Development of a numerical tool for design of the Bottom Hole Assembly (BHA) using Artificial Neural Networks

摘要

This paper describes the use of Artificial Neural Networks in developing a design tool for calculating BHA (Bottom Hole Assembly) parameters. The program streamlines and simplifies BHA design calculations for a planned well curvature by calculating half-wave lengths using nomograms developed by Lubinski-Woods and Muzaparov. Nonlinear normalisation of data was applied on nomogram data to significantly improve accuracy of the modelling and design. The design data calculated by the program were compared with the real data and have been verified on deposits in the Aktobe region in Kazakhstan. It is concluded that the proposed tool provides more accurate results in a shorter time, allowing designers to compare and analyse the influence of design parameters on calculations of half-wave, loading and well curvature.