Petrophysical Properties Modeling Using the Geostatistical Approach: Case Study of Barito Basin, Indonesia

The Salemba Field is the largest productive oil field in Barito Basin. This field is located in the north-eastern area of Barito Basin. An improvement was required for the Field development, either from the geology, reservoir, or production aspect. The aim of this study is to build a reliable static reservoir model that can match the production history when it was simulated in the simulation reservoir. In this study, the targeted reservoirs were in the A and B zones, which have the biggest oil production. The A and B zones represent the most productive units in the synrift filling sequence in Tanjung Formation, which mainly comprises medium to coarse grained sandstone, well to moderately sorted volcanic litharenitic and feldspathic litharenitic sandstone, and volcanic pebble-dominated conglomerates. These reservoir zones are separated by a continuous shale break indicating differences in the depositional event. Modeling the distribution of A and B zones was guided by a detailed well-to-well correlation, tracer data and production history as a data constraint. The integrated interpretation of these data was then used to derive net sand maps which were used as trends to guide geomodel facies and properties modeling. After building the conceptual reservoir element distribution model, geostatistical analysis for facies parameter population was conducted. Probability distributions for net sand maps and vertical proportion curves for facies distribution variability were also constructed. The reservoir rock type is defined by Flow Zone Index (FZI) equation combined with geology facies interpretation. The rock type will guide to generate permeability transform and J-function equation to distribute permeability (k) and water saturation (Sw) in the model. Our experiment shows that the model has a good agreement with the geological interpretation and production data. In addition, the base case model represents the best estimation. This model has been analyzed using a dynamic model and has shown a good simulation.