Flow Diagnostics for Fractured Carbonate Reservoirs

The accurate static and dynamic characterisation of natural fractures is fundamental to successful hydrocarbon recovery from carbonate reservoirs because (i) fractures are ubiquitous in carbonate formations, (ii) fractures can cause early water breakthrough and low final recovery, and (iii) fractures are a primary source of uncertainty. The classical approach to production forecasting in support of reservoir development is to build one or more static models, translate selected models into dynamic models, and perform Black-Oil or compositional reservoir simulations that are tuned to production data. This is a time-consuming approach that often does not address the full range of geological uncertainties due to time limitations inherent to asset teams. Therefore, reservoir models tend to have limited shelve-life and require frequent modifications, leading to missed economic opportunities.

This project will be based at the Carbonate Reservoir Group (http://carbonates.hw.ac.uk/) at the Institute of Petroleum Engineering (IPE) and combines long-standing research efforts from IPE and SINTEF in Norway. Using the state-of-the-art Open Source Matlab Reservoir Simulation Toolbox MRST (http://www.sintef.no/Projectweb/MRST/), the project will extend and apply a novel technology termed "flow diagnostics" to improve classical reservoir characterisation and simulation workflows for fractured carbonate reservoirs. Flow diagnostics calculate and visualise key dynamic reservoir properties (e.g. drained and swept reservoir volumes, time-to-breakthrough, decline rates, sweet spots, well-allocation factors) directly on the grid blocks of multi-million cell static models in negligible time. Flow diagnostic tools hence allow us to quickly approximate the dynamic response of multiple static reservoir models to estimate their production behaviour without running computationally intensive full-physics reservoir simulations. Results from flow diagnostics tools can be used to screen, rank, and cluster static reservoir models based on their dynamic response, offering a robust and efficient way to identify production trends while accounting for uncertainty.