Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] Nuclear Power and the Grid



On 09/09/2011 03:38 PM, Edmiston, Mike wrote:
I strongly agree that all plants should have black-start capability,
and I have been saying that for years.

:-)

I also know that nuclear plants tend to SCRAM when there is trouble
on the grid, and I wonder if someone can explain that to me.

Disclaimer: I am not an expert in this area. Really really not.
I generally don't approve of ill-informed speculation, but I will
try to say a few things that might be in the category of hypotheses
worth considering, worth looking into.

I reckon there are two problems:

1a) Electricity is exceedingly perishable. If you think lettuce
is perishable, or theater tickets are perishable, that's nothing
compared to electricity on the grid. There are very few inductors,
capacitors, flywheels, or anything else in the system that can
store appreciable amounts of energy. There are a few pumped
storage hydropower systems, but they are small, few, and far between.

That means that if your power plant becomes disconnected from the
grid, it will very soon overspeed and destroy itself unless you
take action to throttle back the prime mover.

1b) Throttling back a gas turbine takes only a few seconds. Throttling
back a coal-fired boiler takes somewhat longer, but then again there
is some energy-storage capacity in the boiler itself.

In contrast, the physics of heat generation in a nuclear reactor is
very complicated and very unfriendly. You really don't want it to
be "prompt critical" because that's what makes a nuclear bomb go
boom. So you are relying on /delayed/ radioactive decays. And
they are delayed quite a bit. After you SCRAM a reactor, it continues
to put out a large percentage of its rated power for an hour, and
still-significant percentages for days.

2) The second-worst nightmare scenario for a nuclear power plant is
loss of station power. This is what happened at Fukushima. Without
station power, all the operators can do is watch while the thing
destroys itself ... and the surrounding countryside ... over a period
of a few hours or a few days.

This is relevant because as I understand it, the safety design documents
and operating requirements assume that external grid power is available,
and they rely on this as the primary source of power for the control
room, the cooling pumps, et cetera. Remember (item 1b) that you need
ginormous amounts of cooling even *after* you have scrammed the reactor.

Combining those two thoughts, it makes a certain amount of sense that
if you have any doubt about the continued availability of external grid,
as a power sink *and* as a power source, you want to SCRAM the reactor,
the sooner the better, so as to minimize the amount of power and energy
that will be coming out of the reactor.

This policy seems to have paid off handsomely in Fukushima. The
reactors scrammed at the time of the earthquake (14:46). The first
tsunami did not arrive until about 41 minutes later (15:27). The
really big tsunami that disabled the diesel generators was even later
(15:46). During that hour, the power output of the reactors went down
severalfold. So ... even though what happened was bad enough, it
could have been much worse. I can't be sure, but it seems likely that
it /would/ have been much worse if the reactors had kept operating up
to the point where the tsunami disabled the generators.

To express a slightly more general principle: Nuclear reactors are so
dangerous that it makes a certain amount of sense to have a policy of
SCRAM first, ask questions later.

As I understand it:
-- The Fukushima reactors were down to about 10% of rated power when
cooling was lost.
-- The Three Mile Island reactor was at 100% of rated power when
cooling was lost, but cooling was restored after a few hours.
-- The Chernobyl reactor was at about 1000% percent of rated power
when it blew up.

=========

As a separate issue, at Fukushima there was supposed to be an Nth-order
backup system that used heat coming out of the reactor to run a mini-
generator sufficient to power the control room and some part of the
cooling system. That has a pleasing self-consistency: As long as there
is appreciable power coming out of the reactor, you have a system for
coping with that power. Alas, this system did not function properly
at Fukushima. I don't understand the details on this, but the story is
that this system needed to be turned on, and it wasn't turned on, or
couldn't be turned on because of control room damage and/or loss of
station power. So it wasn't a failsafe system; it failed into the
unsafe state.