Process‐based Modelling of Syn‐depositional Diagenesis

Carbonate rocks are of global importance both as economically significant reservoirs for oil and gas and as aquifers for water supply. On deposition, most carbonate sediments transmit fluids through matrix porosity, and thus they act as single-permeability systems. However, the reactivity of their constituent minerals makes them especially prone to chemical diagenesis – the alteration of minerals and of pore volume and geometry that occurs largely as a result of water:rock interaction. This can involve precipitation of pore-occluding cements or generation of secondary mouldic, linked- vug and channel porosity, and even cave formation (Moore, 1989). Diagenesis can strongly impact the performance of these rocks as reservoirs and aquifers by modifying the spatial distribution of primary (depositional) porosity and permeability that together determine key properties of storage and flow (Skalinski et al., 2013). Even minor diagenetic changes in porosity can result in major shifts in permeability at core scale, and significantly enhance both the heterogeneity and scale dependence of permeability (Whitaker & Smart, 2000). In addition, diagenesis can affect rock strength and ductability, wettability and absorption characteristics, and control their response to acidization and measurement of porosity via wireline logs (Roehl & Choquette, 2012). Developing a predictive understanding of carbonate diagenesis is thus of critical economic importance. Reservoir geologists are often called upon to predict diagenetic modification of rock quality across a carbonate platform based on core samples and downhole data for a limited number of sites, and seismic data that is rarely of sufficient resolution to image most diagenetic features. This relies on a workflow which involves inferring diagenetic process from product, and predicting distributions based on an understanding of controls on process (e.g. Dickson & Kenter, 2014). For example, petrographic and geochemical evidence may suggest a given phase of cement was precipitated close to the water table, and thus, by reference to models of the spatial configuration of modern water tables in carbonate environments we may predict a near-horizontal zone of cementation controlled by the paleo-position of sea-level. This approach has enabled geologists to effectively highlight potential mechanisms and conditions for carbonate diagenesis within the context of a geological history (e.g. Hollis, 1998; Hollis & Walkden 1996). However, it is only relatively recently that the application of process-based models has allowed us to quantitatively assess the relative merits and shortcomings of mechanisms hypothesized by these more inferential methods (e.g. Frazer et al., 2014). This contribution provides a review of the status of process-based forward modelling of syn- depositional diagenesis. The unstable mineralogy and high matrix porosity and permeability means that carbonate sediments are at their most vulnerable to diagenesis soon after deposition. We review the processes of syn-depositional diagenesis and the application of reactive transport models (RTMs) to understand controls on the rates, distribution and nature of alteration. The limitations of simulating reactions within a static sedimentological framework are then explored, and an alternative approach presented that involves explicit coupling of stratigraphic and diagenetic forward models. Applications of this approach to understanding fundamental controls on early meteoric digenesis within a sequences stratigraphic framework are explored using examples of simulations of modern, Cenozoic, Mesozoic and Paleozoic platform carbonates. Learnings for outcrop studies are compared to those for reservoirs where constraints on system behavior are less well known, but where forward stratigraphic–diagenetic modelling offers potential to generate fully 3D models to help reduce uncertainty in prediction of reservoir properties key to efficient extraction of hydrocarbons.