It is generally thought that data from at least 20 or 30 wells are needed to generate the statistics for inter-well interpolation. We show in this paper that it is possible to use as few as two wells to generate a cross-section, and to use as few as three wells to generate a three-dimensional reservoir description. Geostatistics describes a reservoir with two models: a distribution model that deals with data values, and a spatial correlation model (the variogram) that deals with variation from location to location. Examples are considered for the Kern River Field in California to show how the method can be applied. Three major areas of application for geostatistical interpolation include computer visualization of the reservoir, Monte Carlo simulation for sensitivity study, and reservoir heterogeneity characterization for other reservoir models.

You can access this article if you purchase or spend a download.