Application of fuzzy logic for stuck pipe prevention purposes on oil wells drilling operations

Resumo

The drilling of oil wells represents one of the most challenging activities in the oil and gas industry. In
an environment surrounded by uncertainties, one of the most common problems is the stuck pipe, which results in
increases in the cost of building the well and in non-productive time. Considering that factors such as safety and
costs are directly related to the decisions made during operations, the adoption of tools that help the operator to
make the best decisions is highly desirable. In addition, the large amount of information available to the operator
requires its due analysis in real time in order to offer the diagnosis of anomalies. Artificial intelligence techniques
such as artificial neural networks, decision trees and support vector machines have been used to detect stuck pipe.
These strategies, however, have efficiency conditioned to the specific characteristics of the available data set.
Alternatively, this work presents an algorithm for preventing stuck pipe using fuzzy logic. Throughout the paper,
the stuck pipe problem and its impacts will be presented, the characteristics of fuzzy logic and its application to
the case under study, the results obtained from the application to an oil well data and, finally, the job conclusions.