Medium Voltage Closed Transition Interrupt Rating

[diakonos1984]

I've read through the forums here on previous discussions of closed transition transfer schemes and necessary AIC and interrupting ratings.

There were a couple things that no one brought up, but I would still highly value opinions on for my application.

I'm not sure what voltage levels the previous thread was covering, but I'm dealing with a 4160V Main-Tie-Main scheme. Traditionally these use a barely open fast transition (3 cycle) "odd-man-out" scheme for manual transfer, and a voltage supervised open transition for automatic (utility line failure) transfer. This does not provide my client with adequate reliability, so we are looking into alternative transfer schemes.

For either closed transition or a permanent parallel, there is quite a bit of redesigning of the system required, and I've read the Buff Book thoroughly on the subject. However, the AIC and interrupting ratings are still causing me pause. Clearly 2xSC ratings would be required for a permanent parallel.

But for a very brief (a few cycles) closed transition, the previous discussion somewhat hinged on the statements

But basically during that closed transition time your fault current may exceed the AIC ratings of your equipment. This is sometimes accepted as a "calculated risk".

and

If you are paralleling one live source with another live source, what are the odds of a bolted fault occurring during a 1-2sec transfer period?

What no one brought up, as far as I could find, is what if we simply used the fault detection capabilities of a modern multifunction microprocessor relay to prevent closed (or any kind of) transfer if there was a fault inside the system? This is not difficult with today's relays. The system should never transfer for a downstream overcurrent event, and to do so would only compound the event. The system only needs to transfer for an upstream fault (often miles away on the utility grid), and should block any transfers on a downstream fault.

I understand utilities don't like you paralleling their feeds and backfeeding into a fault, but unless I am mistaken (the likelihood of which is why I'm posting here) fault current backfed to the grid will not approach the SC available inside the system. I base this belief on the assumption that impedances on the grid are much higher than inside the system, and that faults on the grid have to be detected by reverse current/power relays because the normal overcurrent relays (set lower than the available SC current) will not usually pickup for reverse current!

So... there's my thought process. I know there are other factors, but speaking only to the AIC and interrupting ratings, am I missing anything?

Thanks!

 

[ron]

The issue is whether the equipment would be properly rated if a fault occurred during the parallel condition, whether 100ms or longer. It is practically impossible for a fault to occur coincidently during the 100ms that you are in parallel, but it is theoretically possible.

You've seen the code proposal (in the link you provided) that asked for that exemption and it was rejected.

 

[ron]

From the committee responses in the previous code proposals that were rejected, it would appear that you could put your license on the line and appeal to the better judgment of the AHJ, but I will keep my seal safely in my drawer and not do that.

 

[kingpb]

A simple sync check relay will keep the tie from closing onto a faulted bus.

Beckwith makes the best fast transfer device. Check this link for a little info.

The Beckwitch is the only thing we specify for power plants.

http://www.beckwithelectric.com/docs/app-notes/appn-19.pdf

 

[rbalex]

The problem isn't transfering into a fault. Until the literal text of the first sentence in Section 110.9 is revised or an Exception is recognized, no amount of relaying will make a Code compliant installation.

Otherwise, the only option is a 90.4 "Special Permission." Since the vast majority of AHJs or their representives aren't qualified to evaluate it or confident enough in your ability to, I wouldn't count on that option either.

 

[rbalex]

You might also want to review this thread.

 

[diakonos1984]

All right, I see the little fine print issue here!

no amount of relaying will make a Code compliant installation.

Thanks everyone!

@kingpb... I was looking into the Beckwith device already... it sounds like your experience has been good? Have you gotten any feedback in the years after the devices have been installed? Since I'm a field service and consulting engineer with a small and regular clientele, if it has one problem in 10 years, I'm still going to be the one dealing with it. :blink:

 

[kingpb]

All right, I see the little fine print issue here!

Thanks everyone!

@kingpb... I was looking into the Beckwith device already... it sounds like your experience has been good? Have you gotten any feedback in the years after the devices have been installed? Since I'm a field service and consulting engineer with a small and regular clientele, if it has one problem in 10 years, I'm still going to be the one dealing with it. :blink:

It works as stated, no issues, and in many cases it is the only thing the Owner's will accept. The Beckwith fst transfer equipment is top notch.

 

[rbalex]

With no disrespect at all to Beckwith, I would point out a few things:

1. The primary application Beckwith is describing is for utility power plants. (Section 3.0 introductory statement) Depending on the set-up they are probably exempt from the NEC anyway.

2. For the transfers you described in the OP, high-speed synchronizing isn?t all that beneficial either.

a. For a voltage supervised transfer, you are already transferring from a ?dead? bus to start with. Although you are legitimately concerned about possibly closing the healthy bus into a fault, there are plenty of ways to detect and block it without a high-speed synch-check relay. (If you look closely at your current scheme, you probably already have it) Besides, you probably want to delay the transfer anyway to allow back-EMF decay from motors on the dead bus.

b. For either a manual maintenance transfer or ?return-to-normal? after a voltage supervised transfer, you already assume the sources are synchronized for a closed transition.
​

In any case, there is no current NEC provision (other than 90.4) for a closed transition of any kind that doesn?t require full interrupting duty of the Mains, the Tie and all downstream OCPs that see an increased available fault caused by paralleling the sources.

BTW the best way to handle 2a above is banked automatic restarts.