Aref Khazaei - Ph.D. Candidate, Industrial Management, Science and Research Branch, Islamic Azad University, Tehran, Iran.
Reza Radfar - Professor, Industrial Management, Science and Research Branch, Islamic Azad University, Tehran, Iran.
Abbas Toloie Eshlaghy - Professor, Industrial Management, Islamic Azad University, Science and Research Branch, Tehran, Iran.

Iran is one of the largest oil and gas producers in the world. Intelligent manufacturing approaches can lead to better performance and lower costs of the well drilling process. One of the most critical issues during the drilling operation is the wellbore stability. Instability of wellbore can occur at different stages of a well life and inflict heavy financial and time damage on companies. A controllable factor can prevent these damages by selecting a proper drilling mud weight. This research presents a drilling mud weight estimator for Iranian wells using deep-learning techniques. Our Iranian data set only contains ۹۰۰ samples, but efficient deep-learning models usually need large amounts of data to obtain acceptable performance. Therefore, the samples of two data sets related to the United Kingdom and Norway fields are also used to extend our data set. Our final data set has contained more than half-million samples that have been compiled from ۱۳۲ wells of three fields. Our presented mud weight estimator is an artificial neural network with ۵ hidden layers and ۲۵۶ nodes in each layer that can estimate the mud weight for new wells and depths with the mean absolute error (MAE) of smaller than ±۰.۰۳۹ pound per gallon (ppg). In this research, the presented model is challenged in real-world conditions, and the results show that our model can be reliable and efficient in the real world.

Artificial Neural Networks, Deep Learning, Drilling Mud Weight, Mean Absolute Error, Smart Manufacturing