摘要
混源油人工配比模拟实验是目前最常使用的定量判识混源油气烃源组成比例的方法。首先总结其一般步骤,提出了该方法应用过程中存在的4个问题:端元油成熟度差异导致生物标志物参数非线性变化、高演化或强降解条件下混源油常规生物标志物参数失效、复杂地质条件下端元样品难以获得、多元混源条件下实验流程及地球化学指标变化规律复杂。在此基础上,结合最新研究进展,提出了针对这4个问题的研究对策:生物标志物定量法、筛选特殊生物标志物、多元数理统计学方法、采用计算机软件协助处理数据法。最后,重点介绍了使用多元数理统计学方法定量判析混源比例的基本方法和原理,指出应用该方法,可以在不知道端元样品、不进行混源实验的情况下,直接计算混源油气中端元的数量、组成和贡献比例,该方法在研究复杂地质条件下混源油气比例时具有特别作用。
Artificial oil mixing experiment is frequently adopted in quantifying mixed oils at present.The author first summarizes the common process of the experiment,and then puts forward 4 questions during experimentation: nonlinear variation of biomarker ratios caused by different maturity of end member oils,invalidation of conventional biomarker parameters under the condition of high maturity or strong degradation degree,being hard to get end member samples in complex geological conditions,complexity of geochemical parameter's changing rules and experimental processes when there are many end member samples.Afterward,solutions to the above mentioned questions are proposed: quantitative analysis of biomarkers,selection of special biomarkers,multivariate data analysis,and computer software data analysis.Finally,a new multivariate data analysis method for quantifying mixed oils is introduced,through which,the numbers,composition and proportion of end member samples could be calculated even without knowing the end member samples or adopting the artificial oil mixing experiment.This method is particularly useful to quantify mixed proportion under complicated geological conditions.
出处
《石油实验地质》
CAS
CSCD
北大核心
2010年第4期377-381,386,共6页
Petroleum Geology & Experiment
基金
国家重点基础研究发展计划(973计划)项目(2005CB422102)
中国石化科技部项目(P07018)资助
关键词
混源油
混源实验
混源比例
多元数理统计学分析
交替最小二乘算法
mixed oils
artificial oil mixing experiment
mixed proportion
multivariate data analysis
alternating least squares