ATM Re: Power systems question

["Bell, Andrew" <AMB2@pge.com>]

Yesterday afternoon, Jim Miller provided a fairly accurate description of
how dependable the 60 Hz grid frequency is, and asked whether we had any
"power guys" on the list.  I can't claim to be one, but will confess to
responsibility for occasionally counting an electron or two.

The short answer to Jim's question is that U.S. grid operators do pay a
great deal of attention to making sure that neighboring power systems are
maintained in phase with each other, but this is done somewhat at the
expense of keeping everybody right at 60 Hz with every tick of the clock.
The power system term used for this is Area Control Error (or "ACE"), and
one of the units that is used to measure this is accumulated deviations from
60 Hz as would be measured by a clock driven at the grid frequency.  The
grid operators agree to make "time error corrections" as needed to bring the
accumulated deviations back to zero, but apparently only begin doing so once
the accumulated error has exceeded some threshold (which can amount to as
much as several seconds).  A standard recovery rate used during these
corrections seems to be 1.2 seconds per hour (e.g., by running 72 cycles per
hour "too fast" if they're trying to make up time, or the same amount "too
slow" if a positive clock error has accumulated).  This means that for some
purposes 60 cycles per second should really be looked at as "60.00 +/- 0.02
Hz."

These corrections are an everyday rather than only occasional occurrence.
For example, calendar year 2000 data reported by the Western Systems
Coordinating Council grid serving the entire Pacific Coast showed 427
separate instances (and 1459 total running hours) where they needed to
intentionally speed up the clock, together with 111 other instances and 341
total running hours to "slow down" the clocks.

While those of us with clocks plugged into a wall socket would probably
never notice the difference here, the time corrections would run a simple
"60 Hz" clock drive 18 arc-seconds too fast per hour of tracking time during
a catch-up cycle, and similarly too slow during a slow-down correction.  I
would have to assume that for professional applications there must be power
conditioning or other corrective steps taken, so that a real big bucks long
exposure would never rely on the wall circuit frequency alone.  For those
interested in the excruciating detail, here are a couple of links:   

National Electric Reliability Council Operating Standards:
http://www.nerc.com/~oc/rs.html

Western Systems Coordinating Council Operations:
http://www.wscc.com/committees/OC/CMOPS/PWG/Meetings
/PWG_March%2001Final%20%20Mtg%20Minutes.pdf

http://www.wscc.com/documents/standards/for_approval
/Procedure_For_Time_Error_Control.pdf

-- Andrew Bell
----------------------------------------------------------
From: "Jim Miller" <jim@jtmiller.com>
Subject: Re: ATM Fw: Posting Assistance

i remember my father once telling me that although the 60hz grid frequency
is maintained reasonably accurately over a few days there can be pretty
significant deviations over a day or so due to the need to supply more power
if needed.  the average is maintained by slowing when necessary to achieve
the desired average.

any power guys online?

jtm