How does the power grid work, and why is it so vulnerable

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Can anyone explain how the power grid works (in layman's terms)? I have been searching the internet for a comprehensive discussion about the grid, but have only found extremely technical information. I have a basic understanding of electricity/electronics (eg ohm's law). If you can provide links, that would be appreciated. Also, what makes it so vulnerable to the y2k problem? Information that I have read suggests that if a substantial percentage of the grid fails, the entire grid could fail. Some posts suggest that it could take months to "restart" if it fails. Thanks in advance for any insights. Also, I commend most of you for providing useful information about other issues related to y2k--most of you seem very sincere and knowledgeable!!!

-- Don Chen (DChen@newbie.xxxcom), February 16, 1999

Answers

From euy2k:

"The Grid" is really a misnomer. It consists of 180 +/- Control Areas which already manage their local generation and transmission needs. These Control Areas are grouped into NERC Regions which manage the security of a larger area.

The US, Canada and part of Mexico are divided up into 5 major Grid regions. This is somewhat arbitrary in that it refers to power sharing arrangements in a given region. It is possible to import/export power across the regional Grids via a DC/DC bridge. There are intrinsic limits on both intra and inter grid power shifting for various reasons including line capacity. On the East Coast, there are areas which are already net importers of power under "normal" conditions. The implication is that relatively small degradations can cause rolling blackouts. South Florida is an example of such a region. Its my opinion that power problems will occur largely because of failures in distribution as opposed to generation. Power networks are heavily dependent on reliable telecomm networks. If the telecomm goes down for long, so does the power.



-- RD. ->H (drherr@erols.com), February 16, 1999.

Yes, euy2k is a good site for info on "electric utilities and the year 2000" and is is run by a GI. <:)= euy2k.com - The Newsroom


-- Sysman (y2kboard@yahoo.com), February 16, 1999.

Don,

I have two links for you. The first is CBN's interview with Rick Cowles, who is very knowledgeable in the field...

http://www.cbn.org/y2k/cowles.htm

...and here's an article I saw just today about Philadelphia's electric power utility. They are replacing software and hardware, but as of the end of 1998, the company had only spent $21 million of the $75 million they plan to spend on Y2K...

http://www.phillynews.com/inquirer/99/Feb/16/front_page/ELEC16.htm

-- Kevin (mixesmusic@worldnet.att.net), February 17, 1999.


The best explanations I've seen are Dick Mills' "powerful prognostications" on the Westerguard 2000 site.

You may think his conclusions are optimistic -- read euy2k as well -- but the explanations of how it works, and the debunking of some oft-repeated falsehoods about the same are masterful.

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), February 17, 1999.


To put the issue into a nutshell the power industry is a big user of embedded systems. Embeddeds have become a very big issue for Y2K due to being overlooked as probable trouble points until very recently (the last year or so). Previously the Y2K emphasis had been on mainframe computer systems and their programs.

To tie this to how Y2K failures could bring down the grid, look at the recent San Francisco blackout, one failure dominoed until there were more than two dozen sub-stations that were knocked-out and two generating units were shut-down as well. And it was because one sub- station had a failure.

If this was a Y2K failure, besides having to have repair crew reset the equipment at the substations and to re-start the generators, there would be the need to find the failed device and replace it, if spares are available. To scale this event up to a Y2K-size event, think of past events.

A similar outage was caused over the entire western US and parts of Canada by just one power line being brought down by a tree. There's the Great Northeast Blackout of the sixties to consider. It was a similar single failure that knocked-out power to several states. I went through a statewide blackout caused by a transformer failure at a power plant which resulted in five generating units at two power plants being tripped off-line. Those five generators were one-third of the power company's capacity and they were lost because of one single failure that cascaded. Imagine just one or two such simultaneous Y2K failures within each region of the national grid and you can see why there are serious concerns about possible regional or nationwide power outages.

Y2K poses peculiar failure opportunities because so many possible combinations of smaller failures that can add up might occur. I"ve got family members who work for the power industry who admit that inside their company "We really don't know what's going to happen." is the "official, not for the public" outlook.

As far as how long power could be out due to Y2K failures, it depends on what kinds of secondary damage any failures might cause. If there is no damage to associated equipment, then some portions of the grid can be restored fairly quickly. But there would still be problems with the actual equipment that failed and needed repairs. If any grid failures cause damage to generating plants or to the transmission and distribution systems, then getting the power back on becomes a more prolonged issue.

If there is serious damage involved, things could take a long time. Some items like replacement tranformers for large substations have six month lead-times to get one from the manufacturers. And that's in good times, in a Y2K timeframe with disruptions and shortages it will doubtless take longer to get some replacement parts.

I hope this provides a starting point for you on this subject.

WW

-- Wildweasel (vtmldm@epix.net), February 17, 1999.



Thank you Sir Weasel,

May I add that the telecom and satellite controls must be used to control the whole thing - because of the instanteous nature of electricity. You can't "store it" and unlike pipes and even a chemical refinery, you can't "monitor" and "control" by watching flow (amp) or pressure (voltage) slowly go up or down, reading dial and "tuning" the input generators and turbines manually. reaction times must be near instaneous - and over a wide area - to control the whole thing.

Manually, certainly you can go into the field (or call over radios) to local substations to individually trip breakers and reset panels. If no damage (breakdowns) has occurred - else the same part thta failed in area will fail again (or simultaneously) in a third, fourth, or fifth, or fiftieth, substation. Then that substation is back on line - until the next fault. then, will it trip automatically, or will you have to call each station and manually take each back off the line again. Then try to restart again.

The immense size of the grid (as-is/as-now) serves as a natural "load" or balance to the biggest power stations - so isolating each area, or even each customer utility, will tend to increase the efect of voltage and frequency fluctuations and variations as big loads come on and off the generator.

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), February 17, 1999.


Moderation questions? read the FAQ