Fast Drilling Optimizer for Drilling Automation

Abstract

Abstract The rate of penetration (ROP) was optimized using a particle swarm optimization algorithm for real-time field data to reduce drilling time and increase efficiency. ROP is directly related to drilling costs and is a major factor in determining mechanical specific energy, which is often used to quantify drilling efficiency. Optimization of ROP can therefore help cut down costs associated with drilling. ROP values were chosen from real-time field data, accounting for weight on bit, bit rotation, flow rate variation along with bit wear. A random forest regressor was used to find correlations between the dependent parameters. The parameters were then optimized for the given constraints to find the optimal solution space. The boundary constraints for the ROP function were determined from the real-time data. The function parameters were optimized using a particle swarm optimization algorithm. This is a meta-heuristic model used to optimize an objective function for its maximum or minimum within given constraints. The optimization method makes use of a population of solution particles which act as the particle swarm. These particles move collectively in the given solution space controlled by a mathematical model based on their position and velocity. This model makes use of the best-known solution for each particle and the global best position of the system to guide the swarm towards the optimal solution. The function was optimized for each well, providing optimal ROP values during real-time drilling. A fast drilling optimizer is crucial to automate and streamline the drilling process. This simultaneous optimization of ROP based on real-time data can be implemented during the process thereby increasing the efficiency of drilling as well as reducing the required drilling time.