谱分解技术在SZ36-1油田时移流体监测中的应用分析

Application of spectral decomposition to detect time-lapse fluid in SZ36-1 oilfield

在渤海SZ36-1油田的不断注水开采过程中,有的区域含水饱和度增加非常大,有的区域则压力下降太多而导致脱气。为了监测流体的变化,对该区域实施了时移地震应用。但在常规的时移地震属性差异体上,很难同时监测注水异常区域和脱气异常区域。利用该地区的储层物性建立了楔状砂体地质模型,正演模拟其时移地震响应差异,利用傅里叶变换分析地震反射波在不同厚度下注水和脱气引起的不同频谱响应特征。对时移地震数据运用谱分解技术,获得了一系列频率的调谐振幅数据体,通过在低频段分析注水引起的负异常和在高频段分析脱气引起的正异常,从而有效地划分了流体异常区域。这对于更好地利用谱分解技术进行储层流体识别,以及时移地震资料的差异分析解释提供了有效地指导。

In the course of water flooding in Pohai SZ36-1 oilfield,water saturation became very large in some regions whereas the pressure in other regions dropped significantly resulting in degassing.Time-lapse seismic analysis was implemented to monitor the fluid changes in the oilfield.The difference between conventional time-lapse seismic attribute volumes is usually not suitable for simultaneously monitoring anomalies from water flooding and degassing.Based on reservoir petrophysics observations,we established a wedge model to simulate the time-lapse seismic reflections.We analyzed the spectral responses of seismic reflections to different thickness in areas of water flooding and degassing by performing Fourier transform on the time-lapse seismic data.Then we used spectral decomposition technique to obtain a series of tuning frequency amplitude volume.We conclude that the negative anomalies in low-frequency end are caused by water flooding whereas the positive anomalies in high-frequency end are caused by degassing.