基于CO2气驱提高页岩气采收率渗流机理及竞争吸附机制的研究

Mechanisms of Percolation and Competitive Adsorption During CO2-based Enhanced Shale Gas Recovery.China holds huge potential for CO2 recovery and storage on its abundant shale gas reservoirs. CO2 injection can not only significantly boost shale gas recovery but also realize greenhouse gas storage in geologic formations, which presents great economic and environment benefits for China’s fossil energy self-supply and greenhouse gas emission reduction..Most of previous research only focused on pure CO2 or CH4 adsorption on shale rocks, or CO2 geological storage in shale gas reservoirs. Limited studies have been carried out on CO2 and CH4 competitive adsorption, desorption, and diffusion in the CO2 injection process. This proposed study aims at investigating fluid mass transfer at micro- and nanometer pore in shale gas reservoirs. With the consideration of CO2 and CH4 competitive adsorption and desorption on shale rocks, deduce the mechanism of competitive adsorption for CO2 and CH4 during CO2 gas drive process. Diffusion coefficients of CO2 and CH4 on CH4-saturated shale will be obtained through solving mass balance and diffusion equations. Then, such diffusion equation will be coupled with isothermal adsorption/desorption model for exploring the mechanisms of percolation..Furthermore, this project will study practical CO2 injection process through well-designed laboratory core displacement tests. Methodology on CO2–shale gas long-term exposure experiments will be established to investigate effects of geochemical reactions on petrophysical properties, adsorption/desorption, and mass transfer. Complex geochemical interactions of CO2–formation water–shale minerals will be systematically studied, and these effects will be incorporated into adjusting prediction strategy on shale gas production.

中国页岩气藏具有可观的CO2封存潜力，注入CO2既能实现地质封存又能提高页岩气采收率，对实现我国能源自给和CO2减排具有重大意义，具备良好的经济环境效益。现有研究多关注于单一气体在页岩中的吸附或地质封存，对真实页岩条件下CO2和CH4竞争吸附扩散的作用机制尚未明确。本项目拟通过建立溶蚀实验前后储层微观结构分析的实验方法，探究影响CO2和CH4吸附解吸的作用因素，推导CO2气驱过程中CO2和CH4的竞争吸附机制；通过耦合CO2和CH4扩散系数建立页岩气的解吸-扩散-渗流模型，探究CO2气驱提高页岩气采收率的渗流机理；拟将CO2注入页岩的实际过程与实验结合起来，考察CO2流动过程中与岩心的地球化学反应及其对孔渗参数和吸附传质的影响规律，深入研究CO2-地层水-页岩矿物之间的水岩反应机制，进而对由此造成的产能变化提出预测方案，为CO2的地下封存及提高页岩气采收率提供基础与理论指导。

我国页岩气藏具有可观的CO2封存潜力，注入CO2既能实现地质封存又能提高页岩气采收率，对实现我国能源自给和CO2减排具有重大意义，具备良好的经济环境效益。现有研究多关注于单一气体在页岩中的吸附或地质封存，对真实页岩条件下CO2和CH4竞争吸附扩散的作用机制尚未明确。本项目模拟岩芯吸附解吸装置得到的CO2和CH4竞争吸附实验数据构建出数学模型，建立CO2和CH4在页岩中的扩散方程，并在已有模型的基础上建立页岩气的吸附-扩散-渗流模型。同时，该项目引入了人工智能中的机器学习算法，结合甲烷气体（或二氧化碳）比例、含水量、TOC含量、温度和压力五个输入变量，对页岩体系中的甲烷和二氧化碳的过剩吸附量进行预测。针对甲烷过剩吸附量数据集，R2的大小顺序为：XGBoost梯度提升树回归（0.999和0.985）>多元非线性回归（0.994和0.981）>人工神经网络（0.929和0.873）>支持向量机回归（-0.328和-1.09）。XGBoost梯度提升树回归算法中，n_estimators取值为62，λ取值为1.2。针对二氧化碳过剩吸附数据集，R2的大小顺序为：XGBoost梯度提升树回归（0.999和0.966）>人工神经网络（0.991和0.955）>多元非线性回归（0.967和0.936）>支持向量机回归（0.621和0.540），XGBoost梯度提升树回归算法中，n_estimators取值为60，λ取值为1.0。机器学习算法适用范围广，且可以考虑多种地质和环境因素，本项目建立的页岩对甲烷和二氧化碳的过剩吸附量预测模型精确度较高，具有一定的可行性，可以为评估页岩地质中实际甲烷和二氧化碳吸附量提供一些见解。