Introduction and review

The need to appraise and develop smaller and technically challenging fields, and to maximize recovery from mature fields, requires a high level of geoscientific input and application of innovative methods. Optimal positioning and geosteering of wells, and drilling and completion of ‘intelligent’ wells, are only successful if the relevant reservoir data are available and implemented in reservoir models. Modern seismic data reveal ever more information about reservoir heterogeneity and fluid distribution, often allowing definition of intra-reservoir features. In mature fields, production data frequently identify reservoir compartmentalization and validate models of geological heterogeneity. Recognizing that existing reservoir models may be inconsistent with current data is an important motivation for revisiting existing concepts and hypotheses. Recognition that all reservoir models are approximations of reality produced from incomplete data leads to the use of probabilistic methods to estimate uncertainty. As new data are acquired, new methods of processing and analysing data are developed, or new concepts are applied to interpretation of data, the need to change models becomes apparent. New concepts may come from reworking and re-examination of existing data or from acquisition of new data. Sometimes changes are enforced as predictions of reservoir behaviour prove erroneous and remedial action is required to secure profitability or to cut further losses. The stimulation to carry out reservoir studies may occur at any time during field appraisal and development. For example, a study may be carried out to assess uncertainty and reduce risk during planning of field development or, to extend the life of a . . .