Cost Prediction of Sidetracking Operation in Gachsaran Oil Field Using Machine Learning

Sidetracking is one of the operations that may be performed during the drilling of oil and gas wells for various reasons, including directional drilling, failure of the fishing operation and well control. Estimating the sidetracking cost can help the driller and the operator in identifying the solution when the fishing is unsuccessful. It is economical to carry out the fishing operation until its cost is less than or at most equal to the cost of sidetracking, therefore predicting the cost of sidetracking is very important to save the cost and time of drilling. Many statistical analyzes have been used to predict sidetracking costs, but due to insufficient accuracy and the uniqueness of calculations for a region, these calculations have not been widely and comprehensively implemented. The sidetracking cost forecast is felt for Iran, that's why the purpose of this article is to estimate the sidetracking cost. Today, machine learning (ML) has played a significant role in estimating and predicting events, for this reason, machine learning, which is a sub-branch of artificial intelligence (AI), has been used to estimate the cost of sidetracking. For this purpose, first, all the data were collected and checked, and then different regression algorithms were taught with the input data of the cost rate and depths of the fishing and the output data of the sidetracking cost. Random Forest Regression and Decision Tree algorithms have shown the best performance for predicting sidetracking cost with an error of less than ۲%.

کلیدواژه ها:

Sidetracking operation ، Predicting ، Machine Learning ، Gachsaran oil field