论文详情
非常规油气储层精细描述是实现目标油藏精细开发的关键, 而现有测井曲线自动分层算法很难适用于单砂体级非常规油气储层的精细描述。为此, 针对井间薄砂体储层结构划分问题, 提出了利用测井曲线构建地质描述不变特征以有效挖掘井间多曲线不变关联信息, 进而, 在不变特征的支撑下, 设计了分层能量约束的无监督集成聚类区域生长模型, 实现了具有显著多井一致性的目标储层精细自动分层。以自然伽马(GR)、井径(CAL)、自然电位(SP)、声波时差(AC)、浅侧向电阻率(LLS)、深侧向电阻率(LLD)、密度(DEN)共7条常规测井曲线为输入, 对大庆油田陆相坳陷盆地齐家凹陷工区多口井的实际测井数据进行了自动分层实验, 实验结果表明, 该方法比传统分层方法更加准确高效, 特别是对于工区内的薄层和薄互层储层, 利用该方法进行储层划分的结果与油田专业人员人工划分的结果吻合度较高, 能准确实现0.375 m以上的薄层分层, 且具有了很强的井间一致性。所提方法对于非常规油气藏精细描述具有一定的应用价值。
Fine-scale characterization of unconventional oil and gas reservoirs is one of the most important factors for enhancing the corresponding recovery capability.However, it is difficult to realize fine-scale reservoir characterization tasks using traditional automatic layering methods with well logging.To address this problem, a geophysical structure invariant feature of well-logging curves was proposed to reveal invariant correlation information among wells efficiently.Further, with the obtained invariant features, a novel unsupervised integrated clustering regional growth model with layered energy constraints was designed to perform automatic layering using well logging for fine-scale reservoir characterization.The proposed method was validated using real data from several wells of the Qi Jia Sag of the Continental Depression Basin in the Daqing Oilfield.Specifically, seven conventional loggings, including natural gamma, well diameter, and spontaneous potential, were selected as the inputs of the method.The experimental results illustrated that, compared with traditional methods, the proposed method can realize automatic layering above 0.35 m.Furthermore, the layering results agreed well with the manual results provided by professional experts, which can achieve strong inter-well consistency.The proposed layering method is valuable for the fine-scale characterization of unconventional oil and gas reservoirs.