How accurate are Capacitance Model connectivity estimates?

A number of methods have been proposed to evaluate well pair connectivity, including the Capacitance Model (CM). Since its development in 2006, numerous published reports using the CM (also called the Capacitance-Resistance Model) have appeared. Here, we take a closer look at the CM and evaluate its performance for evaluating interwell connectivity. is a tool to analyze injection and production rate data for interwell connectivity. It is derived from a linear productivity model in the pseudo steady state regime for slightly compressible fluids. Therefore, we expect within a specific range of reservoir and fluid properties such as permeability, compressibility, viscosity, porosity, and number of wells per area, the results are accurate and repeatable but these ranges have not hitherto been quantified. Furthermore, there is a statistical side of CM performance involving the number of sampling data, sampling interval, and amount of noise which affects the accuracy of the results. Before applying the CM on field data, a sensitivity analysis on the accuracy of the expected results is desirable. ried out a sensitivity analysis on the listed parameters by changing each of them and keeping the others constant during simulations. From the analysis, we define a dimensionless number, the CM number (C), to specify a range in which we can apply the CM with accurate and repeatable results (0.3<C<10). However, we may not have some of the information needed to evaluate C while working on field data. To solve the problem, we applied the bootstrap to evaluate the uncertainty and provide unbiased estimates and find that 50 iterations are sufficient to evaluate CM results accuracy. Finally, by calculating C and L (the ratio of number of samples to the number of model parameters) from the available information, we estimated the uncertainty of evaluation for 11 published field cases. ndings in this work not only help us to estimate the accuracy of the CM before applying it to field data, but also assist the CM users to assess uncertainties of the results and understand the effects of noise and number of measurements after applying the model.