Togliatti State University (TSU) in collaboration with Samara State Technical University (SamSTU) have developed innovative solutions to improve the reliability of oil well equipment. The methodology, based on artificial intelligence, is successfully applied in one of the oil producing companies.
Scientists have created neural network models that analyze data on submersible electric motor (ESP) failures. The methodology allows predicting possible malfunctions and making decisions about repairing or replacing equipment. According to TSU Associate Professor Vladimir Romanov, ESPs often fail due to high loads and aggressive environments.
Experts have developed an integral reliability indicator that helps companies assess the risk of failure of each ESP. It takes into account the history of failures, the operating mode and the cost of repair. The data is entered into a special software system, which forms a list of devices by degree of risk.
This approach reduces the likelihood of accidents and downtime, and also reduces maintenance costs. Now companies can focus on repairing the most vulnerable devices, optimizing the operating mode of the rest. This is much more effective than traditional planned maintenance methods.
Read more on the topic:
AI created "smart molecules" to increase oil production at fields in Russia
More than 40% of companies in the fuel and energy sector use artificial intelligence technologies
