Wacky Weekend: NERC to relax power grid frequency strictures

Alex Rubenstein alex at corp.nac.net
Sun Jun 26 14:28:28 UTC 2011

> I think we're missing something, which is where these ATS's are
> installed.
> I don't think most utilities allow (largeish) ATS's to do a closed
> transition from a genset to the utility grid, but I may be wrong.
> There may be other ATS's in your facility that do a closed transition
> though.  For instance, consider this (somewhat simplified) dual UPS
> design:
>    Utility  Generator
>       |   \/   |
>       |   /\   |
>    ATS #1a  ATS #1b
>       |        |
>    UPS #1    UPS #2
>       |        |
>        \      /
>         \    /
>         ATS #2
>           |
>         Load

IMHO, this design is overly complex, and will lead to the most usually form of failure - human. There is lot to think about in the above, especially during maintenance.

You also have an interested situation on retransfer (gen -> normal) when 1a and 1b do not transition at precisely the same exact moment, which almost never happen. And if one of 1a or 1b gets 'stuck', you have UPS's paralleled being fed from unsynchronized sources, which leads to other problems (such as bypass not being available, etc).

> ATS's 1a, 1b, sense utility power for quality.  Should the utility
> power quality not meet specs (e.g. go out), they disconnect from
> utility, tell the generator to spin up, wait 5-15 seconds for the
> generator(s) to spin up and then close to the generator.  They are in
> an open state for perhaps 20+ seconds, generators are never closed to
> the utility.  Going back the drop may be shorter, perhaps
> 10 seconds, but there's still a long-ish open gap.   Definately not
> sub-second.

That is not typical. The normal contactor/breaker doesn't usually open until the emergency source is available. It remains closed to the dead normal until emergency is ready to be transferred to. 

> ATS #2 takes the dual UPS output (from synchronized UPS's) and does a
> closed transition between the two sources.  Indeed, a previous employer
> had ATS's at this location that could switch between sources in less
> than 1/4 wave, the equipment never knew the differenece.  Very
> impressive.

So are you saying that the load was directly connected to utility (no UPS protection) until utility had a problem? 

> It's not that you couldn't install a closed transition ATS in the ATS
> 1a/1b location from an electrical point of view, but I don't think
> codes, power companies, or common sense make it a good idea.
> As others have pointed out, the grid can do weird things because your
> neighbors did something stupid, or a car hit a power pole and shorted 3
> phases together.  Syncing to it is, well, crazy.

Closed transition should not really be thought of as syncing to the grid. Closed transition infers a very, very short overlap. A few cycles. Mainly so that downstream load does not get interrupted on the retransfer.

More information about the NANOG mailing list