Reply
Wed 5 Nov, 2003 07:52 am
The Northeast Blackout In August
Have authorities come up with a coherent explanation for the Northeast blackout in August?
The last explanations I heard were still speculative...
I have not seen or heard any definitive answer.
I suspect that improper switching procedures were used.
Also, in Cleveland, Ohio there were some massive failures an hour or
so before the big cascade of power outage. The combination of the
local failure and the badly followed power switching protocol,
adding to this a massive instantaneous demand, all contributed
to a cascade of the rest of the failing grid.
~The theory of the grid is my town is temporarily using more
electricity than it is producing, so the grid automatically makes
up the deficit with less than a second delay. Other towns are
producing a small surplus and my town pays for the electricity we
bought automatically. As long as the surpluses and shortfalls are
small, the grid performs reliably, but suppose a big generator is
supplying 101% of it's rated output. This can do millions of
dollars worth of damage if the excess gets bigger, so the
automatic cut outs do what they are supposed to do: they protect
the generator by disconnecting it from the grid. This is not
supposed to happen, but likely Homer Simpson was eating a donut
and did not notice the alarm telling him that the output just
passed 99%. If Homer had not been goofing off; he would have
reduced the output voltage slightly and remotely started a gas
turbine generator or two, which can be put on the grid in
seconds. As the gas turbines began to put power on the grid,
Homer would ease the output of the big generator back to the
proper voltage as he watched the big generator fall back to a
more comfortable 98%.
You can probably guess what happens on the grid when that one
gigawatt generator trips off line. It creates a massive negative
spike, which propagates at about 100,000 MPH throughout the grid.
Before Ohio hears Homer say "oops" thousands of miles of
transmission line are carrying more amps than they are designed
to carry even briefly, trying to replace the missing gigawatt in
the Springfield area. Overload breakers trip. If they have been
bypassed, the wire melts and a strand of high tension line drops
from a giant pole. Gas turbine generators and perhaps some big
plants pass 101% and trip off the grid. A few humans see what is
coming and cut their town off the grid, but they will probably be
fired if they guessed wrong and the grid does not fail as it did
in August 2003. Unfortunately there are 10,000 other scenarios
that can bring down the grid. It is a miracle it does not fail
more often.
The answer is not to practice brinkmanship. Generators should
rarely supply 98% of their rated output nor should lines be
pushed to their limits except in emergencies. Locals that are not
rigorous in providing a comfortable margin and avoiding goofs need
to be disconnected automatically from the grid at the first 1/2
cycle that is a few parts per million lower than the optimum
voltage. The local grids of those who are incompetent will fail,
but the main part of the grid will survive. Please embellish, refute or comment. Neil~
The problems, as I understand them, were, Poorly maintained transmission lines that were over heating due to weather. A computer monitoring system that was receiving insufficient data (some of them were apparently down at the time) An automatic response system that began to shut down generators when it sensed the system was out of phase. Part of the problem seems to have been (and I do not understand this) that there are two kinds of electric waves transmitted. One which carries power, the second which cost power to generate and transmit but is necessary for the proper functioning of the grid. Public utility generators produce both kinds but private utility generators tend not to as a profit margin consideration. There was a lack of the second kind of power on the grid as public utility generators began to shut down, which led to more shut downs.
Last year I realized that matching phase on the grid could be difficult. I found no confirmation and forgot until you mentioned this problem. Each transmission line acts as a delay line. Typically a millisecond or there abouts for a 50 mile long power line . If city a is tied to city b, one city can change the phase so they match at the point where city c ties into the line. then city c changes it's phase so the power is in phase at the triple junction point. I think you can see that as the gid starts looking like a web there will be no practical way to match some connection we would like to make. I presume a phase miss-match of 1/2 millisecond causes only minor power losses, but a 2 milliseconds or more miss-match can bring down the grid if other types of brinkmanship are being practiced. Can someone add details? Neil
Stumble It!
Tweet This
Bookmark on Delicious
Share on Facebook
Share on MySpace