[142439] in North American Network Operators' Group
RE: Wacky Weekend: NERC to relax power grid frequency strictures
daemon@ATHENA.MIT.EDU (Alex Rubenstein)
Sun Jun 26 00:24:09 2011
From: Alex Rubenstein <alex@corp.nac.net>
To: Michael DeMan <nanog@deman.com>, NANOG list <nanog@nanog.org>
Date: Sun, 26 Jun 2011 00:23:47 -0400
In-Reply-To: <4F3FF32B-AC21-4505-91E1-3EEA21374994@deman.com>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
> It ismy understanding also that most commercial grade gensets have
> built into the ATS logic that when utility power comesback online, that
> the transfer back to utility power is coordinated with the ATS driving
> the generator until both frequency and phases are within a user
> specified range?
Well, that depends.
If you have a open-transition ATS, where there is a 'neutral' (read: not co=
nnected to any source) position, it doesn't matter (much). Well, it matters=
a little. There is really two types of open transition.
Something called "open transition" will provide a transfer by going closed-=
open-closed (in both directions). The issue is the open portion of that tra=
nsfer can be very short; sometimes only a few cycles at 60 hz. If you have =
an electric motor connected as load (fan, compressor, whatever), if the sou=
rces are out of phase, it can be an interesting event for said motor. Typic=
ally, a open transition switch will wait until the phasing is 'close enough=
' (usually programmable by way of degrees.). We have an old russ electric a=
ts somewhere that is happy at about 15 degrees +/-.
There is also a type, "delayed transition", which is closed-open-wait-close=
d. Wait is typically programmable, it may be 500 msec, it might be a minute=
. It's up to the user. This is regarded as the safest type of switch (imho)=
because you do not run the risk of any of the above mentioned badness. How=
ever, in a datacenter scenario, you do have a battery hit (ranging from ten=
s or hundreds of millisecond to many seconds depending on what you want). H=
ow good is your UPS and battery plant? Will your fans inertia keep air movi=
ng for a little while? All things to consider.
If you have a closed-transition switch, typically the retransfer from emerg=
ency to normal is closed-closed, meaning that emergency gen, normal utility=
, and load are all connected together for a short while. Typically in the t=
ens or hundreds of msecs. Anything longer than that kinda falls into the co=
generation category. That is another discussion.
At least here in JCPL territory (northern NJ), closed transition is frowned=
upon. Too much risk, they think. They are correct, really, but the risk is=
mostly yours. If you lock to the utility out-of-phase, you will surely los=
e and they will surely win. The fault you create that they will see will pr=
obably not hurt them. Unless it is extraordinarily large and you are very c=
lose to the nearest substation. You must really trust your utility and your=
transfer gear and your generators to do this. Personally, I'm not a huge f=
an of this, but that is just religion.
Personally, I like delayed transition, and that is what we do on anything r=
ecent. Short, usually, like 3 or 5 seconds. If anyone wants a demonstration=
, let me know. Long enough for motor controls to say "oh, hey, we lost powe=
r so let's do a nice soft restart of motors" and compressor controls can do=
delayed restarts as well. Works quite well, in practice.
Much is overlooked in this discussion, as to things people should do about =
ATS and UPS programming.. but it is outside of the scope of NANOG unfortuna=
tely. Perhaps we need a NADCOG or something.=20
What does this have to do with the whole 60 hz discussion and clocks? Not m=
uch. Other than I will have to rely on the cell phone more and the microwav=
e less for time.
