Thursday, March 24, 2011

I'm not impressed when people tell me they think something has been "over-engineered." Here's why not.


Dear Readers,

Below is a news item from 2008 regarding negligent behavior in the workplace at San Onofre Nuclear Generating Station. You probably had to already be paying attention to the plant's problems when it happened to have noticed it. And even that might not have helped, since at the time, it was just one more incident on top of many. And things have only gotten worse at the plant since then.

But it's an incident that SHOULD be remembered, in light of the ongoing events at the Fukushima Dai-ichi nuclear power station.

A battery that was supposed to power safety systems in an emergency was inoperable for four years. Ho hum, there were others. Enough others? And in a separate incident in the same article, a worker "skipped hourly rounds" for five years (and, of course, falsified reports). And that's just one of many articles about problems at San Onofre. It's an accident waiting to happen.

TEPCO was a typical nuclear power company before the disaster. They were even being courted to build new reactors in Texas. (Now I think that deal will fall through.) TEPCO wasn't above suspicion, but who is? Certainly not the workers at San Onofre!

The fact is, what happened in Japan -- OR WORSE -- can happen anywhere. And don't think there's always time to "call in the Marines", assuming they'll come, even if they're right nearby -- and don't expect them to have every potential spare part a nuclear power plant might need, anyway.

But even if we had perfect human operators and emergency responders, and perfect weather, and no terrorists to worry about, and lots of spare parts, and the Emergency Core Cooling Systems all work when necessary from now on, and no asteroids (!) smash into the spent fuel pools either, which would be a real planet-killer... even then, there's still design failures which, even after 40 years, might not have appeared yet. Perhaps a study of what went wrong in Japan will reveal many unknown shortcomings of the GE Mark 1 BWR reactor design. More likely it would confirm shortcomings that were already pointed out and ignored. However, if they bury the stricken reactors in concrete it might be hard to know what went wrong. And it may be difficult to fix things if they go wrong again under all that concrete.

Of course, if we shut all the other reactors it won't matter so much what went wrong. But we haven't done that yet, have we?

In any event, we know that a lot of things failed that the public was promised would always work flawlessly. This would never happen, we were told. The public believed it.

Many people believe San Onofre was properly designed, built, and maintained. We won't know the truth until it's tested in an earthquake. But we do know that supposedly properly-built modern buildings fell (more modern than San Onofre) near the epicenters of recent earthquakes. Buildings which should have stood up.

When asked about severe earthquakes, the plant's owners have always simply assured the media and the public that San Onofre has undoubtedly been "over-engineered" (whatever THAT means) by factors of three, or five, or ten, etc.. Whatever number they need to close the gap.

But really, whatever "over-engineering" is, you can bet it costs money. But that's not the real problem. The problem is that over-engineering doesn't happen. And shouldn't.

When companies bid to build a proposed project, they don't bid on building something better than the specifications. Other companies bid on the projects as well -- competition is considered healthy. Then the lowest qualified bid is chosen, and the project begins. Where does the money for "over-engineering" then come from?

Instead what happens is you design and build things according to the design specifications. As an engineer, you develop your designs based on stone-cold calculations. You don't just "wing it"!

Thus, the final project is not "over-engineered" and if everyone tried to simply "over-engineer" everything, what would happen?

Ships would sink and buildings would topple. And perhaps worse.

Let's say your gut feeling is that the bolts should be heavier, they're not strong enough. So, believing in yourself, you put bigger bolts in. Nice move!

Ah, but you can't just do that, of course. You have to drill out bigger holes. But the space between the holes is what holds the I-beam together. You've weakened that. The building falls down in a strong hurricane. Blame mother nature if you want to. Or blame "over-engineering." Or you put the bigger bolts in, the boat becomes too heavy, and sinks in a storm. Blame mother nature if you want to. Or blame "over-engineering."

Okay, so you just buy a stronger steel than was specified. Same size, just stronger. Only you don't know that a certain chemical will be used in that building, and only very specific metals must be used.

Some metals react chemically to electrical charges that run through them. Did everyone think of what would happen when a large utility box had its main cable bolted to the floor under its own metal stand? It took more than 30 years for that to short out. But it happened, and they had to SCRAM the reactor when it did.

