[Lowfer] Re: Power Outages (and frequeny control)

[Clint Turner]

In a previous life, I worked for a company that 
designed/installed/maintained SCADA systems (Supervisory Control And 
Data Acquisition.)

Back then - and now - the Western Grid, WAPA (Western Area Power 
Administration - a government-related organization) was "in charge" of 
maintaining frequency.

How this is done is rather tricky - but then it was largely with larger 
hydroelectric plants, etc.  Back then, I believe that there was some 
sort of schedule as well as updates pertaining to frequency control to 
make sure that the long-term average maintained 60.0 Hz.  (Some of these 
things are alluded to at this page:  
http://www.wecc.biz/documents/NERC/IOSS%205-7-03%20Notes.pdf  - which 
doesn't directly cover the topic, but refers to some of the effects on 
frequency vs. load. )

Because the loads are generally very predictable - so is the line 
frequency (which typically varies about 0.05 Hz over the course of the 
day - higher at night when there is less load, and lower in peak load 
times) it was mostly a matter of a few phone calls to "nudge" the 
frequency in the right direction for longer-term corrections.

Nowadays, I'm not sure what is done.  I suspect that GPS timing is used 
- but not so much for power frequency control as to give utilities a 
common reference:  The powerline frequency almost qualifies for that 
"unstoppable object" that was mentioned in physics lectures - I really 
doubt that the actual means of management have changed all that much.

Suffice it to say that a generator puts power into the grid by "pushing" 
it - that is, advancing the phase just slightly, so the "outlying" areas 
can be slightly advanced in phase as compared to the major pools - which 
can cause problems if you have multiple interties at various points.  
This also occurs between pools - and can cause problems if not very 
carefully managed, as you can imagine.

Generally, two things were done about this:

- Large phase-shift networks exist.  I believe that one of the largest 
ones is in the 4-corners region (Utah, Colorado, Arizona, New Mexico) 
where these differentials may be "tuned out" between segments of the 
power pools.  More interties have been constructed in more recent years 
to allow more tight coupling of the Eastern and Western pools.
- The use of DC powerlines.

The last one is of particular interest.  My former boss had told me 
about a problem that existed between the Pacific Midwest pool (Oregon, 
Washington - lots of hyrdo) and Southern California.  Apparently, in the 
60's, this pools would occasionally destabilize, with the northern and 
southern pools going into some sort of slow oscillation, with several 
gigawatts "sloshing" back and forth, driving the system crazy.  One of 
the solutions was to construct a DC powerline with the Pacific Northwest 
feeding DC down to southern Cal, ending up in, as I recall, Sylmar in a 
coliseum-sized building with extremely large mercury vapor switches.  
Since the DC is reconverted to AC at the far end, any phasing/frequency 
problems are irrelevant.  Also, this saved a fair amount of energy - 
especially considering that about 18-20% is lost in distribution.  
(Don't forget that, at 60 Hz, the skin effect penetration depth is about 
6mm or so...)

DC is used several other places - between the IPP power plant near 
Delta, Utah (+- 500 KV) which goes to S. Cal. and (I believe) from the
Nuclear Power Plant in Arizona to Phoenix - and other places where it 
make sense to do so.  (It's interesting to drive near/under those lines 
while watching a compass...)

***

Shortly after the Northridge California earthquake (early 90's) I was 
visiting a local (Utah) municipal power plant and one of the operators 
copied for me a strip chart recording of the frequency versus time of 
that event.  At the moment of the earthquake several GW were shed - and 
the line frequency - which had been wobbling at about 60.05 Hz suddenly 
shot up to nearly 61.5 Hz.  It then swung down to about 59.0 Hz or so 
and continued oscillating wildly like a bell (but with decreasing 
amplitude) for about 8 hours until it finally stabilized at about 59.8 
Hz - and then, over the next several days, it slowly drifted up to 60.0 
Hz as the loads got reattached and (presumably) the WAPA did its work to 
make sure that the "long term average" maintained 60.0 Hz.  (Note that 
the long-term average had been skewed by the brief excursion to >61 Hz...)

In talking to the operators at the power plant (who had spent plenty of 
time on the phone at the time of the event) what apparently happened was 
that when several gigawatts dropped off, the source of power (which was 
not necessarily located where the load was shed) had to go somewhere - 
and the frequency shot way up.  When the frequency shot up to about 61.5 
Hz, a bunch of other power generating facilities that would not have 
otherwise been affected tripped out on overfrequency - as well as a few 
other utilities - which caused blackouts in areas otherwise unaffected.

The loss of generating capacity (due to high frequency tripout) then 
caused the frequency to drop again to about 59.0 Hz - causing some 
existing systems to become overloaded and trip off - causing even more 
scattered outage - which reduced the load again, causing the frequency 
to go over 60 Hz yet again.  Fortunately, at about this point, the human 
operators were getting clued in, taking manual control, isolating 
segments, and "massaging" load management - and the amplitude of the 
"ringing bell" greatly diminished.

***

One final "frequency control" anecdote:  Once, while I lived in 
Flagstaff, AZ (late 1986, early '87.)  I was on 75 meters with a bunch 
of other hams across the western U.S.  I noticed then that the 
incandescent lights were slowly pulsing - about once every 5-10 
seconds.  Grabbing a voltmeter, I could see the line voltage swing up 
and down about about 5 volts.  After several minutes of this, I 
mentioned it on the air and another ham in Salt Lake (about 385 miles or 
625 km away)  mentioned that he, too, was watching his plate voltage on 
his amplifier swing up and down a - by this time - hundred volts or so - 
and it was precisely in sync with what I was seeing.

This continued for a few minutes more - and the "oscillation" frequency 
gradually increased to nearly 2 Hz - when suddenly it stopped - without 
the blackout we were expecting...  I Never saw anything in the news 
about it - but I wondered who was "motoring" what and where?

Clint
KA7OEI

P.S. For an interesting list of "Oopses" go to 
http://www.nerc.com/~dawg/database/dawg-97.html  - and related pages...