【作者】 孙睿；

【导师】 段振豪；

【作者基本信息】 中国地质大学 ， 矿产普查与勘探， 2003， 博士

【摘要】 流体参与了大多数地质过程并在其中起着重要的作用。研究表明，H2O，CH4，CO2，N2和NaCl是许多富水地质流体体系最重要的组分。H2O-CH4-CO2-N2-NaCl体系（水-气-盐流体体系）在低温条件下（-20~+200℃）存在多种相平衡关系，其中重要的是水-气-盐流体体系的气液相平衡和水合物相平衡。研究H2O-CH4-CO2-N2-NaCl体系的低温相平衡性质对于研究地质流体在地球表面和地壳浅部的地质作用过程中所起的作用，对于指导石油天然气资源和天然气水合物资源的勘探和开发有着重要的意义。对于化学工业和石油天然气工业中的许多分离过程设计，了解H2O-CH4-CO2-N2-NaCl体系在低温条件下的气液相平衡性质也是必不可少的。由于实验数据的有限性，热力学模型是研究流体相平衡性质和其它热力学性质的最重要的工具。虽然迄今为止，前人建立了许多预测水-气-盐流体体系气液相平衡的热力学模型，但没有一个模型能够精确预测H2O-CO2和H2O-CH4体系在0-100℃的低温条件下的气液相平衡，没有一个模型能够精确预测低温条件下CO2和N2在NaCl水溶液中的溶解度。前人建立的甲烷水合物相平衡模型适用的温压范围较小，而且模型存在较大的理论缺陷。本研究的目标是建立精确预测水-气-盐体系低温气液相平衡和甲烷水合物相平衡的热力学模型。本研究的主要成果如下：（1）建立了一个精确预测CH4-H2O体系低温气液相平衡的状态方程。该方程由两部分组成——“参考流体”部分和“微扰”校正部分。借鉴Anderko和Pitzer（1993）处理“参考流体”的方法，采用“硬球”模型和“偶极硬球”模型描述“参考流体”的贡献；而“微扰”校正项则采用新的多参数维里型展开式处理。模型的经验参数通过拟合纯组分的PVT数据和混合体系的气液相平衡数据得到。通过与实验数据的对比，证实该方程能够精确预测CH4和H2O在273-623 K，0-3000 bar的温压范围内的PVT性质，平均误差约为0.3%。更重要的是，该方程能够精确预测CH4-H2O体系在273-383 K，0-1000 bar的温压范围内的气液相平衡，预测的液相组成平均误差小于4%，明显优于前人的模型。本模型的创新主要体现在：（1）建立了一个新的状态方程，将Anderko和Pitzer（1993）处理含水体系的方法推广到低温条件；（2）第一个能够精确预测CH4-H2O体系低温（0-100℃）气液相平衡的状态方程；（3）相对于大多数只能预测气液相平衡的热力学模型，该方程还能精确预测CH4-H2O体系的PVT性质。（2）建立了一个能够精确预测甲烷水合物在H2O-CH4-NaCl体系中的相平衡条件（生成压力）的热力学模型，模型适用于243-318 K，0-3000 bar的温度、压力范围，预测精度与实验数据的精度相当。本模型以Van der Waals-Platteeuw模型（1959）为基础，采用Exp-6势能函数计算Langmuir常数，采用Englezos-Bishnoi模型（1988）考虑NaCl对水的活度的影响。相对于前人的模型，本模型的主要创新和改进体现在：（1）采用Exp-6势能函数拟合Cao等（2001）的量子力学研究成果（从头计算法计算CH4-H2O之间的相互作用势能），<WP=7>势能参数随温度变化，而大多数水合物模型采用的Kihara势能函数并不能准确地描述CH4-H2O之间的相互作用势能（相互作用力）；（2）计算Langmuir常数时考虑多层（超过10层）水分子与甲烷分子的相互作用对Langmuir常数的影响，而前人的模型一般只考虑一层或三层水分子的影响；（3）采用第三章建立的高精度的状态方程计算甲烷的逸度系数和逸度，而前人模型采用的S-RK和P-R方程对纯甲烷逸度的预测误差较大。（3）建立了一个预测CO2和N2在纯水和氯化钠水溶液中的溶解度的热力学模型。模型参照Duan等（1992a）的处理方法，采用Pitzer特征相互作用模型处理液相，采用高精度的状态方程处理气相——CO2气相采用DMW-92方程（Duan等，1992b），N2气相采用DMW-96方程（Duan等，1996）。本模型能够精确预测CO2在273-533 K，0-2000bar，0-4.3 m NaCl的T-P-m范围内的溶解度；N2在273-623 K，0-600 bar，0-4 m NaCl的T-P-m范围内的溶解度。预测精度与实验数据的精度相当（7%）。对比前人的模型，本模型适用的温压范围更宽广，预测的精度更高。虽然本模型的参数仅从H2O-CO2-NaCl和H2O-N2-NaCl体系的实验数据拟合得到，参照Duan等（1992a）的处理方法，我们将本模型外延到复杂的含多种可溶盐的卤水体系。通过与实验数据的对比，证实本模型能够精确预测CO2和N2在海水中的溶解度。本研究的主要创新在于拓展了Duan-92（Duan等，1992a）模型，建立了精确预测低温条件下（T<200℃）CO2和N2溶解度的模型。更多还原

