1 ! ! " # $" % # &"’"(" ) 1 EC DG Joint Research Centre, Energy Systems Evaluation Unit, Westerduinweg 3, 1755 ZG, Petten, Netherlands. 2 TU Delft, Faculty of Technology, Policy and Management, Jaffalaan 5, 2628 BX, Delft, Netherlands. E/mails: alessandra.colli@jrc.nl dan.serbanescu@jrc.nl b.j.m.ale@tbm.tudelft.nl * Risk indicators, risk characterization, relative ranking, comparison, fossil fuels % The primary goal of the paper is to illustrate and apply a methodology to compare and rank risks from different energy systems. The application takes into account three fossil fuels, coal, natural gas, and oil. Comparative analyses are based on a set of Risk Characterization Indicators (RCIs) that consists of 17 indicators to highlight different risk aspects, developed on the backbone of a causal structure for accidental events in the energy sector. Although RCIs are mainly event/oriented, they can be used in different contexts: (1) analysis of single events, (2) comparison of events with similar characteristics, (3) comparison of different energy technologies based on aggregated values, and (4) use within Multi/Criteria Decision Analysis (MCDA). The objectives of this paper are threefold. First, a concise overview of the RCIs, and the associated grouping and ranking methodology, is given. Second, few illustrative examples of accidental events from coal, natural gas, and oil in open/source form are chosen from the database ENSAD (Energy/related Severe Accident Database), which since its establishment in the early 1990s has been continuously maintained and extended by the Paul Scherrer Institut. The events are processed by the RCIs and ranked according the proposed methodology. This allows evaluating the applicability and usefulness of the RCIs in regard to different energy technologies. The evaluation is done considering the logical development of the method and using the feedback of existing literature on risks of fossil energy chains to confirm the validity of the results. Third, the benefits and limits of the energy risks comparative methodology are discussed, considering its potential to become a future support to decisions on energy/related risk aspects. #" This paper focuses on the validation of a set of Risk Characterisation Indicators (RCIs), which have been developed to assist in the comparison of energy risk expressions from different energy systems, and mainly wants to show the methodology to group them and subsequently to rank events to which the RCIs are applied. The present work brings further what already initiated in (Colli et al., 2008), enhancing the set of RCIs with an innovative grouping and ranking approach based on PRA/related concepts and logic. Few selected events from the ENSAD (Energy/related Severe Accident Database) database / in limited open/ access form / are used to validate the process. They cover accidental events (not terrorist attacks) of significant outcomes from the coal, natural gas, and oil chains, for the extraction, treatment, and regional transportation stages of the fuel cycle. The results obtained are discussed in view of the existing literature on fossil fuels. In addition, two alternative evaluation paths for the RCIs are introduced and evaluated on the base of the same group of ENSAD cases in comparison with the main method in the paper. To be complete, the presented methodology needs the evaluation of the quality of the process. This will be obtained through the evaluation and quantification of the uncertainty of the process, based on the contribution of the various uncertainties coming from each single step performed. This part is under development. The paper is divided into nine sections. After the introduction, the second section recalls useful information about the set of RCIs (Colli et al., 2008). The third section gives a theoretical description of the innovative grouping and ranking methodology. The fourth section applies the risk comparative methodology to a selected number of accidental events affecting fossil fuels extracted from the ENSAD database. The fifth section discusses the results obtained for the single steps of the fuel cycle with possibility of analysis. The