油气化探精查中的矿置异常特征与模式识别

FEATURES OF ANOMALIES RELATED TO RESERVOIR AND ANOMALY PATTERN RECOGNITION IN INTENSIVE OIL AND GAS GEOCHEMICAL EXPLORATION

油气化探精查是近年来开发完善的新技术,一般采用高密测网、轻烃检测新技术及非常规数值处理技术研究近地表烃类组成和分布特征,评价圈闭的含油气性或油气聚集的有利部位。化探精查矿置异常特征总体上属顶端晕类型,主要表现为复杂的块、环结合状,油田内外轻烃组构具有内重外轻的分异特征,烃类组分(特别是重组分)的浓集中心与深部油气聚集体具有较好的对应性。油气化探精查工作中应谨慎“圈环”,重视对异常特征的研究,并借助物化探多参数综合判别、分形理论、神经网络等评价技术进行含油气性预测,提高油气化探解决复杂地质问题的能力。以HBG油田为例,采用3层BP神经网络技术对化探精查异常进行模式识别,预测的有利部位得到钻探证实,表明在矿置异常识别和评价时,采用浓度异常提取与模式识别类异常评价技术相结合进行含油气性预测是可行且有效的。

Intensive oil and gas geochemical exploration is a new technique recently developed by former Hefei Institute of Petroleum Geochemical Exploration. Based on the research results of the near-surface hydrocarbon compositions and their distribution characteristics acquiring through the new light-hydrocar-bon detection method, the high-density sampling grid and the unconventional data-processing methods, this technique can evaluate the oil potential of a trap or the favorable areas of oil and gas accumulation. The hydrocarbon anomaly related to reservoir recognized by intensive geochemical exploration normally occurs as apical anomaly being composed of the combination of block anomaly and halo anomaly. The fabric of the light hydrocarbons ischaracterized by heavy inside and light outside an existing oilfield. The concentration center of the hydrocarbons, especially the heavy hydrocarbons, coincides well with deep oil and gas accumulation. In order to further improve the ability of the oil and gas geochemical exploration technique to resolve the complicated geological problems, it is proposed that, during intensive geochemical exploration, anomaly rings should be cautiously delineated while researches on anomaly characteristics should be paid more attention to. Evaluation techniques like multi-parameter geophysical and geochemical recognition, fractal theory and neural network should be used in predicting oil and gas potential. The favorable site predicted by using the 3-layer BP neural network technique to recognize the geochemical anomal patterns in HBG oilfield has been verified by the consequent drillings. This result shows that in the recognition and the evaluation of the geochemical anomaly related to reservoir, integrating hydrocarbon concentra-tion anomaly and pattern recognition technique is feasible and effective for oil and gas potential prediction.