RE: Nuclear Power

Sweitzer, Dennis (
Fri, 10 May 96 07:54:00 EST

I wrote
>> Maybe it is time to redebate the issue, since this is 1996 and Three Mile
Island & Chernobyl have occured in the meantime. Not to mention safety is
only part of the issue. Economics are changing quite a bit as well.

Joel responded,
>Three Mile Island happened prior to (and inspired) the discussion at the
ASA meeting and the articles in the Journal. I know of no subsequent
findings would change the way I and the technical community would view that

I didn't remember when it was, and I wasn't part of ASA back then.

> Significantly, the Browns Ferry accident which to my mind (and to a
number of other people) posed a much more serious risk to the public
had also happened. My talk dealt with these two accidents. My writing
skills were such that the talk was not published.

I'm unfamilar with the Browns Ferry accident. Feel free to explain.

> As I stated earlier, to people familiar with nuclear power, Chernobyl
differs technically (it was a graphite moderated, gas cooled reactor in
contrast to being water cooled, and it had no containment--the
domes you see at some nuclear plants), and culturally in the way Russians
approached safety. I heard while in grad school long before the accident
about the way the Russians did safety analysis (Assumption: There will be no

I've heard a joke about the differences between TMI and Chernobyl: At TMI,
the operators should have been awake, and at Chernobyl, they should have
been asleep.

My understanding about TMI is the following (feel free to correct me). At
some point, someone closed water valves for a back up system (routine
maintainance), but negelected to reopen them. Operators were
sleepy--prehaps some were asleep. Water levels went low, alarms went off,
people woke up, but a sticky gauge showed water in the reactor
core--contrary to all other indications. Automatic systems weren't quite
working properly (which happens). Eventually, they flooded the core. Later
examination of the core showed that it had collapsed and would have melted
through the bottom in another 15 seconds.

The fundamental problem with boiling water reactors is that you have 90
seconds to react once the coolant is lost from the core. The nice thing
about gas cooled reactors is that you can have hours to respond (depending
on the design).

My understanding about Chernobyl is the following. Late one night the
operators did some unauthorized experiments, involving low power output of
the reactor. The operators were not nuclear engineers/scientists and did
not know--until they experimentally verified it--that these reactors had a
quirkly flaw that would cause an huge power surge under the circumstances
they did.

At TMI the containment dome did keep in the radiation. But in another 15
seconds the core could have melted through the bottom rendering it nearly
useless. I'm not sure that a containment dome would have worked at
Chernobyl (it was a tremendous explosion).

In either case, operator error were the major factors. The human subsystem
is often the hardest to engineer, which explains a lot about TMI, Chernobyl,
and the fact that one of the most common causes of airline crashes is
"controlled flight into terrain".

>I have no quarrel with your economic statistics although I am surprised by
them and wonder whether they would withstand more critical scrutiny than I
can give them. They don't come from Amory Lovins do they?

They suprise me too. None are directly from the Lovins, but we may be
ultimately referencing some of the same sources. I've been taking notes
from a variety of sources for a long time. The Lovins have optimistic
figures, and I have to question some of their assumptions. I would regard
them as vanguard visionaries--their analyses are useful to get the ball
rolling, but others will need to do the many detailed analyses necessary to
bring it into reality.

Traditional corporate planning and forecasting almost inevitably is based on
what happened in the past. The future is a lot trickier, and requires a
different (and broader) skill set. Consider that ARCO was the only major
oil company to anticipate OPEC and the oil embargoes. And now ARCO is (or
was) making solar cells (they may have sold off that division to Siemans, or
are doing a joint venture. I'll have to check my notes).

> I am not so enamored with nuclear power that I want to spend more to have
it. It is also my fuzzy recollection that other developed countries can
build reactors safely, ... for something
like half of what it costs here ....

If European nuclear electricity is $0.04/kwhr after 40 yrs of heavily
subsidized developement, and that from new wind technologies is at $0.05
(and falling) after <20 yrs of fragmented development, what can we expect in
another 10-20 years?

I guess my point is that the broad trend is toward renewable energy sources,
largely because of intrinsic economic factors and the usual pattern of
technological development. It's safe to say that once solar power drops
sufficiently far, it's cost effectiveness will be more a function of the
cost of the power line than of what is feeding into the power line.

>The differences were attributed to differences in the regulatory and
licensing processes (I am for licensing, regulation, and public input
but my time in the nuclear industry lets me see the other side of the
fence; it can be done sensibly and can be done insensibly).

Nuclear is intrinsically dangerous. Otherwise, we wouldn't have to build
massive "armored" containment domes and worry about storing spent fuel for
10,000 years. It takes a lot of work to make it safe. And once we make it
as safe as is humanly possible, is that safe enough? The odds of TMI
happening may have been predicted at 1 in a million before it happened.
Based on observation, it would seem that the odds of TMI occuring were much

Can we make nuclear safe enough? On average, yes. Is that safe enough?

In risk analysis, one must consider both the average expected hazards and
the worse case scenarios. The nuclear industry focused on average expected
hazards, and the anti-nuke folks focused on the worse case scenarios. The
former depends too much on accurately assessing the probabilities of the
various disaster scenarios--probabilities that can easily be way off due to
wrong assumptions and Murphy's law.

One last comment. I really like nuclear fusion power, but one thing has
struck me. It could well be that when it is finally developed in 50 yrs or
so, it won't be able to compete with renewable energy sources. I find that
very ironic.

Grace & peace,

Dennis Sweitzer