Carbonate reservoirs host significant hydrocarbon, groundwater, geothermal and CO2- and H2-storage resources. However, their complex depositional, tectonic and diagenetic histories make it challenging to efficiently characterize and predict their flow behavior. Here, we use a novel, rapid and efficient screening methodology that integrates experimental design, the construction of three-dimensional (3D) reservoir models via sketch-based methods and single-phase flow diagnostics to investigate the impact of geological heterogeneity on flow patterns and displacement in an ultra-deep (>7 km) carbonate reservoir in the north-central Tarim Basin, northwest China. Eight heterogeneities are investigated: (1) strike-slip fault zone width; (2) complexity of flower structure configuration; (3) continuity of fault core lithology; fault zone rock properties related to (4) karstification and (5) late, post-karstification cementation; (6) the occurrence of fault-perpendicular fracture corridors; (7) connectivity of fracture corridors to fault zones and (8) variability in host rock porosity and permeability. Fault zone width has the most significant impact on reservoir properties, with wider fault zones increasing effective horizontal permeability along fault zone strike. Fracture corridor occurrence and connectivity to the fault zone are the principal heterogeneities controlling effective permeability perpendicular to fault zone strike. Fault zone rock properties reflecting karstification and late cementation also significantly impact effective permeability in all directions. Other heterogeneities have little effect on effective permeability and well performance. However, simulated wells in negative flower structures and the main fault zone have higher productivities on average than simulated wells in positive flower structures and the host rock, similar to published production data from the ultra-deep reservoir. This study demonstrates the value of the screening methodology for assessing the effects of uncertainty in the interpretation of geological heterogeneities in complex carbonate reservoirs, in order to narrow the focus of future, more comprehensive reservoir simulations. The screening methodology is directly transferable to low-carbon energy applications in settings with sparse data.

Net-transgressive, shallow-marine sandstone reservoirs overlain by thick mudstone seals are prime candidates for storage of CO2, H2 and thermal energy. Although these reservoirs have high net-to-gross ratios, analogous outcrops demonstrate a wide range of sedimentological heterogeneities that are sampled only sparsely or at low resolution in subsurface data. We use a combination of outcrop data, sketch-based reservoir modelling and flow diagnostics to assess the impact of sedimentological heterogeneities on subsurface storage.

The Cliff House Sandstone outcrop example comprises wave-dominated shoreface sandstones arranged in aggradationally-to-retrogradationally stacked parasequences, which overlie and pass up depositional dip into mudstone-dominated coastal plain, lagoonal and tidal flat deposits that encase channelized tidal and tidally influenced fluvial sandbodies. Reservoir models of this outcrop example demonstrate that effective horizontal permeability, flow patterns and displacement, and stratigraphic trapping potential are controlled by: (1) the packaging of shoreface sandstones into laterally extensive parasequences bounded by offshore mudstones; (2) the spatial distribution, connectivity and permeability of channelized sandbodies; and (3) the localized connections between channelized sandbodies and shoreface sandstones. The last two parameters are likely to be poorly constrained in subsurface seismic and well data, and their potential effects require evaluation in reservoir modelling studies.

We describe an efficient and novel method to characterise multiscale geological heterogeneity and its effects on fluid flow using an open-source, sketch-based modelling and flow diagnostics tool (Rapid Reservoir Modelling. RRM). The method has three aims: (1) to generate a nested hierarchy of geologically accurate models of sedimentological heterogeneity at different scales; (2) to determine the representative elementary volume (REV) for each heterogeneity style, in order to calculate effective properties for larger scale models; and (3) to evaluate the impact of sedimentological heterogeneity on fluid flow and trapping in the hierarchy of geological models. The sketch-based modelling approach enables the construction of multiple geometrically accurate geological models and allows us to analyse them quickly using flow diagnostics and simulation tools. We illustrate this approach using examples from Triassic fluvial sandstones of the UK (Bunter Sandstone and Sherwood Sandstone), which host groundwater and geothermal resources and are targets for carbon capture and storage (CCS).