http://onlinelibrary.wiley.com/doi/10.1111/risa.12587/full
For example, the Fukushima accident and the Chernobyl accident are rated 7 -- the maximum severity level -- on the INES scale. However, Fukushima alone would need a score of between 10 and 11 to represent the true magnitude of consequences, the researchers said.
To remove a possibility of such disasters would likely require enormous changes to the current fleet of reactors, which is predominantly second-generation technology, Wheatley noted.
But, "even if we introduce new nuclear technology, as long as older facilities remain operational -- likely, given recent trends to extend permits and relicense existing reactors -- their risks, and the aggregate risk of operating the global nuclear fleet, remain," Sovacool said.
The results were published in the journals Energy Research & Social Science and Risk
Analysis
And another great article from the same authors.
http://www.sciencedirect.com/science/article/pii/S2214629615301067
--------------------------------------------------------------------
An interesting query with many links to nuclear risk analysis
https://www.elsevier.com/search-results?query=risk%20of%20nuclear%20accident&labels=all
---------------------------------------------------------------------
You have to be careful when reading "risk analysis" reports. In most cases, the assessmentof likelihoods are not very scientific.
ReplyDeleteAdditionally, the methodologies of "social sciences" are not rigorous enough. Ill have to logon today and read the article more closely. Two data points that are disparate as Chernobyl and Fukushima need reconciling. A pdf for extremely unlikely events includes Gomperz, which the authors dont mention in their abstract. Also, in examining all nuclear mishaps, the frequency is heavily skewed the first 30 years. The last 30 years, the frequency drops off to 1/10. But frequency isnt likelihood because the Uniform Distribution doesnt apply. Additionally, the authors may not realize that the existing modality for nuclear, e.g Gen 3 reactors, will be gone the next 30 yrs. Technical unnovation, and walk away reactor safety would drive likelihoods even to remote.
ReplyDeleteRisk will always be the product of likelihood times consequenses. Social scientists will always misrepresent both terms in an effort to mischaracterize overall risk. If their methodologies are transparent, then one should be able to determine risks of many activities, and then compare with nuclear. Yet they seem to conviently leave that out of their work.
Indeed, no matter how small the risk, if the outcome is infinitely negative, then the risk is infinitely unacceptable.
DeleteZero times infinity is still zero. Nothing has a zero likelihood. The only infinite consequence is death. Ask Jose Fernandez.
DeleteDeath of our species, still infinite. Nothing else can do it, nuclear can.
DeleteGuess you havent watched the many zombie apocalypse movies and tv shows. Bioweapons are more likely to destroy mankind than nuclear. You realize no expert gives a Carrington event as a cause of any meltdown. Reactors that trip on undervoltage go into RCP coastdown then nat circ.
DeleteYet focusing on consequenses without considering likelihood isnt risk assessment but risk aversion as you noted. People think that just because consequences are large, that risk is large. Need 2 components. Likelihood a priori isnt a frequentist exercise.
ReplyDeleteBullcrap comment, shame on you.
DeleteNo its true. Risk by definition isnt exclusively consequences but the product of likelihood and consequences. You know the insurance industry and actuary tables are based on this metric.
DeleteYou cant redefine what risk means to suit your needs. The safety engineering discipline, which is not exclusive to nuclear, has this definition, not yours.