基于厚度约束的匹配追踪算法识别石灰岩顶界地层

塔里木盆地塔中地区奥陶纪石灰岩地层与泥岩地层界面有两种特殊的地质现象:1泥岩沉积期局部发育石灰岩,与下伏石灰岩沉积形成假连续沉积;2石灰岩顶不整合面发育岩溶储层,储层波阻抗与上覆泥岩地层相近。因此在地震记录上可能会出现"穿时"的地震响应特征而掩盖真实的储层信息和地层信息。为此,提出了基于厚度约束的匹配追踪算法,以挖掘出更多、更准确的地震信息去表征所要关注的地质体。首先将原始地震数据分解成不同频率的多个原子;然后以地质目标体厚度所对应的频率作为约束阈值,剔除分辨率低于地质目标体的低频原子,从剩余原子中筛选出最能体现地质目标体的原子来做地震重构;最后利用重构地震数据完成储层预测和石灰岩地层分布预测。该方法在R井区的实际应用中,重构的地震剖面还原了"串珠"顶部的真实位置,结合地震反演后成功识别出泥岩底部局部沉积的石灰岩地层,为地震解释方案的制定和探井深度设计提供了参考依据。

Carbonate reservoir top boundary layer identification based on thickness-restricted matching pursuit algorithm

There are two peculiar geological phenomena in the interface between an Ordovician carbonate reservoir and clastic rock strata in the central Tarim Basin: limestone is developed locally in the period of shale deposition presenting paracontinuity with the underlying limestone deposition; karst carbonate reservoirs develop on the unconformity surface of limestone top layers, and the reservoir impedance is similar to that of those overlying shale layers. That's why "diachronism" may occur in the seismic record, which covers the actual reservoir and strata information. For this reason, the residual signal matching pursuit algorithm was proposed based on thickness restriction to explore much more accurate seismic information for the characterization of geological bodies. First, the original seismic data were decomposed into many atoms with different frequencies as the constraint to the thickness of the geological body. Then, the decomposed low-frequency atoms were rejected with their resolution ratio lower than the geological body, while the top-class atoms were screened from the left characterizing the geological body to make seismic signal reconstruction. Finally, the re-obtained seismic data were applied to predict the reservoirs and limestone strata distribution. From the re-constructed seismic profile in the practical application of this method in R well area, the real location of the top of "string beads" was revivified, and in combination with the seismic inversion, the carbonate reservoirs deposited partly at the bottom of mudstones were successfully identified. This method provides evidence for further seismic data interpretation planning and exploratory well depth designing.