Fluid discrimination is challenging for reservoir prediction, especially for tight sandstones with special petrophysical properties. In this paper, we first review the effective medium models that are widely used in seismic exploration and a variety of inversion methods and reservoir prediction strategies in reservoir prediction. Rock physics modeling takes an important role in reservoir prediction by linking petrophysical properties and elastic parameters. We also review the theoretical implications for different rock physics models that are based on the inclusion-based method, focusing specifically on the modeling workflow for conventional sand-shale reservoirs and two models for tight sandstones. The applicability of the conventional fluid substitution equations is analyzed in detail. Then, a new inclusion-based rock physics model for tight sandstones is proposed by considering the fluid pressure ratio between cracks and stiff pores. The proposed model helps to highlight the difference between different pores and present reasonable fluid information. In the application, a detailed prediction process for fluid discrimination is given, in which the Bayes posterior prediction framework is adopted to provide the maximum posterior probability solution and its posterior probability. Field data applications demonstrate the effectiveness of the proposed method.

流体判别对于储层预测具有挑战性，特别是对于具有特殊岩石物理特性的致密砂岩。在本文中，我们首先回顾了地震勘探中广泛使用的有效介质模型以及储层预测中的各种反演方法和储层预测策略。岩石物理建模通过将岩石物理性质和弹性参数联系起来，在储层预测中发挥着重要作用。我们还回顾了基于夹杂物方法的不同岩石物理模型的理论意义，特别关注常规砂页岩储层的建模工作流程和致密砂岩的两个模型。详细分析了常规流体置换方程的适用性。然后，考虑裂缝和硬孔隙之间的流体压力比，提出了一种新的基于包裹体的致密砂岩岩石物理模型。所提出的模型有助于突出不同孔隙之间的差异并提供合理的流体信息。在应用中，给出了流体判别的详细预测过程，其中采用贝叶斯后验预测框架来提供最大后验概率解及其后验概率。现场数据应用证明了所提出方法的有效性。给出了流体判别的详细预测过程，其中采用贝叶斯后验预测框架提供最大后验概率解及其后验概率。现场数据应用证明了所提出方法的有效性。给出了流体判别的详细预测过程，其中采用贝叶斯后验预测框架提供最大后验概率解及其后验概率。现场数据应用证明了所提出方法的有效性。