Vector-based three-dimensional (3D) well-path optimization assisted by geological modelling and borehole-log extraction

This paper describes a novel strategy to optimize the drilling time of three-dimensional (3D) directional wellbore trajectories using a vector-based approach subject to drilling and geological constraints. Many existing well-path models require manual entry for certain geological constraints such as formation dip or kick-off limit. In contrast, this vector-based approach ensures that geological constraints are automatically satisfied by building a geological model, and extracting a borehole log of key points along the well-path. The presented approach applies and compares a deterministic optimization technique known as Constrained Optimization by Linear Approximation (COBYLA) with a Genetic-Algorithm (GA) global optimization to determine the optimum 3D well path to drill the target. While optimizing the path, the model determines the optimum kick-off point based on the subsurface-formation strength and depth subject to predetermined doglog severity, inclination and azimuth angles. The methodology is applied to well paths with different number of build-up and drop sections in unconstrained and constrained geological settings. Results show that COBYLA and GA are comparable when not using geological modelling while GA is superior for complex well-path geology-assisted optimization problems. The technique is applicable for a single well path planning, and can be expanded to a set of wells being optimized during Field Development Planning (FDP).