A Geostatistical Method for Characterizing Superpermeability From Flow-Meter Data: Application to Ghawar Field

This paper presents the application of multiple point geostatistical modeling, combined with a new technique for model perturbation, to characterize abnormally high flow capacity geologic features, super-k, within an 11-well study area in a large, Middle Eastern, carbonate reservoir. The reservoir characterization adequately history matches well flow-meter data measured in 2 surveys, taken over a 3 year period, obtained in 2 of the study area wells. Characterization of the abnormal flow features is desired for future placement of water injection or production wells, to decrease the probability of wells being completed in the abnormal features, avoiding quick water breakthrough. The high conductivity features are an aggregation of two principle geologic components: high permeability beds of various lithologies, and high conductivity discrete fractures. These components form a flow network of unusually high conductivity, often much higher than predicted from flow capacity calculations in wells intersecting the network. Simulation of flow in the discrete fractures is performed using well models. Well models offer several advantages over fracture discretization and dual porosity models, for flow simulations performed as part of an optimization algorithm. A training image of facies architecture is developed based on conceptual depositional models of the reservoir. Using multiple-point geostatistics, various geologically realistic reservoir models can be generated, constrained to facies data from wells. Discrete fractures are associated with certain elements of the facies model, and therefore are added stochastically. A history match of flow-meter data is obtained with a single parameter optimization of the geostatistical model realizations, using a technique in which facies probabilities, rather than facies events, are perturbed. This method maintains integrity to the underlying geological model, while history matching, hence making prediction of future super-k occurrences possible. Introduction This paper summarizes the characterization of the spatial distribution of permeability, in a small area within the Ghawar Field, Saudi Arabia. Fig. 1 shows the study area within the Ghawar Field, the largest of a number of oil fields in the region, and, in fact, the largest field in the world. Fig. 1. Study area location. SPE 84279 A Geostatistical Method for Characterizing Superpermeability From Flow-Meter Data: Application to Ghawar Field Joe Voelker, SPE, Stanford University; Jim Liu, SPE, Saudi Arabian Oil Company; Jef Caers, SPE, Stanford University