AI-Driven Zero Emission Autonomous Well Systems for Remote Areas Via Renewable Energy Integration

Abstract Remote wells using conventional fuel in off-grid areas cost dozens of times more than grid-connected wells. Integrating renewable energy into oil and gas operations can drastically cut costs, reduce emissions, and lower the consumption of produced hydrocarbons. This paper presents a comprehensive microgrid solution that utilizes renewables, energy storage, and optimization algorithms, integrated with remote wells for large-scale applications. Onsite energy consumption data, electric loads, and solar radiation are collected to estimate an appropriate mix of renewable energy. We build multiple models using machine learning algorithms to forecast electric loads and renewable energy generation. Various optimized operational strategies for energy storage and solar power have also been proposed to enhance resiliency and reduce costs. For some mature wells, we have optimized the wells’ production plans and cycles in accordance with renewable energy generation to achieve pure green electricity fulfillment and minimize operational costs. We have already tested these newly developed microgrid solutions on several oilfields of PetroChina. The field test results show that the off-grid microgrid, integrated with renewable energy and well operations, has succeeded. For single and multiple wells, they operate very well under the established strategies, which maximize the value of renewable energy, and the percentage of green electricity can reach over 90%. The optimized well production plan also demonstrates strong performance during the daytime, aligned with the capacity of solar power and energy storage based on forecasting models. The integrated solutions have also proven to be a viable substitute in harsh environments of high altitude and high temperature. These models and solutions developed with machine learning and optimization algorithms will be important components in the low-carbon energy transition. The experience acquired in well operation scenarios can be applied to more sophisticated upstream businesses to accelerate decarbonization.