摘要
根据30组烃类气体-原油体系最小混相压力细管实验数据(8组贫气驱,22组富气驱),运用修正共轭梯度和全局优化算法,回归拟合了一个新的统一的计算烃类气体-原油体系最小混相压力经验关联式。该关联式是一个关于油藏温度、原油中C7+相对分子质量、原油中挥发组分(CH4)、中间组分(CO2,H2S和C2—C6)和注入气中C2—C5的摩尔百分含量、C2—C5相对分子质量的函数,是一个统一的可以用来计算干气和富气驱油最小混相压力的经验关联式。对比了文献上3种(Kuo、Firoozabadi-Azia和Pedrood)常用的精度较高的烃类气体-原油体系最小混相压力经验关系式,结果表明,该经验关联式计算结果与文献上的实验数据更为接近,相对误差绝对值的平均值为6.11%,相对误差的绝对值最大为16.39%.为进一步验证新的经验关联式的预测能力,用其他6组最小混相压力组细管实验测试的数据进行检验,新的经验关联式预测精度也相对较高,相对误差绝对值的平均值为1.94%,相对误差的绝对值最大为4.10%.
Based on 30 groups of slim-tube experimental data of minimum miscibility pressure （MMP） in hydrocarbon gas-crude oil flood- ing system, including 8 groups of lean gas drive data and 22 groups of rich gas drive data, this paper makes regression and simulation of a new and universal empirical correlation expression for calculation of the hydrocarbon gas-crude oil flooding system MMP. This expression is a flmetion about reservoir temperature, C7＋ relative molecular mass of crude oil, volatile component in crude oil （CH4）, intermediate com- ponents （CO2, H2S and C2-C6）, the molar percentage and relative molecular mass of C2-C5, in injected gas, which can be used to universally calculate the MMP for dry gas and rich gas flooding. In this paper, the three empirical correlation expressions reported in the literature by Kuo, Firoozabadi-Azia and Pedrood that commonly used to calculate high precision of hydrocarbon gas-oil flooding system＇ s MMP are com- pared. The results show that the calculation resuhs are more closed to the experimental data from the literature： the average of relative er- ror＇ s absolute value is 6.11%, the maximum value is 16.39%. In order to further verify the new empirical expression＇ s prediction ability, other 6 groups of MMP slim-tube experimental data are applied, and the results are relatively high in accuracy, with the absolute value of relative error averages 1.94%, the maximum value is 4.10%.
出处
《新疆石油地质》
CAS
CSCD
北大核心
2013年第6期684-688,共5页
Xinjiang Petroleum Geology
基金
国家科技重大专项(2011ZX5016-005-2)
关键词
最小混相压力
烃类混相驱
优化算法
预测模型
minimum miscibility pressure （MMP）
hydrocarbon gas miscible flooding
optimization algorithm
prediction model
