Intelligent Optimization of Drilling Parameters Based on Multiple Drilling Agents

Abstract

Abstract Drilling parameter optimization is one of the important ways to improve drilling efficiency and reduce drilling cost. Multi objective intelligent optimization technology has the characteristics of comprehensive consideration of factors and fast solution speed, and is suitable for drilling parameter optimization scenarios. Based on the drilling data of a vertical well in Tarim Oilfield, the rate of penetration (ROP) prediction agent, pipe string friction calculation agent and bottom hole cleanliness agent are established by using artificial intelligence or mechanism methods. Based on non-dominated sorting genetic algorithm II(NSGA-II), several established drilling agents are combined, and intelligent optimization model of drilling parameters is established. The evaluation function of the drilling scheme is set up to select the optimal drilling parameter optimization scheme from numerous Pareto solutions under the condition of ensuring operation safety, and the scheme is applied and verified in this well for Kangcun Formation and Jidik Formation with low ROP. According to the optimization results of drilling parameters, under the condition of ensuring the bottom hole cleaning and normal drilling, compared with the previous drilling plan, the ROP of Kangcun formation is increased from 3.43m/h to 7.72m/h; The ROP of Jidik formation is increased from 4.76m/h to 8.84m/h. The ROP of the two formations has been increased by 125.1% and 85.7% respectively, which can shorten the drilling cycle and reduce the drilling cost to a certain extent.