基于叠前多参数敏感因子融合的浊积岩储层识别技术

Turbidite reservoir identification technology based on prestack multi-parameter sensitivity factor fusion

浊积岩油藏一直是济阳坳陷重要的勘探类型,经过多年勘探开发,目前面临的是“异质同像”型浊积岩,这类浊积岩的砂岩储层与非储层具有相近的速度、密度以及相似的地震波形特征,常规地震属性和叠后波阻抗识别难度大。由此建立了基于叠前多参数敏感因子融合的储层描述方法。该方法主要包含3个部分:①分析了影响横波估算精度的主要因素,建立了基于修正Xu-White模型的多矿物组分横波预测技术,提高了横波预测精度,为弹性参数的精准预测奠定基础;②提出了基于反射系数比的敏感因子定量评价方法,得到Murho、Lambrho和POIS这3个敏感弹性参数,应用3个弹性参数构建了敏感因子融合指数F,降低单参数的多解性,准确识别岩性;③开展叠前反演技术,反演敏感弹性参数,应用基于RGB三元色信息融合模型对3个敏感参数进行砂岩信息融合,实现岩性的精细预测。该技术在济阳坳陷坨71井区深水浊积岩油藏勘探中进行了应用,精细预测了研究区深水浊积扇体储层展布,预测结果与实钻井吻合度达到85%,提高了储层识别及描述精度。应用研究成果在该区描述砂体有利面积9.5 km^(2),部署探井和开发井10余口,其中5口井完钻投产后均获工业油流,预计新建产能2×10^(4)t。

Turbidite reservoirs have always been an important exploration type in the Jiyang depression.After years of exploration and development,the turbidites are mainly of the heterogeneous isomorphic type.The sandstone reservoirs of this type of turbidites have similar velocity,density,and seismic waveforms to those of non-reservoirs and thus are difficult to identify using conventional seismic attributes and poststack impedance.Therefore,a reservoir description method based on prestack multi-parameter sensitivity factor fusion was established.This method mainly included three steps.Firstly,major factors affecting the accuracy of shear wave estimation were analyzed,and then the multi-mineral-component shear wave prediction technology based on a modified xu-white model was established to improve the accuracy of shear wave prediction and lay a foundation for the accurate prediction of elastic parameters.Secondly,a quantitative evaluation method of sensitivity factors was proposed based on reflection coefficient ratios to obtain three sensitive elastic parameters,namely Murho,Lambrho,and POIS.The fusion index F of sensitivity factors was constructed by using the three elastic parameters.The purpose is to reduce the strong multiplicity of solutions of a single parameter and accurately identify rock properties.Thirdly,the prestack inversion technology was used for the inversion of sensitive elastic parameters.The three sensitivity parameters of sandstone information were fused using the fusion model of the RGB primary color information to realize a fine-scale prediction of lithology.This method was applied to the exploration of a deep-water turbidite reservoir around well-Tuo-71 in the Jiyang depression.The distribution of deep-water turbidite fan reservoirs in the study area was accurately predicted.The coincidence degree between the prediction results and the actual drilling reached 85%,indicating the improved accuracy of reservoir identification and description.The results of this study have contributed to an interpreted favorable sand body area of 9.5 km^(2)and the deployment of more than 10 exploration and development wells.Among these wells,five have yielded industrial oil flow after competition and being put into operation,and their new production capacity is expected to be 2×10^(4)t。