The heterogeneity characteristics of sandstone reservoirs lead to variations in composition, thickness, and porosity, making them more complex and uneven. This complexity presents a significant challenge in the development of sandstone reservoirs, which is further compounded by contamination factors. To characterize reservoirs comprehensively, the Hydraulic Flow Unit (HFU) method is commonly used. This method can describe reservoir characteristics and determine rock typing distribution based on HFU calculations. In this study, the HFU method is combined with petrophysical parameters and the calculation of oil reserves in the reservoir with the code “MZG.” The HFU calculations revealed four rock types, namely RT-1, RT-2, RT-3, and RT-4, in order of decreasing quality from RT-1 to RT-4. Furthermore, this HFU method is integrated with the Multi-Resolution Graph-Based Clustering (MRGC) approach, which serves to spread the available data to wells with limited data, thereby obtaining results that correlate well based on the combination of HFU-derived rock typing and petrophysical parameters. Overall, RT-1 and RT-2 show very good reservoir quality, while RT-3 and RT-4 represent reservoirs with poor quality. This classification has been validated using both qualitative and quantitative approaches based on petrophysical parameters. The correlation between predicted permeability using the GFU (Generalized Flow Unit) method and core data permeability was performed, yielding correlation coefficients of 0.9482 and 0.458, respectively. Oil reserve calculations were carried out based on the geological structure obtained from the structure map derived from seismic picking. The structure reveals an anticline, which was further analyzed considering the impact of petrophysical parameters. Petrophysical calculations using a trial-and-error approach on tortuosity factor (a), cementation exponent (m), and water saturation exponent (n) based on SCAL and RCAL data showed significant results. With parameters a, m, and n set to 1, 2, and 2, respectively, there was an increase in reserves of 31% in tank 1 and 43% in tank 2. This increase was achieved by optimizing higher parameter values based on the SCAL and RCAL data available.