Knowledge of initial fluids saturation has great importance in hydrocarbon reservoir analysis and modelling. Distribution of initial water saturation (Swi) in 3D models dictates the original oil in place (STOIIP), which consequently influences reserve estimation and dynamic modelling. Calculation of initial water saturation in heterogeneous carbonate reservoirs always is a challenging task, because these reservoirs have complex depositional and diagenetic history with a complex pore network. This paper aims to model the initial water saturation in a pore facies framework, in a heterogeneous carbonate reservoir. Petrographic studies were accomplished to define depositional facies, diagenetic features and pore types. Accordingly, isolated pores are dominant in the upper parts, while the lower intervals contain more interconnected interparticle pore types. Generally, in the upper and middle parts of the reservoir, diagenetic alterations such as cementation and compaction decreased the primary reservoir potential. However, in the lower interval, which mainly includes high-energy shoal facies, high reservoir quality was formed by primary interparticle pores and secondary dissolution moulds and vugs. Using huge number of primary drainage mercury injection capillary pressure tests, we evaluate the ability of FZI, r35Winland, r35Pittman, FZI* and Lucia’s petrophysical classes in definition of rock types. Results show that recently introduced rock typing method is an efficient way to classify samples into petrophysical rock types with same pore characteristics. Moreover, as in this study MICP data were available from every one meter of reservoir interval, results show that using FZI* method much more representative sample can be selected for SCAL laboratory tests, in case of limitation in number of SCAL tests samples. Integration of petrographic analyses with routine (RCAL) and special (SCAL) core data resulted in recognition of four pore facies in the studied reservoir. Finally, in order to model initial water saturation, capillary pressure data were averaged in each pore facies which was defined by FZI* method and using a nonlinear curve fitting approach, fitting parameters (M and C) were extracted. Finally, relationship between fitting parameters and porosity in core samples was used to model initial water saturation in wells and between wells. As permeability prediction and reservoir rock typing are challenging tasks, findings of this study help to model initial water saturation using log-derived porosity.

初始流体饱和度的知识在油气藏分析和建模中非常重要。3D模型中的初始含水饱和度（Swi）的分布决定了原位原油（STOIIP），因此会影响储量估算和动态建模。非均质碳酸盐岩储层的初始含水饱和度的计算始终是一项艰巨的任务，因为这些储层具有复杂的沉积和成岩史，具有复杂的孔隙网络。本文旨在模拟非均质碳酸盐岩储层孔隙相框架中的初始含水饱和度。通过岩相学研究来确定沉积相，成岩特征和孔隙类型。因此，孤立的毛孔在上部占主导地位，而较低的间隔包含更多的相互连接的粒子间孔类型。通常，在储层的上部和中部，成岩作用如胶结作用和压实作用降低了主要储层的潜力。然而，在较低的时间段（主要包括高能浅滩相）中，较高的储层质量是由初级颗粒间孔隙和次级溶解霉菌和孔洞形成的。使用大量的一次排水汞注入毛细管压力测试，我们评估了岩石类型定义中FZI，r35Winland，r35Pittman，FZI *和Lucia的岩石物理分类的能力。结果表明，最近引入的岩石分型方法是将样品分类为具有相同孔隙特征的岩石物理类型的有效方法。而且，正如本研究一样，每隔一米的储层间隔即可获得MICP数据，结果表明，在SCAL测试样品数量有限的情况下，使用FZI *方法可以为SCAL实验室测试选择更多代表性样品。岩石学分析与常规（RCAL）和特殊（SCAL）岩心数据的集成导致对所研究储层中的四个孔隙相的识别。最后，为了模拟初始含水饱和度，将每个孔隙相的毛细管压力数据取平均值，并用FZI *方法定义，并使用非线性曲线拟合方法，提取拟合参数（M和C）。最后，使用拟合参数和岩心样品中孔隙率之间的关系来模拟井和井之间的初始水饱和度。