Let's say you find the perfect better bolt. All it cost was money -- your time to find it, and maybe it cost a lot more too. After all, it's perfect. Nuclear reactor components routinely cost ten or twenty or more times what a run-of-the-mill unit of the same specification, with the same warrantee, would cost.

But anyway, let's say you got what you wanted. Again and again, you get your changes approved. Then what happens? Well, two bad things might happen. You might make a mistake because you don't really know what you're doing, or you become complacent. After all, nobody's checking your work against the specifications, because if they did, they'd find you aren't following them. The other bad thing that can happen is, at the end of the project, you run out of money. Then what do you do? Just over-spend, as if money grows on trees? Your checks will bounce and you'll go to jail. Cut corners somewhere YOU don't think is important? At least then nobody goes to jail, right? So that's probably what you do.

When I hear the operators at San Onofre tell me they believe the facility has been "over-engineered" I cringe and so should everyone else. There's no easy way to get from a 7.0 to a 7.5 level of protection.

Instead, you do that by engineering to a 7.5 specification, not by "over-engineering" (whatever that means) a 7.0 specification just because you're altruistic and assume 7.0 isn't good enough, and no one can stop you. It just doesn't happen that way! It can't! Technology is a pretty tight box to fit in. Bridge spans aren't really supposed to fall down at all, but they do.

Over-engineered? I'm not even impressed that they try to claim that it is! And I certainly don't believe it.

They should be saying San Onofre is 7.0 -resistant, and that's all they can assume based on the specifications that it was built to. Or was it 6.9? I've heard the REAL value isn't even 7.0, it's 6.9. Small difference? If so, then it's all guesswork, isn't it?

And even the 7.0 (or 6.9) figure assumes everyone did their job properly. We have no evidence to support such an assumption. Time and again at the facility, corners were cut and problems were covered up. We know that.

San Onofre looks a lot better than it is. And it doesn't look very good.


Ace Hoffman Carlsbad, CA


At 12:03 PM 12/23/2008 -0800, MoJo <> sent:

> > > >San Onofre nuke plant battery inoperable 4 years > >The Associated Press >Posted: 12/22/2008 10:36:01 AM PST > >SAN CLEMENTE, Calif.—Federal regulators are ramping up inspections at the San Onofre nuclear plant after it was discovered a battery used to supply power to safety systems didn't work for four years. > >Nuclear Regulatory Commission regional administrator Elmo Collins said in a statement Monday that the problem stemmed from poor maintenance and the lapse is troubling because the condition persisted for so long. > >Other emergency batteries were available to supply power to plant safety systems, but the NRC said the loss of even one emergency battery reduced the plant's safety margins. It was inoperable from 2004 to 2008. > >San Onofre, on the coast between Los Angeles and San Diego, is operated by Rosemead-based Southern California Edison. > >The utility said in a statement that it accepted the outcome of the NRC's special inspection and will focus on preventing a recurrence. > >"We agree with the NRC regarding the rigor needed in problem identification and resolution. To that end, we have made significant leadership and organizational changes at San Onofre to ensure that plant performance keeps pace with continuously rising nuclear industry standards," the statement said. > >The commission said the San Onofre Nuclear Generating Station will receive additional oversight because of the "white" inspection finding. The problem was discovered on March 25. > >NRC color-coded safety inspection categories range from green to white, yellow and red. >White is a low to moderate safety significance finding. > >NRC inspectors found loose electrical connections to the battery were caused by inadequate maintenance. The regulatory commission also noted seven findings determined to be of low safety significance. > >The NRC report said the San Onofre plant's "ability to effectively evaluate problems has been, and continues to be, a concern to the NRC." > >The commission plans additional oversight at the San Onofre plant "until sustained improvements are recognized." > >The San Onofre plant has troubled the NRC for some time. > >In September, Edison executives and NRC officials met in San Clemente to discuss safety lapses at the northern San Diego County plant. > >The commission disclosed in January that Edison fired or disciplined seven employees over the last two years for safety and security violations, including one worker who skipped hourly rounds for five years and falsified hourly logs. > >Edison then appointed a site manager who replaced a half-dozen plant managers and began a new accountability training. > >——— ----------------------------------------------

Ace Hoffman
Author, The Code Killers: An Expose
Carlsbad, CA