基于三参数匹配追踪的油气检测技术与应用

匹配追踪时频谱分解技术具有较高的时频分辨率。近年来, 基于匹配追踪方法的储层含油气性检测技术得到广泛应用, 但传统的匹配追踪技术因计算效率低和时频原子分解精度不够等因素影响了流体识别的效果。为此, 研究了一种三参数动态扫描的快速匹配追踪方法。该方法以Morlet小波为时频原子, 引入三瞬属性作为初始值, 从而缩小参数扫描范围, 只对尺度、相位和频率进行局部扫描, 在瞬时相位的局部单调区间内直接获得最佳时移参数, 减少搜索参数的同时提高了分解的稳定性。单道地震信号试验表明, 该方法能够有效提高匹配追踪的计算效率和分解精度。该方法应用于CD油田沙一段储层油气检测的结果表明, 利用该方法提取的吸收衰减属性可以有效地区分油气层与水层和干层, 预测的油气区展布符合沙一段储集体发育规律和油气成藏规律。

With the continuous advancement of oil and gas exploration, fluid identification of subtle reservoirs is vital for oil field development.Because the matching pursuit method has a higher resolution than other time-frequency analysis methods, fluid identification techniques based on this method have been broadly developed in recent years.However, the fluid identification results are affected by problems such as the low-computational efficiency of the traditional matching pursuit method and inaccuracy of time-frequency atom decomposition.In this study, a high-accuracy and fast matching pursuit method is proposed to improve computational cost and precision.This method takes the Morlet wavelet as a time-frequency atom and introduces three instantaneous properties to reduce the range of scanning parameters, which locally scans the scale, phase, and frequency independently.The improved matching pursuit method directly obtains the best time-shift parameters in the local monotonic interval of the instantaneous phase.Single trace tests validate the computational cost and decomposition accuracy of this method.The application of field data at the Sha 1 formation in the CD Oilfield shows that the absorption and attenuation properties extracted by this method can effectively distinguish between oil and gas, water, and dry layers.The predicted area is consistent with the oil and gas reservoir distribution and hydrocarbon accumulation in the Sha 1 formation.