Risk Assessment Model for Gas Storage Injection and Production Wells based on Random Forest

Abstract The wellbore of gas storage reservoir will degrade its strength or reliability over time due to aging, corrosion, wear, vibration, underground complex environment, etc. Under the high-speed strong injection and production wells, the wellbore will have defects such as corrosion or cracks, which will make its integrity fail, and bring risks to the operation of injection and production wells in the gas storage reservoirs, so in this paper, firstly, we establish the limit state equation of the failure of the injection and production wells, and then we use the Monte Carlo method to obtain a large amount of data of the failure probability of injection and production wells. This paper firstly establishes the limit state equation of injection and production well failure, then uses the Monte Carlo method to calculate the limit state equation of injection and production well failure to obtain a large amount of data on the probability of failure of injection and production wells, and at the same time, selects the 16 factors of the limit state equation of injection and production well failure as the indicators of the injection and production well risk assessment, and then combines the concept of fuzzy to quantify the level of risk, and then establishes a risk assessment database. An example application of the model was carried out on a gas reservoir injection and production well, and it was found that the risk level of the reservoir injection and production well was 6, which was very dangerous and should be immediately reported to the oilfield company for repairing the well or implementing effective risk mitigation measures. The model is able to assess the risk of gas storage reservoir injection and production wells, which facilitates the managers to quickly grasp the risk status of gas storage reservoir injection and production wells, and provides a decision-making basis for the safe operation of gas storage reservoirs.