New correlations to calculate vertical sweep efficiency in oil reservoirs using nonlinear multiple regression and artificial neural network

Abstract

Estimating oil recovery factors and oil reserves depend primarily on vertical, areal, and displacement efficiencies. This work presents a new accurate correlation to estimate the efficiency of the vertical sweep relevant to water/oil ratio (WOR), reservoir permeability variation (V), and water-to-oil mobility ratio (M). This correlation covers the mobility ratio of (0 ≤ M ≤ 10) and permeability variation of (0.3 ≤ V ≤ 0.8). The proposed mathematical correlation to estimate the efficiency of vertical sweep in the above range achieves a high correlation coefficient of 0.9999 with an average percentage error of −0.05 %, an average absolute percentage error of 0.53 %, and a standard deviation of 1.37 %. Furthermore, new graphical and mathematical correlations were proposed to determine the efficiency of vertical sweep and the correlating parameter (Y) covering a wide range of 0 ≤ M ≤ 50 and 0.1 ≤ V ≤ 0.9. These correlations achieve a high determination coefficient of 0.99 with an average percentage error of −0.039 %, an average absolute percentage error of 2.49 %, a root mean square error (RMSE) of 0.018 and a standard deviation of 3.5. In addition to the proposed correlations, this work presents an artificial neural network model to calculate vertical sweep efficiency for the wide data range as a direct function of M, WOR, and V, with high accuracy as deduced from the calculated statistical error analysis. To the best of our knowledge, no published mathematical correlation covers this range of data presented in this work to calculate the vertical sweep efficiency.