"First, due to changing environmental conditions (sea level rise, subsidence, changing storm activity, etc.), historical records may no longer be reliable predictors for future risks.
For example, the summer of 2003 was unusually hot. Many of the French nuclear power stations are cooled by river water. But, in 2003, the rivers were so warm, they couldn’t be used to cool as normal. That caused the powering down or shutting off of 17 French nuclear reactors. It cost the French utilities hundreds of millions of dollars to buy power from neighboring countries.
This ‘anomaly’ happened again in the summers of 2006 and 2009, again causing powering downs at French nuclear facilities. According to the Hadley Center, by 2040, it will be ‘commonplace‘ for European summer temperatures to reach 2003 levels.
This new environmental change variable is often left out of risk assessments for all manner of new infrastructure builds. Dams in India are seeing reduced generation capacity as a result of shifting monsoons. Sections of oil and gas infrastructure along the U.S. Gulf Coast are suffering repeated shutdowns due to flooding, hurricanes and subsidence. Homes in the US and UK are already being built on actual floodplains, let alone areas that are likely to become floodplains."
One question I have for Cleo (and have posted on Huffpost) draws from her statements about risk assessment. She gives high marks to actuaries and the insurance companies. But, we should be very careful about attributing perfect knowledge to private-sector actors, while going against government. Certainly, the lessons of AIG show that the insurance industry is not always very good at assessing risk. Maybe structural and engineering risk is different from financial risk, but there are still questions here.
Also - it is important to note that the Indian parliament decided not to allow the liability cap on nuclear power. The implications of that, though, are far from clear. The result very well could be that companies from the United States are prevented from investing there, because of (justifiable) investor concerns about being stuck with a potentially unlimited bill for a nuclear accident, while Russian nuclear companies face no such concerns - because their decisions are made for political, not economic reasons. The likely result here would be Indian investment in reactors that are not as high a standard as possible. So - we have to be careful about creating perverse incentives with policy decisions like this.
Hey Andrew, thanks for the useful additions to the conversation, and good to hear from you. I was constrained by length over at HuffPo, so am glad for the opportunity to expand/clarify/move the discussion forward. Below is what I posted at HuffPo in reply. Thanks again, Cleo
ReplyDeleteThanks, Andrew! Very useful comments.
You are right that the insurance sector may not always be perfect at calculatin g exact risk -- or even close to it especially as, theoretica lly at least, they are supposed to hedge their bets. However when it comes to infrastruc ture, they usually think they err on the side of caution (whether they do or not, especially as they rarely take environmen tal change in to account, is another issue).
What would be ideal is a series of checks-and -balances including insurance, regulators , area specialist s, etc. Essentiall y people who will have to pay if things go wrong over the long-term.
What I was trying to get at is that, in certain cases, the 'check' provided by insurance has been fatally distorted by various interests. Another example is the way U.S. government flood insurance allows for building in areas that the private sector has deemed too risky. Or allowing BP to self-insur e if private sector says Deepwater Horizon is too big a risk. Insurance isn't the stand-alon e solution, but it can be useful if not distorted.
I read the Indian liability cap was passed: http://www .hindu.com /2010/08/2 6/stories/ 2010082653 670100.htm
Has something changed?
That doesn't invalidate your larger point though, there are lots of other examples of 'perverse incentives ' that need to be avoided when structures are put in place.