Geostatistical Reservoir Modeling - Clayton v. Deutsch - (Nafti - Ir) [PDF]

Citation preview

GEOSTATISTICAL RESERVOIR MODELING



APPLIED GEOSTATISTICS SERIES General Editor Andre G. Joumel Clayton V. Deutsch, Geostorisrical Resen'oir Modeling



Jean·Laurent :vtallcl, Geolllodeling Pierre Goo\'aens. CCos/l//I.nicJ for Nawraf Resources Em/l/arioll Clayton V. Deutsch and Andre G. Journel. GSLlB: Library a"d u.ta".( Guide. Secollt! Edilioll



Geosrn.'i.~fic{/} SOjrl\'(1/"(!



Edward H. Isaak and R. i\loh:ln Sn\'3!>1,3\'a. All Illtrodllctiulll(l.4,pphed



Geostallsrio



GEOSTATISTICAL RESERVOIR MODELING Clayton V. Deutsch



.j()l-:'L~:l.?~ cJl0 \', se" The parameters of the geostatistical modeling technique are also uncertain and they could be made \·ariable, that is, "randomized," to lead to a larger and possibly more realistic assessment of uncertainty. The mode!~ ing approach itself could be considered uncertain and alternative approaches or geological scenarios considered. At some point this circular quest for a realistic assessment of uncertainty must be stopped [1491, Uniqueness and Smoothing Conventional mapping algorithms were devised to create smooth maps to reveal large-scale geologic trends; they are low-pass filters that remove highfrequency property variations. The goal of conventional mapping algorithms such as kriging, splines, inverse distance, and contouring algorithms is Tlot to show the full spectrum of patterns or variability of the property being mapped. For Auid Aow problems, however, the spatial patterns of extreme high and low values of permeability often have a large effect on the flow response. Geostatistical simulation techniques, conversely, are de\'ised with the goal of introducing the full variability, creating maps or realizations that are neither unique nor smooth. Although the small-scale variability of these realizations may mask large-scale trends, geostatistical simulation is more appropriate for predicting flow performance and modeling uncertainty, Analogue Data There are rarely enough data to provide reliable statistics, especially horizontal measures of continuity. for this reason, data from analogue outcrops and similar more densely drilled resen'oirs are used to help infer spatial statistics that are impossible to calculate from the present subsurface reservoir data. \Ve acknowledge that there are general features of certain geological settings that can be transported to other reservoirs, provided they originate from similar geological processes. Although the use of analogue data is essential in reservoir modeling, it should be critically evaluated and adapted to fit any hard data from the reservoir being studied. Data Integration The goal of geostatistical reservoir modeling is the creation of detailed numerical 3-D geologic models that simultaneously account for (honor) a wide range of relevant geological, geophysical, and engineering data of varying degrees of resolution, quality, and certainty. This data integration must be accomplished by construction rather than by sele