[time-nuts] Yukon Energy causes time sync problems
bill at iaxs.net
Thu Apr 8 22:48:03 UTC 2010
There are some interesting misconceptions here.
Yukon Power did not cause time to slow down. They did what every generating
station does, which is to adjust drive power to make a synchronous power
clock match a precision reference clock. The tolerance is seconds because
means for controlling drive power has a time constant of several seconds.
The real problem is the way demand varies. If you pour mechanical power into
a generator, it will speed up when lightly loaded or slow down when heavily
loaded. This is less of a problem when many generators are tied together by
a power grid, as they are all synchronous machines. Central dispatching
stations compare line and reference clocks, and direct plants with capacity
to do so to make up lost cycles, or buy less from the most expensive sources
when extra cycles are generated.
If you have a 5 digit counter (or more) tied to a computer, you can plot the
deviation of line frequency for 24 hour intervals. TVB had this on his site.
What I saw in MN was that generators speeded up in the early morning to make
up cycles so there was no reference error at 6 AM. Then the loads turned on
and the cycles fell behind and recovered as power was dispatched, within +/-
6 seconds. This is good enough for social time, where the mundanes don't
The Alaskan network is probably too sparse for central direction, so each
power plant makes its own adjustments. Note that this doesn't necessarily
produce stable control, ever.
In this case, the reference clock appears to refer to GPS satellite time,
uses a standard wall clock to display it. It is the reference clock that
slowed down when it should have failed to work at all. Perhaps the wall
(maybe it was really a HP 113) needed oil. There's the real question for
nuts: How did the reference clock slow down?
The first comment to the article shows what happens when your ego fails to
shame you into silence when you don't know what you're talking about:
"I don't understand how the amount being generated has anything to do with
what happens to household electronics." [see above]
"It would make more sense if the plant was generating at 55Hz versus 60Hz
as some electronics will use the line frequency rather than integrated
oscillators to set clocks." [The plant probably has breakers that take it
off line when the frequency gets below 58 cycles, to keep it from dragging
the network down.] [I thought that all electronics today converted the line
to DC without sampling it, and ran timing from a crystal. Anyone know?]
"Regardless, it shows YEC continues to be a bunch of bumbling oafs." [People
who live in glass houses shouldn't stow thrones, or something like that.]
Thanks, Brooke. I had some fun explaining all this.
From: Brooke Clarke
Sent: Thursday, April 08, 2010 12:58 PM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Yukon Energy causes time sync problems
More information about the time-nuts