Dealing with uncertainty on parameters elicited from a pool of experts for CCS risk assessment

Handling potentially conflicting information coming from experts when other available data are scarce or nonexistent is an important issue for subsurface operations, especially for projects in the early stages of development. The problem tackled in this study concerns the CO2 geological storage technology and more specifically the estimation, for a specific storage project, of the injected CO2 plume extent in a reservoir. For that purpose, information on the minimum, median and maximum values of the porosity parameter have been collected from 14 different experts. A classical multi-step method has been followed. It consists of first representing the data given by the experts, then aggregating the information provided by the 14 experts and finally propagating the resulting uncertainty within a dedicated CO2-migration mathematical model. Within this method, two different approaches have been considered to deal with uncertainties: a classical probabilistic approach and an approach based on the Dempster-Shafer theory. The outcomes of this study show that, in a context of high conflict, the aggregation of the experts’ data with the classical probabilistic approach and with the classical disjunctive rule within Dempster-Shafer theory might not reflect the proportion of uncertainty linked with the conflict among the experts. In addition, while the former approach may lead to specific but not conservative results, the latter may lead to too poorly informative results. In that context, the application of a trade-off rule appeared necessary to provide the decision makers with additional outcomes. In this paper, the application of a newly published experts’ information fusion rule is proposed and its ability to help decision-making is discussed.