Abstract:

Natural gas is a kind of relatively clean and low-carbon energy, the proportion of primary energy consumption in China continues to increase, and its accurate physical property prediction plays an important role in the process of natural gas gathering, transportation and utilization. By collecting the measured data of natural gas properties at different temperature and pressure in literature, we compared and analyzed the prediction accuracy of molecular dynamics simulations and commonly used empirical models to calculate the density and viscosity of natural gas, and clarified the best physical property prediction models suitable for natural gas with different compositions. The results show that molecular dynamics simulation has strong applicability in predicting the viscosity of natural gas, especially at high temperature and pressure, when the temperature is 444.4 K, the average absolute error is less than 5%. For natural gas density, it is more suitable to use the relatively mature empirical model, while several force field models using molecular dynamics simulation methods are not accurate in predicting it. In addition, the accuracy of the prediction model of natural gas density and viscosity is not only affected by the temperature and pressure range, but also by the composition of natural gas.

Key words: natural gas, mixtures, density, viscosity, molecular simulation

摘要：

天然气作为相对清洁、低碳的能源，在我国的一次能源消费占比持续增长，其物性的精准预测对天然气集输利用过程分析至关重要。通过收集文献中天然气在不同温度和压力下的物性实测数据，对比分析了分子动力学模拟和常用经验模型计算天然气密度和黏度的预测精度，明确了不同组成天然气适用的最佳物性预测模型。结果表明，分子动力学模拟在预测天然气黏度方面具有较强的适用性，尤其是在高温高压条件下，当温度为444.4 K时，平均相对误差小于5%；天然气密度则采用现有相对成熟的经验模型更合适，而采用分子动力学模拟方法的几种力场模型预测其精度均不高；天然气密度和黏度预测模型的预测精度，不仅受温度和压力范围的影响，还与天然气组成有关。

关键词: 天然气, 混合物, 密度, 黏度, 分子模拟