QUANTITATIVE EVALUATION ON THE UNCERTAINTY OF RESERVOIR GEOLOGICAL MODEL WITH DISTANCE METHOD

Aiming at the main problems in the common evaluations on the uncertainty of reservoir geological model, this paper presents a method to calculate the differences between two models with distance function, then estimate the degree of uncertainty based on the differences. Taking the western WZ Oilfield for example, the facies-controlled physical parameters modeling technology and sequential Gauss stimulation method are used to establish 3D permeability models. Comparing and analyzing various distance functions, the results show that the Manhattan and Euclidean distance functions can describe the differences between models better. The principle of Euclidean distance function is calculating the quadratic sum of permeability value in each grid node between two models first, and then taking the square root to get a dissimilarity matrix characterized the differences between models. The difference between models can be directly obtained with the matrix:The greater the difference between models is, the greater the uncertainty of simulation result is. Finally, the comparative analysis results of intersection map and distance matrix show that this method is effective in evaluation of the uncertainty of reservoir geological model generated by stochastic simulation.