does anyone have authoritative knowledge as to how long it would generally take to safely shut down a typical nuclear power plant in the U.S.A.? A recent article claimed that such shutdowns can require four months. Any truth at all to this?

-- michael dorfman (dorfmanki@home.com), January 20, 1999

Answers

Latest on csy2k is 5 months.

-- Andy (2000EOD@prodigy.net), January 20, 1999.

I think it depends on what "shut down" means. I have it on good authority that it takes 6 months to a year to "de-commision" a nuke, i.e., shut it down for good. This same source said they could take it off-line and put it safely in an "idle mode" within 24 hours.

Maybe "shut down", in the context of Y2k uncertainties, means placing the reactor in such a state that, if it were never used again, no long-term physical or environmental damage could occur. In other words, sort of a "temporary" de-commision.

-- Nathan (nospam@all.com), January 20, 1999.

Latest on csy2k is 5 months. For core cooling with guaranteed (ha!) juice to do it...

-- Andy (2000EOD@prodigy.net), January 20, 1999.

You can SCRAM a plant like right now. BUT should you want to shutdown, cooldown, depressurize, well you might take a couple years plus or minus depending on how the plant was operated. How old is it and how many Fission products does it have?

-- nine (nine_fingers@hotmail.com), January 20, 1999.

Technically what ought to worry most is keeping the core cool after you've terminated all fission chain reactions, which you can do in milliseconds by dropping all the control rods.

Trouble is, this doesn't stop radioactive decay of unstable isotopes formed during the previous operation of the reactor. This means the fuel rods are still giving out heat. The amout of heat falls as the isotopes decay.

For some amount of time, you have to be able to pump cooling water around the core. Someone might like to find out whether they can do that long enough without any electricity supplies from offsite? Without any oil supplies from offsite? If the answer to the latter is no, they should damn well stockpile more oil (diesel?) now, while it's cheap and plentiful and transport arrangements aren't a problem.

What this time is will depend, amongst other things, on how long the fuel rods have been in the reactor for.

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), January 20, 1999.

I've been wondering about this myself. I have been reading on threads 4 months to safely cool the core and 5 months to cool the fuel rods. And that "safely cool" means there must be a (long-term) alternative source of energy available. What gives? In the event of energy rationing, I would expect nuclear plants (like hospitals and 911 forces) to be given priority. I'm more concerned about the possibility for the plants to shut down automatically (and simply not be available to us) at the slightest embedded chip problem.

-- Brooks (brooksbie@hotmail.com), January 20, 1999.

My husband was a quality control inspector for General Electric. I just picked his brain: An emergency shut down is called a SCRAM. It means that each rod, powered by a 15 horsepower motor, is jammed into the core. All stop. That process keeps the reactor from getting out of control, but it also requires months and months to repair, as they have to remove all the rods. If the company needs to shut a reactor down voluntarily, it takes a few weeks to put everything on hold. If the company needs to shut the reactor dowm permanently, it takes about 5 months.

Incidentally, the reactors in this country are very safe. They have many redundant systems built into them, for safety. Whereas, the Soviet reactors eliminated many safety measures, and that is why they had Chernobyl.

Hope this helps.

Mary

-- Mary Phillips (blufrogg@garlic.com), January 20, 1999.

Nope - shutdown by scram = rapid insertion of rods (by springs, obviously assisted by gravity "down" into the core, no power needed. If power lost (even if millisecond) locking springs "pull out of the way, rods are forced all the way into the core.) -

Motors "pull" them up against spring pressure when restarting. No power, no motor, no restart obviously. Lots of other pre-requisites and checks before restarting. Scramming reactor is simple and "routine" event, not a catastrophy, and the reactor can be restarted without "repairs" or other delays. I can't tell why you're getting poor info from a source supposedly from GE, but that a scram always requires "repairs" info is not right. (You are required to fix whatever caused the shutdown before restarting, else get BIG fines, criminal penalties from NRC. Maybe that is what the story is talking about.)

Time to shutdown and cooldown to room temperature (routine maintenance or fuel reload) is 2-3 days. Thermal load after shutdown depends on time - and is expontially decreasing.

Minutes after shutdown = about 2.0 - 3.0 percent of full previous power.

One day after shutdown = about 0.5 percent.

One week after shutdown = about 0.02 percent of previous power, then more slowly declining. Realistically, use the 0.02 percent as a constant. (0.0002 times max power.)

Before startup can begin, and while the reactor is operating, there must be at least two independent backup power sources for cooling - these two sources must be independent of site or off-site power. Normal backup are paired (sometimes triple) diesel gen sets at the site, each driving a different power and control "train" supplying different pumps on site. To refill a "fuel pool" with cooling water - if the level ever gets low, there are up to six other sources of clean water on site. Various alternative ways of getting cooling water into the pool - up to and including fire pumps.

Note - no fossil power plant has any mandatory backups, nor are any fossil plants legislatively required to have any on-site reserve power to allow a "black" startup. Some do, but they are the exception.

If the grid goes down, it could be very, very difficult to restart a large fossil plant from zero - include train unloading stations, barge cranes, coal belts, and coal crushers through exhaust emission and filters.

Hope this answers your questions. Do not accept the "four months" time frame for shutdown. It is often repeated, but is equally wrong.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), January 20, 1999.

There's an excellent discussion that verifies Robert's explanation at:

http://www.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=000C f9

-- Dan Webster (cant@spamme.com), January 20, 1999.