Vicious words mark the war between pro and anti-nuclear environmentalists

[kristopher]

29. Beyond our imagination: Fukushima and the problem of assessing risk

Beyond our imagination: Fukushima and the problem of assessing risk
BY M. V. RAMANA | 19 APRIL 2011
Article Highlights
- Severe accidents at nuclear reactors have occurred much more frequently than what risk-assessment models predicted.
- The probabilistic risk assessment method does a poor job of anticipating accidents in which a single event, such as a tsunami, causes failures in multiple safety systems.
- Catastrophic nuclear accidents are inevitable, because designers and risk modelers cannot envision all possible ways in which complex systems can fail.

The multiple and ongoing accidents at the Fukushima reactors come as a reminder of the hazards associated with nuclear power. As with the earlier severe accidents at Chernobyl and Three Mile Island, it will take a long time before the full extent of what happened at Fukushima becomes clear. Even now, though, Fukushima sheds light on the troublesome and important question of whether nuclear reactors can ever be operated safely.

Engineers and other technical experts have two approaches for making nuclear reactors safe: The first is to design the reactor so that it is likely to recover from various initiating failures -- lowering the probability that the damage will spread, even in the absence of any protective actions, automatic or deliberate. The second approach, used in addition to the first, is to incorporate multiple protective systems, all of which would have to fail before a radioactive release could occur. This latter approach is known as "defense-in-depth," and it is often advertised as an assurance of nuclear safety. The World Nuclear Association, for example, claims that "reactors in the western world" use defense-in-depth "to achieve optimum safety."

Within this perspective, accidents are usually blamed, at least in part, on a lack of properly functioning safety systems, or on poor technical design. For example, analysts typically traced the catastrophic impacts of the Chernobyl accident to the reactor's lack of containment and its behavior when being operated at low power. Similarly, in response to the current Fukushima accidents, many analysts have focused on the weaknesses of the reactors' Mark 1 containment system.

Unfortunately, focusing on individual components -- rather than the system as a whole -- gives analysts a false sense of security. Here's how their thinking goes: For each safety system, there is only a small chance of failure at any given time, so it's exceedingly unlikely that more than one safety system will fail at the same moment. A severe accident can't happen unless multiple safety systems fail simultaneously or sequentially. Therefore, a severe accident is exceedingly unlikely.

Unfortunately, there are occasions when multiple safety systems do fail at the same time -- and these occur far more frequently than analysts assume. This is what happened at Fukushima. Accidents can also happen when the failure of one safety component triggers failures in other components. And in some cases, individual components work properly but the system as a whole fails. An example PDF is the Mars Polar Lander accident of 1999, when the lander's software -- working as designed -- interpreted transient signals as confirmation that the space vehicle had touched down. The software then turned off the descent engines prematurely, causing the vehicle to crash on Mars' surface. Such failure modes are hard to model within the mechanistic framework adopted by most safety analysts.

Calculating risk...

http://thebulletin.org/web-edition/features/beyond-our-imagination-fukushima-and-the-problem-of-assessing-risk


M. V. Ramana
A physicist, Ramana is currently appointed jointly with the Nuclear Futures Laboratory and the Program on Science and Global Security, both at Princeton University, and works on the future of nuclear energy in the context of climate change and nuclear disarmament. He is the author of The Power of Promise: Examining Nuclear Energy in India, to be published later this year by Penguin Books. Ramana is a member of the Bulletin's Science and Security Board.