Integration of Geostatistics and Well Test to Validate a Priori Geological Models for the Dynamic Simulation: Case Study

Abstract

Integration of Geostatistics and Well Test to Validate a Priori Geological Models for the Dynamic Simulation: Case Study. Abstract To estimate the economics of marginal development with large uncertainties due to a reduced appreciation, wells performances should be evaluated as precisely as possible. This cannot be achieved without a correct quantification of the heterogeneities governing the dynamic processes. in this perspective, geostatistical models have to be built using, as usual, static data, but also dynamic data: well test response should be taken into account as constraint for the geostatistical modelling in an integrated study combining geological and dynamic considerations. A refined 3D geological model built using the geostatistical object method is validated by dynamic data when completely matching the corresponding well test response. The petrophysical values assigned to each facies are found by the matching process. The Bouboute field is a marginal structure appraised with one well, containing hydrocarbons in the Z Formation. This formation is characterised by heterogeneities coming from different depositional environments, with large uncertainties on the sizes and shapes of sedimentary bodies. Because of this geological complexity, the geostatistical approach constrained by well test data is performed. A good match is obtained with 20 – 60 m reservoir bodies while supposed much bigger initially. It allowed to perform a full field simulation on the upscaled model with sensitivity cases. An effect of acceleration of the production due to better Productivity Index (PI) and a second effect of production mechanism due to the percentage of good facies is observed on these sensitivity cases. The integration of dynamic data to constrain geostatistical modelling is thus of great interest to provide supplementary information in this kind of development study and should, especially in the appraisal phase, be used when possible to validate a priori static models. Introduction The appraisal phase is a very important time in a life of a field, especially when the range of reserves uncertainty is close to economic viability. The uncertainties on the HIP and the reserves are often due to the presence of heterogeneities in the reservoir. The distribution of such heterogeneities is difficult to predict but conducts the well performances. In such a context all the data, either of static or dynamic nature, providing information about the reservoir model of heterogeneities should be integrated in the reservoir modelling to reduce the range of uncertainty of the field production. In this perspective, different geostatistical techniques have been developed during the last ten years for modelling heterogeneities, quantifying uncertainties and integrating data of different kind seismic as well as sedimentology, diagenesis or well tests. Furthermore, the fields which provide such kinds of economic viability problems are usually small structures, where the appraisal phase is reduced because of the associated costs. Only few data are thus available to define the optimal development of the field : The seismic at large scale, the logs and the cores at small scale. Both these data are of static nature. Only the Well Test is a technique which provides dynamic information at a medium scale. Only some of the main characteristics of it are usually used in a classical study: distances to limits, permeabilities, skin or partial penetration. In the case of large uncertainties on the geological model, the whole pressure response should be used to validate the a priori geological scenarios, by integrating the dynamic data to constrain geostatistical realisations. The aim of this study is to evaluate the reservoir connectivity and productivity potential of a very heterogeneous field in a small structure. This is performed through a geostatistical modelling of the reservoir, based on the geological scheme of deposition. P. 491^