Reservoir characterization and recognition of the main factors controlling the reservoir quality has a significant share in presenting a real picture of the reservoir properties. In this study, by examining a set of reservoir well log data and their clustering based on mathematical and statistical principles, reservoir electrofacies are identified for the tight sands of Willespie Formation in five wells from the Whicher Range Field in Western Australia. These internally homogenous and externally heterogeneous clusters are truly a reflection of geological (sedimentary and diagenetic) and petrophysical properties of the reservoir. In this study, reservoir electrofacies (EF) were determined using clustering of well logs and along with porosity and permeability data in two individual sedimentary and petrophysical approaches. In the sedimentary approach, three EFs were identified that coincide with clean sands, dirty sands and shaly intervals. They are mostly related to the sedimentary texture and lithological properties indicating high reservoir heterogeneity. In the second approach, four EFs were determined in relation to the petrophysical properties of sandy packages within the reservoir. In addition, in order to find a good connection between EFs and production zones within sandy packages, five hydraulic flow units (HFUs) were introduced. The results show that low quality HFUs (i.e. A, B and to some extent C) in all petrophysical electrofacies are mainly related to the fine grained facies deposited in low energy systems. In contrast, high quality HFUs (i.e. D and E) are related to the medium to coarse and very coarse grained facies, showing different petrophysical properties, although they have generally low permeability and they are considered as tight sands. Thus, three sand types, depending on the severity of diagenetic effects, known as type 1 or very tight sands, type 2 or tight sands and type 3 or sub-tight sands were differentiated. In comparison, sandstones in type 3 with higher porosity and permeability have the best reservoir quality, whereas low reservoir quality sandstones in type 1 have been strongly affected by diagenetic effects (quartz and clay cementation). The results indicate a good agreement between core derived rock types and log facies. Using the methodology described in this study, it is possible to track reservoir flow units from well log responses.

Keywords