复杂油水层中流体对测井响应贡献小,常规测井流体识别难度大,核磁共振长短等待时间(双TW)测井及其时间域分析(TDA)方法在流体识别方面具有明显的优越性。但是,现有的TDA方法通常认为孔隙中的水完全极化且流体弛豫参数是已知的,这并不适用所有勘探区块,因此需要根据油水层情况对时间域分析方法加以改进。改进包括两个方面,一是针对水湿特征推导了核磁共振双TW观测模式下油水的弛豫机理,二是考虑到大孔隙中水的不完全极化问题,建立了回波串差与流体组分及其弛豫特性的关系。针对上述非线性问题,利用遗传算法实现了核磁共振双TW测井数据的非线性反演,计算得到冲洗带含油体积和含油饱和度。理论模型数值模拟结果表明,利用改进方法计算得到的油的横向、纵向弛豫时间以及含油体积与预设模型一致性好,比采用TDA分析方法得到的结果精度高。在鄂尔多斯盆地测井实例中,采用改进方法计算得到冲洗带含油体积和含油饱和度,计算结果与测试结果吻合,为有效识别储层含油性提供了准确信息。
Fluids in complex oil and water layers contribute little to logging response.Therefore,fluid identification by conventional logging is difficult.The nuclear magnetic resonance (NMR) double TW action and its time domain analysis (TDA) method proved to have good potential for fluid identification.However,the original TDA method usually assumes that water in the pores is completely polarized,and the fluid properties are known.However,these assumptions are not always applicable in exploration blocks,thus the TDA method needs to be improved.The proposed improvement includes two aspects.First,the relaxation mechanism of oil and water is deduced in the double TW observation mode of NMR according to the water-wet characteristics.Second,the incomplete polarization of water in large pores is accounted for,and the relationship between echo train difference and fluid composition and its relaxation characteristics is established.The fluid volume calculation is a nonlinear problem when the oil and water relaxation properties are unknown.In this study,a genetic algorithm was used to realize the nonlinear inversion of dual TW NMR logging data,and the oil volume and oil saturation of the flushing zone were calculated.The numerical simulation results showed that the improved TDA method was in good agreement with the model,demonstrating that the improved TDA method is more accurate than the original TDA method.Taking the Ordos Basin as a case study,it was shown that the improved TDA method could calculate the oil volume and oil saturation of the flushing zone,and the results were consistent with the oil testing results.It can be concluded that the proposed method can provide accurate information for effectively identifying fluids in reservoirs.
国家自然科学基金(41774144)、国家科技重大专项(2016ZX05050)共同资助。