Assessment of Cutting and Drilling Mud Heavy Metals and Organic Matter Contamination Using Limit Learning Regression Algorithm Technique of Artificial Intelligence in one of the Oil Fields of Southern Iran

The process of extraction and exploitation of oil and gas resources requires the cycle of production, sending, and recycling of drilling mud or drilling fluid, so achieving the right combination of drilling mud and its recycling is an essential and fundamental matter in the industrial oil and gas and also the environment.

Determining the level of contamination of heavy metals and organic matter in the drilling mud and drilling cuttings can be necessary so that intelligent methods to estimate these contaminants can be indirectly effective. This study tried to estimate the contamination rate of drilling cuttings, despite the formation parameters of 10 oil wells drilled at different depths (66 data sets), using the regression learning limit of an artificial neural network.

A total of 60 data sets were prepared to estimate the rate of change in the concentration of heavy metals, polycyclic aromatic hydrocarbons in the learning and testing process, and another six sets of data related to a well that was randomly selected and used in the artificial neural network validation process. Limit learning regression algorithm for ten heavy elements and ten aromatic compounds contaminating cutting and drilling mud on two different data sets in a drilling area in one of the oil fields in southern Iran was evaluated.

The results are suitable for estimating the contamination of drilling cuttings and subsequent environmental protection processes. Such processes of contamination and recycling of drilling mud will play an efficient role.