【Abstract】 Fluids take part in many geological processes and play important roles in them. Previous researches indicate that, water (H2O), methane (CH4), carbon dioxide (CO2), nitrogen (N2), and sodium Chloride (NaCl) are major components of many Geo-fluids. In H2O-CH4-CO2-N2-NaCl fluid systems (water-gases-salts fluid systems), several phase equilibria may occur at low temperatures (from -20 to 200℃). Among of them, Gas-Liquid Equilibrium and Gas hydrates-Gas-Liquid Equilibrium are the most important. To study phase equilibria of water-gases-salts fluid systems at low temperatures is important both to investigate the role of Geo-fluids play in the geological processes that take place in the surface and upper Crust of the Earth and to guide the prospecting and exploitation of oil and gases and natural gas hydrates resources. The knowledge on phase equilibria of water-gases-salts fluid systems at low temperatures is also essential for designing certain separation equipment in the chemical or oil-related industries.Due to the scarcity of the experimental data, thermodynamic models are the most important tools to study phase equilibria and other thermodynamic properties of fluids. Many thermo- dynamic models have been developed to predict gas-liquid equilibria of water-gases-salts fluid systems by previous researchers. However, none of them can predict gas-liquid equilibria of H2O-CH4 system and H2O-CO2 system at low temperatures between 0 and 100℃ accurately. None of them can predict carbon dioxide solubility and nitrogen solubility in pure water and aqueous NaCl solutions at temperatures below 200℃ accurately. Previous thermodynamic models for clathrate hydrates can’t predict phase equilibrium of methane hydrate at pressures more than 500 bar with sufficient accuracy. Furthermore, there still exist some theoretical flaws in these models. The aim of this study is to establish new thermodynamic models to predict gas-liquid equilibrium of water-gases-salts fluid systems and methane hydrate phase equilibrium at low temperatures. The following are the major achievements of this study.(1) A new equation of state (EOS) has been developed to predict Gas-Liquid Equilibrium and volumetric properties of CH4-H2O system at low temperatures. The equation of state consists of a reference part and a perturbation contribution. Following the approach of Anderko and Pitzer (1993), we adopt hard sphere model and dipolar hard sphere model to account for the contribution of "Reference Fluid". A new virial-type expansion was adopted to deal with perturbation contribution. The adjustable parameters of the equation of state were evaluated from experimental PVT data of pure water and methane and phase equilibrium data of CH4-H2O system. Comparison of the model predictions with experimental data demonstrates that this equation of state can predict PVT data of pure water and methane from 273 to 623 K, and from 0 to 3000 bar with an average error about 0.3%. Moreover, the equation of state can predict<WP=9>gas-liquid equilibrium of CH4-H2O system accurately from 273 to 373 K, and from 0 to 1000 bar. The deviation of this model is within experimental uncertainty. The main advantages of this model are: (a) The accuracy of this model is much better than previous models; (b) This model can predict volumetric properties of liquid water and aqueous solutions accurately, while most of models to deal with gas-liquid equilibria of fluids can’t predict volumetric properties of liquid water and aqueous solutions quantitatively. (2) A new model to predict phase equilibrium of methane hydrates has been established. This model is based on the Van der Waals-Platteeuw Model (1959). Langmuir constant, the key parameter of Van der Waals-Platteeuw model, is calculated from exp-6 potential. Englezos- Bishnoi model (1988) was adopted to account for the effect of NaCl on the activity of water. The improvements of this study on Van der Waals-Platteeuw model are: (a) this model adopted exp-6 potential to calculate Langmuir constant. The parameters of exp-6 poten更多还原