A quantitative hydrocarbon favorability evaluation was desirable for enhancing oil and gas exploration of the first member of Dainan Formation at Yong'an area, Gaoyou Sag. After subdividing the evaluation units and screening the evalua-tion variables in terms of the current exploration results, a new model for quantitative hydrocarbon favorability evaluation was built by employing a logistic regression method to find the relationship between the hydrocarbon occurrence and geological variables. Based on the established model, multi-level hydrocarbon favorability was then characterized systematically and some exploration strategies were promoted combining the cognition of petroleum geology. The results show that the logistic re-gression model of hydrocarbon favorability built in application is reliable with model accuracy up to 89%. The deep part of the study area is deemed favorable for future hydrocarbon exploration. The logistic regression method presented here has good adaptability without heavy restrictions. This hydrocarbon favorability evaluation using logistic regression method can not only derive the probability of hydrocarbon occurrence directly, but also reveal the variations of hydrocarbon favorability convenient-ly, which would be useful for exploration planning.%开展含油气性定量评价成为有效推动高邮凹陷永安地区戴一段油气勘探的重要工作之一.根据已有勘探成果对含油气性评价单元进行细分并筛选出主要含油气性评价指标,利用Logistic回归方法拟合油气分布与评价指标之间的关系建立研究区含油气性定量评价模型;应用该模型对永安地区戴一段含油气性进行多层次的系统量化表征,并结合油气地质认识提供下一步的勘探建议.结果表明:所建立的含油气回归模型总体准确率为89%,具有较高的可靠性;研究区深部地区的勘探潜力值得探索,是未来拓展勘探的有利区域;基于Logistic回归的含油气性量化评价方法所受客观限制小、适应性较强,能够直接计算得到含油气概率值,并可以方便地刻画含油气性特征以指导油气勘探.
展开▼
