Keele Research Repository

The recoverable proportion of known mobile resources from fluvial multi-storey sandbodies (MSBs) reservoirs is relatively low. The low recovery proportion can be attributed to a lack of a three-dimensional understanding of the reservoir architecture, a lack of consideration of meso-scale heterogeneity, and a lack of geological realism within reservoir models. This thesis considers the architectural nature of unconfined and confined MSBs. The work will: 1) develop an understanding of the main controls upon preservation and deposition of architectural elements within unconfined and confined MSBs and determine key diagnostic features attributed to them. Finally, this work will determine best-practice stochastic reservoir modelling practices for the simulation of unconfined and confined MSBs.
The works presented here use terrestrial photogrammetry, sedimentary logging and palaeohydrodynamic reconstructions from unconfined and confined multi-storey sandbodies to determine key architectural diagnostic criteria. Multi-point statistics, object-based models and sequential indicator simulation are used to determine the most appropriate algorithm to represent the two systems.
Results indicate that unconfined MSBs show complex barform architectures with compound bar formation. This is highlighted by the presence of upstream accretion elements, developed by variable discharge. In confined MSBs large water depths relative to sediment supply indicate that local accommodation is not being filled and that compound bar formation and upstream accretion elements are impeded. It is therefore proposed that unconfined MSBs can be identified by the presense of complex barform geometries and upstream accretion elements, whereas confined MSBs, due to their larger water depths can be identified by the lack of upstream accretion elements and a lack of correlation between maximum flow depth reconstructions and the thickness of bars. Reservoir model tests suggest that the best method of representing these systems is through the use of three-dimensional multi-point statistics simulations, which provide the most realistic statistical and visual representation of the fluvial environment.