lets the manufacturer build a cheaper transformer on a three legged core rather than building a two winding transformer on a 4 or 5 legged core.

On the other hand, the control system should be such that once the decision is made to go to generator, a return to utility has to wait for at least 30 minutes after restoration of utility power except for loss of the generator.

 

[killerwatt]

Thanks davidbeach. Although I have not personally inspected the distr. transformers, I highly doubt that any are YDY connected.

As for delay time, yes..the control allows a certain period of time during transfer. I think it is around 20-30 minutes.

 

[waross]

I agree with davidbeache's
comments, particularly the mandatory run time.
However, at the risk of offending you, has someone connected the sensing inputs to the wrong side of the transfer switch?
This is so obvious that it may have been assumed to be correctly connected and not even checked but it fits the symptoms.
If that's okay, then;
Is the voltage detection circuit a true phase monitor or does it just detect loss of voltage on one phase. It may be responding to an induced voltage on the open primary line if it is not checking phase angles as well as voltage.
Is the metering transformer connected delta?
Trouble shooting. I would go to generator power manually.
Now pull one fuse on the primary, and check the incoming voltages to the control circuit. If you are seeing a voltage on the "dead" phase, find a multi trace scope and check the phase relationships. If the "Phantom voltage is induced, it will most likely be in phase or close to in phase with one of the good phases. Upgrade your sensing circuit. You may be able to add a good quality phase monitor to replace the existing circuit.
If the "phantom" voltage is in the proper phase relation ship look for a delta transformer bank somewhere. It may be metering transformers, it may be you sensing transformers, it may be an industrial plant.
Don't forget to pull all three incoming fuses to double check for connections on the wrong side of the switch.
respectfully

 

[itsmoked]

Boy small world.. I am working on this exact same problem right now. Car hits power pole. Phase lost. Generators, 50kW and 200kW, both start. Transfer switches never transfer.... Dooh!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[alehman]

Also see

http://www.basler.com/php/download.php?file=practical.pdf&id=204&type=res

 

[davidbeach]

Oil filled or dry type transformers? If dry, and nobody insisted on two winding, 4 or 5 legged cores, the odds are very high that you have YDYs. If oil filled, I don't know. The ABB transformer factory that used to be the Westinghouse transformer factory has been making those for at least 15 years. One nasty mess about that long ago at a hospital in Portland, OR, when the delta windings kept a primary transfer scheme from switching from the preferred to the alternate feeder.

 

[LWS]

Very interesting problem. With regard to Waross's comments that the voltage sensing transformers may be on the wrong side, he may be right track on that and it should be checked out. You say you have 20 - 30 minutes for retransfer (to allow utility to stabilize). Why isn't it working? Why is the ATS retransferring? If the PTs are on the load side (erroneously), then I would think they that you would get a consistent retransfer back to utility in whatever the defined retransfer timer is set to even if there were no downstream fuse blowing.

You probably have two sets of PTs - one on the line input, one on the generator input. Reason being, you don't want to transfer if generator fails to start (depends on how control power is derived to operate the ATS control circuitry). So, once generator is up to speed and frequency, THEN it will transfer to generator power.

You mentioned that the customer distribution transformers were connected grd wye-wye. Based on facts presented, the fact that you haven't physically inspected them, and my experience, I would say there are 3-phase banks connected delta-delta (what's your secondary service voltage to the nursing home facilities?) and you have two things going on:

1. Possible ferroresonance when the single-phase fuse blows. Depending upon the length of the circuit from the cutout to the transformer, interruption of one of the phases will result in a ferroresonant condition due to the RLC circuit created by this. This will cause severe overvoltages on the remaining two phases that can blow surge arresters or, in this case, wreak havoc on sensing transformers. There's an IEEE Standard (I don't have it at my fingertips) that addresses ferroresonance as a function of various transformer bank connections. This may cause the ATS sensing circuit to believe the generator is unstable and initiate retransfer.

2. Couple that with sensing PTs possibly connected incorrectly, that could exacerbate the situation.

The fact that this occurs only after a transformer cutout blows leads me to strong suspect the ferroresonance theory. Do you have overhead surge arresters? If so, what voltage/class? Any history of arresters blowing? Have they even been inspected to see if they're blown (easy to tell - usually ground-side jumper is blown out of the arrester).

Interested in the outcome of this unusual problem.

LWS

 

[waross]

Hello killerwatt;
Some of the transfer switches that I install have a "fast retransfer" feature. If utility current returns within a few seconds the switch will bypass the normal timing functions and reconnect to the mains immediately. That may explain some of the symptoms.
If the problem is a delta transformer connection as davidbeach suggests, and it may well be, the transfer switch will transfer back almost immediately. If it is a problem with the sensing connections, the switch will not transfer until it sees power from the generator. Also, if the connections are on the wrong side of the switch, the transfer switch would have to be connected to the utility manually if the generator was not running.
The more I think about it, the more I support davidbeach's suggestion. There may also be a nearby facility using a delta power system. If a fuse opened in a line serving both your facility and a small delta power bank, you may have similar symptoms.
respectfully