Application of genetic algorithms for formation of a drilling carpet and risk assessment of drilling new wells and sidetracking in conditions of geological uncertainty

Abstract

Aim. Study of the use of genetic algorithms to solve the problem of finding the optimal version of the drilling/ sidetracking carpet.
Materials and methods. To justify the profitability of drilling, a python algorithm has been developed to assess and reduce the risks of financial losses when deciding to drill new wells by choosing the optimal drilling option from all possible ones. To optimize the drilling decision-making process, a genetic algorithm is proposed, based on a combination of production profiles in such a way as to simultaneously maintain a production plateau at the stage of stabilization of production levels and minimize capital costs for commissioning new wells/laterals in conditions of geological uncertainties.
Results. The use of the proposed algorithm made it possible to significantly simplify the decision-making process on drilling new wells due to the absence of manual enumeration of all possible options, as well as to create a well-founded methodology for selecting a drilling carpet. To make a competent decision on drilling new wells in an existing field, it is necessary to take into account several parameters simultaneously. This approach is called multidimensional optimization, and its result is the selection of one or more best options. The most interesting are the wells that have the highest planned flow rate (that is, ultimately the greatest profit) and entail the least amount of costs.
Conclusion. The idea of multidimensional optimization is implemented using a genetic algorithm, which is based on the theory of probability and the law of large numbers and allows you to make the optimal choice in favor of a particular strategy for drilling a pool of new wells. The proposed approach can be successfully applied to solve optimization problems of various types.