Slashdot Mirror
25th Anniversary Of Three Mile Island
fbform writes "March 28, 2004 is the 25th anniversary of the Loss Of Coolant Accident (LOCA) at the nuclear power plant on Three Mile Island, Pennsylvania. It's a good time to reflect on the impact it has had on our nuclear safety policy and interface design in general."
Oblinks to related wikipedia articles:
Three Mile Island
List of nuclear accidents
Especially pebble bed reactors.
Compare the Soviets worst accident- dozens dead in the short term, thousands dead early from long term effects- with the United States worst accident- Three Mile Island. The radioactivity release from TMI was projected to cause less than 1 premature death from the hundreds of thousands of people potentially exposed to anything, and in twenty five years since, no one has been able to prove that they were adversly affected by the accident, healthwise.
Including the people who work there.
Nuclear Power is perfectly safe when done right, and it's done right in the US. The worst that could happen in the US in an accident condition is that parts of the power plant are destroyed. And for even that to happen, so many very closely watched things would have to go wrong that it's basically not going to happen.
So shut off your lights if you don't like nuclear power, and go back to your cave.
Alcohol, Tobacco and Firearms should be the name of a store, not a government agency.
People do not realize that:
1) 80% of electricity in France is nuclear (Paris vacation, anyone?)
2) There is more radiation in the U.S. Congress due to its granite construction than is permitted outside a nuclear plant
3) If you take 4 cross-country airplane trips, you get more radiation than allowed at nuclear plants
4) If you live in mountains (Colorado) you also get more radiation, due to the altitude
5) Best estimates are for 325 long term general population deaths arising out of the Chernobyl radiation escape. Guess how many cancers due to oil/coal burning plants elsewhere?
6) Current nuke plant designs have a bias for automatically stopping the reaction at the slightest or even gravest out of spec situation. Imagine your car's engine designed to stop every time you rev up/speed/your dome light burns out.
Fact is, greenies have scared the public, we are currently poisoning our air with oil/coal power plants, creating thousands of new cancers every year. Thanks, tree-huggers.
The next pasture is always greener
The parent is correct.. if worded badly.
TMI was a case of automatic safety systems being overrided by undertrained human operators. As the story paragraph mentioned, TMI was a stark lesson in control systems design.
In the control room the operators had no feedback about how much water was in the reactor core, just one gauge showing the level of water in the pressurizer tank near the top of the system. When a valve near the top of the pressurizer stuck open (referred to as the PORV or pressure operated relief value) the steam that normally kept the water near the bottom of the pressuizer tank started leaking out. More water flashes to steam.... and TMI is now losing water. The operators saw the opposite, the water level was rising on the level gauge for the pressurizer and they started reducing and eventually draining water out of the system thinking some malfunction was causing water to be introduced. None of the operators was able to step back from the initial theory that water levels were rising, despite large amounts of contradicting information. (Hours into the incident an off-duty operator arrived and with a fresh set of eyes figured out what was happening)
There are a lot more things that went wrong that night... (the initial shutdown was caused by water accidentially getting into the compressed air supply for the pneumatic control systems in the steam room, a valve closed at the wrong time and burst one of the steam lines to the power turbines)
TMI is a fascinating example of how multiple redundant systems still can fail, given a long string of "coincidences" One can argue that failures of this type are like winning the lotto, their is little chance of it happening on on particular day, but given enough days it is certain to happen to someone. Hence the need for "fail safe" designs.
For years now, the folks guarding nuclear plants have been armed to the teeth.
They have no-fly zones around nuclear plants now. Not really because flying inside the line gets you shot down, but so they can aim a SAM at an incoming threat without worrying about hitting the wrong plane (not that they're worried about hitting the wrong plane - it's really that they're worried about missing the right one).
So let's pretend we're mad as hell and not going to take it any more. What's the plan?
9/11 style air attack won't work. You'll either get SAMed or the containment building will likely survive the impact.
Armed assault will be met with armed resistance. The minute the attack starts, someone presses the panic button and the cavalry arrives.
No, the only credible terrorist threat in my mind is an inside job - someone gets a job as a plant worker and sabotages the plant. If the plant were a fail-safe design, however (as a previous posted pointed out, current plants are designed with redundant systems, but are not fail-safe), the worst the criminal could do is shut the plant down and perhaps try and disperse the fuel with explosives (note that due to a failsafe system, he won't get any help dispersing the fuel from the plant itself). His ability to smuggle explosives into the plant without being detected will limit the effectiveness of that plan. Never mind that he'd have to be able to breach the containment building (yes, even a fail-safe reactor will likely have one).
Sabotage is certainly a threat at current nuclear facilities, just as it is a threat at, for example, petrolium refineries (I'd actually put Richmond, CA ahead of, say, San Onofre on a threat list). Better design mitigates that risk, just as it mitigates so many other risks.