Abstract
Gas lift is an extensively used artificial lift technique in the oil and gas industry. However, operators often shift to intermittent gas lift for low productivity index, reservoir pressure or high-water cuts, or due to the limited availability of gas (IGL). These wells have unique dynamic behavior and are generally not high producers, often lacking monitoring and surveillance, which leads to suboptimal production. Using a simulation-based method can incur excessive computing expenses that sometimes might not be justified. This paper presents an approach that combines data-based models, numerical methods, and machine learning-based algorithms to monitor and optimize IGL operations.
This process uses readily available operational data to estimate the well rates using API RP 11 V10 recommended methodologies and allows for careful monitoring of injection and shut-in phases in IGL wells while coordinating injection schedules between neighboring wells. The solution primarily consists of the following: an empirical equation with adaptable parameters that are designed to forecast the daily production rate of IGL wells, a numerical algorithm that proposes the optimum shut-in and injection period to maximize the production rate of an IGL well, and a genetic algorithm that minimizes the disturbance to a compressor that supplies compressed gas to the adjacent wells by synchronizing their injection schedule. The advanced genetic algorithm can also conduct binary-based scheduling to achieve a more streamlined control process that minimizes the total number of simultaneous injections.
The implementation of this solution has led to significant time savings, delivering a considerable improvement in operational efficiency and a noticeable increase in production. Real-time monitoring of the parameters provides numerous insights for IGL surveillance. Previously, the operator's approach was more reactive, and field engineers had to depend on ad-hoc hit and trial methods upon encountering such issues. The tests that could provide diagnostic signatures about the well performance used to occur occasionally, sometimes only once a year. These tests involved two or three pen recorders, tagging the fluid level using wireline tools, carrying out long cycle studies, bottom hole gradient surveys, etc. Now, using the real-time data monitoring and outputs from the solution, operators can troubleshoot and optimize the IGL operations remotely, which has led to proactive operations management, limiting the number of field visits. These operations have reduced the operational cost, HSE risk exposures, and carbon footprints.
This solution combines various technologies and methods to address challenges in the surveillance and optimization of intermittent gas-lift wells. The solution enables users to monitor important parameters, estimate daily production rates, determine optimum gas injection cycles, and schedule gas injection cycles to minimize gas injection overlap between the wells.
