Fluvial reservoirs are important hydrocarbon reservoirs worldwide. It is the fluvial depositional characteristics that give rise to the complex reservoir architectures/geometries, spatial distributions/patterns, internal heterogeneities/petrophysical properties as well as the connectivity between flow units/channel sands that combine to give great uncertainty in characterising the effective reservoir properties.

Well testing, which measures the dynamic response of the reservoir, is potentially a very important tool for investigating these properties in fluvial reservoir systems. Through the integration of geoscience and engineering, the uncertainty resulting from reservoir description and well test analysis in such heterogeneous systems can be substantially reduced.

This paper reviews the latest techniques developed from the integration of geology and well testing for fluvial reservoir characterisation. Starting from the classification of fluvial systems, deterministic geological models, based on the two end members of fluvial systems (Meandering and Braided), have been mapped out.

A well test interpretation model for meandering channel reservoirs, the Pseudo-channel model has been distilled, which removes the assumption of the uniform formation thickness. Numerical solutions, termed Geotype curves have been derived. A new measure for the reservoir heterogeneity leading to an Effective flow interval has been developed, which shows a significant improvement from the conventional core analysis method. Geoskin as a geological phenomena has been identified and the technique quantifying its value with respect to geology has been developed. A wealth of geological data sets, derived from ancient (outcrops) and modern rivers, have been integrated. Fluvial Flow System Diagnostic Plots (FFSDP) have been developed, with which ten well tests of Tertiary channel sand reservoirs from the Gulf of Thailand have been evaluated.

Based on these studies, a Two-stage diagnostic procedure for well test analysis has been developed. A confident interpretation requires that the tested system should be clearly mapped out prior to the transient data analysis in ensuring the correct selection of the interpretation model. Then the meaningful reservoir parameters can be inverted. The final match to the tested data should be made by the Numerical solution from the defined reservoir system, rather than "force" an analytical model, which are mostly idealised, for the match.

An improved understanding on the scale and limits of the disciplines involved in reservoir characterisation is crucial for the integrated studies. This study gives insights into the integration and scaling of measurements as well as the need for improved geological, petrophysical and dynamic descriptions.

You can access this article if you purchase or spend a download.