Study On Carbon Dioxide Sequestration Characters In Depleted Natural Gas Hydrate Reservoirs

Global warming is becoming more and more serious,thus reduction of carbon dioxide emission cannot be delayed.Carbon dioxide sequestration is considered to be a time-sensitive method to reduce carbon dioxide in the atmosphere.Simultaneously,natural gas hydrate mining technology is developing rapidly,the disposal of depleted gas hydrate reservoirs has been put on the agenda.Focusing on the above two issues,this study proposed to store carbon dioxide in depleted natural gas hydrate reservoirs.In this research,the phase equilibrium conditions of carbon dioxide hydrate were predicted firstly,and based on this,carbon dioxide storage characteristics under the gas migration conditions in the depleted gas hydrate reservoir were investigated,and finally carbon dioxide storage with different nitrogen concentrations in depleted natural gas hydrate reservoirs was investigated in this study.By the way,magnetic resonance imaging(MRI)was employed to monitor the entire experimental process.Clathrate hydrate phase equilibrium is the key property to determine whether carbon dioxide hydrate can exist stably in the depleted gas hydrate reservoir.This is also the premise for determining whether the hydrate reservoir is suitable for carbon dioxide storage.Based on the machine learning(neural network algorithm and gradient boosted regression tree algorithm)algorithm,this study established hydrate phase equilibrium prediction models to predict hydrate phase equilibrium conditions with multiple components in the presence of various salts,organics or pure water.The prediction results showed that the prediction accuracy and success rate of the gradient boosted regression tree pressure model were higher than those of the neural network model,but the neural network model has better generalization.Specifically,for the gradient boosted regression tree model,the prediction accuracy of the gradient boosted regression tree pressure model was lower than that of the gradient boosted regression tree temperature model.Thus,the prediction of the temperature model was the best in this study.Furthermore,the accuracy of gradient boosted regression tree temperature model was satisfactory compared to the values in the literature.In addition,the phase equilibrium conditions of carbon dioxide hydrate could be predicted perfectly.Based on the research of clathrate hydrate phase equilibrium model,the carbon dioxide storage under gas migration conditions in depleted gas hydrate reservoirs was carried out using MRI.The experimental results showed that in water-saturated depleted gas hydrate reservoirs,initial water saturation was not the main controlling factor for CO2 hydrate formation during carbon dioxide gas flow;the pressure of carbon dioxide gas flow was the key control factor for carbon dioxide hydrate formation during carbon dioxide gas flow.The optimal flow pressure was 3.5 MPa in this study;compared with the flow process,the constant pressure process was more suitable for carbon dioxide hydrate formation.In order to further evaluate the effect of carbon dioxide storage in depleted gas hydrate reservoirs comprehensively,carbon dioxide and nitrogen gas mixture storage in depleted gas hydrate reservoirs was investigated;and the effect of different carbon dioxide and nitrogen concentrations and constant pressure on the carbon dioxide hydrate formation characters were analyzed.During the storage process,carbon dioxide hydrate was not uniformly formed in all regions at the same time.In the process of storage,3.2 MPa and 3.5 MPa had similar effects on carbon dioxide storage,and both are better than 3.0 MPa in the presence of nitrogen.Under pure carbon dioxide conditions,the higher the pressure,the greater the impact on carbon dioxide storage.The optimum pressure was 3.5 MPa.In general,nitrogen could accelerate the formation of carbon dioxide hydrates during the initial storage period;however,in the long run,pure carbon dioxide was more suitable for carbon dioxide sequestration.