Optimal discretization of grounding systems applying Maxwell's subareas method

This paper presents a method for evaluating the optimal number n of equivalent sources needed for simulating grounding systems by the Maxwell's subareas method. It is well known that the number of elements in which electrodes are subdivided plays a role on the accuracy and reliability of results (as well as on computational time). Previous studies, accomplished through iterative calculations (performed with different segmentations), led mostly to some recommended practices for the identification of lower and upper bounds for n. The procedure proposed in this paper allows for predicting the optimal n in a single process. The method starts from the identification of a set of appropriate scalar functions, which heuristically express a relation between the number of subareas and the accuracy of the results (earth resistance and earth surface voltages) computed applying the Maxwell's subareas method. Then, a multi-objective optimization process evaluates the number n* that maximizes that accuracy.