Transparency in risk assessments - Presenting the 'expectation value' of post-closure risks from radioactive waste repositories

The Environment Agency of England and Wales (the 'Agency') has an extremely broad regulatory remit covering aspects of flood defence, integrated pollution control, water quality, waste management, abstraction control, navigation, fisheries, conservation and recreation. Risk assessment, as a regulatory and management tool plays an essential role in the targeting and prioritisation of this activity, as well as in aiding site-specific decisions on authorisations for abstraction, discharge and/or disposal. From a regulatory perspective, the majority of the Agency's risk assessment activity is focused on critically reviewing risk assessments submitted to the Agency in support of requests for authorisation. With increasing calls for openness in all areas of regulatory decision-making, new demands are being placed on risk assessments with a view to allowing far more transparency and traceability of 'process' and 'content' than has historically been the case. The Agency is responsible for the licensing of radioactive waste disposal facilities in England and Wales. It has issued guidance on what is expected of an application for an authorisation to dispose of low and intermediate level radioactive waste to land - the 'Guidance on Requirements for Authorisation' (the 'GRA'). The GRA includes a risk target and places a strong emphasis on confidence-building during the preparation and assessment of post-closure safety cases for such facilities. In this paper we discuss a recent study commissioned by the Agency which has examined the use of expectation value of risk in assessments and considered ways of improving transparency. The study has concluded that the expectation value is an appropriate measure of risk for comparison with a single-value criterion, provided that the scope of the assessment does not involve undue speculation regarding the FEPs (Features, Events and Processes) to be included. Low probability or speculative events and processes for which no data can be derived from site characterisation or laboratory experiments might be more suitably assessed through 'what-if' calculations outside the risk assessment. Transparency might be enhanced by a wide-ranging debate, involving developer, regulator and other stakeholders, about the types of low probability or speculative events and processes to be examined in this way. If there was a consensus view that a particular event or process of this nature should be given greater emphasis, steps might be taken to modify the safety case, and possibly the design of the facility. The developer would maintain a basis for proceeding, while stakeholders would have the assurance that a mechanism existed for addressing their concerns in the overall safety case presented to the regulator. The study also concluded that the uncertainties concerning future populations were likely to be too great for assessments of group risk to be useful in regulatory decision-making. Comparative analyses of group risk may however be beneficial in making and explaining design decisions