The 46th STIG Policy Platform Seminar: Bayesian and Non-Bayesian methods for the assessment of the risks of major nuclear accidents

June 7, 2016


Type Lecture
Intended for General public / Enrolled students / International students / Alumni / Companies / University students
Date(s) June 10, 2016 10:00 — 11:30
Location Hongo Area Campus
Venue Seminar room 2, Economics Research Annex (Kojima Hall)
Capacity 50 people
Entrance Fee No charge
Registration Method Advance registration required
Contact Science, Technology, and Innovation Governance (STIG), The University of Tokyo

How do past observations inform us of the future risks of major nuclear accidents? How did the catastrophe at the Fukushima Dai-ichi nuclear power plant change the expected frequency for such events? There has been little consensus in answering these questions. While opponents of nuclear power claim that the probability of a serious accident is very high, the industry ensures that it is negligible. Furthermore, when facing such ambiguity, or multiple sources of information, how should policy-makers behave regarding these rare but catastrophic risks? The aim of the presentation is to present two methods developed in CERNA-Mines ParisTech that try to shed light on these questions. We will first present a Bayesian method which tries to determine the effect of the Fukushima Dai-ichi accident on the probability of witnessing future major nuclear accidents. Second, we will present a non-Bayesian method which tries to account for the ambiguity that characterizes the risks of nuclear power accidents.

Related links

Access Map
Kashiwa Campus
Hongo Campus
Komaba Campus