Most Naturally Fractured Reservoirs (NFRs) form a major challenge for the hydrocarbon as well as geothermal industry, because their flow and mechanic behavior is strongly dependent on the geometry, topology and physical characteristics of fracture networks. Predicting fracture networks in the subsurface is difficult, as only very limited direct information can be acquired from wells, leaving the large volume between wells basically unsampled. The challenges posed by NFR are even larger in the case of carbonate rocks. Carbonate rocks typically experience early diagenesis becoming prone to fracturing already at an early stage and often experience dissolution and precipitation. One of the most obvious expressions of dissolution being stylolites, up to several tens of meters long pressure solution features common in carbonates and especially in chalks similar to the ones addressed by this study. Up to tens of meters large caves have also been reported, which formed as a consequence of hydrothermal fluids moving across a fractured rock column. At present, we only have limited tools to predict discontinuity networks in the subsurface and, thereby, predictions of flow and mechanic behaviour of carbonate NFR reservoirs are affected by large uncertainties. More specifically, these are caused by our poor ability to:
- predict the topology and reservoir-scale spatial distribution of discontinuity networks (stylolites, sedimentary bedding and fractures) in the absence and presence of fluids.
- include pre-existing geologically-realistic anisotropies related to distributed fractures specific faults in the model.
- perform flow simulations which are at the same time accurate enough to honour reservoir complexity and computationally inexpensive.
This results in lack of recognition of possibly important hydrocarbon plays, unsatisfactory reservoir sweep in existing systems and also led to substantial geo-engineering challenges in the past. Major dissolution (karsts) has caused the abandonment of wells in, for instance, the sub-salt of Brazil as well as in the Netherlands (California geothermal field). The peculiar geomechanic behaviour of chalks reservoir during depletion has caused major engineering problems in the well-known Ekofisk field (e.g. Zoback and Zinke, 2002) and well integrity disruption and consequent gas leakage above chalk reservoirs in the UK offshore (Elgin field).
The general strategy
The goal of CarbFrac is to develop models to predict the mechanic and flow behavior of carbonate (chalk) hydrocarbon and geothermal systems, honoring their complex geologic structure. This results in better production strategies, opening of new exploration opportunities and decreasing risk during production. CarbFrac builds on the cooperation and knowledge developed in the 2F2S program and adopts a multidisciplinary approach integrating the analysis of outcropping analogs, numerical modeling, and mechanic experiments through three overarching challenges.
Provides a set of rules to predict the type and spatial organization of discontinuity networks (inclusive of fractures, stylolites and bedding surfaces) in subsurface reservoirs. This is needed to populate mechanic and flow models of targeted reservoirs. We achieve this goal through an innovative integration of outcrop and modelling studies.
Tests the knowledge gathered in challenge 1 adopting innovative simulators to predict production scenarios in a specific block, which will act as the CarbFrac natural laboratory.
Predicts the response of reservoir and seal to changes in reservoir conditions as a consequence of depletion and/or stimulation (hydrofacturing). The experiments are conditioned by the geological and physical results obtained from challenges 1 and 2, and will be carried out under boundary conditions corresponding to the identified evolutionary stages.