Heriot-Watt University (UK)
Flow diagnostics for fractured reservoirs: An innovative way to account for geological and geomechanical uncertainty in modern reservoir modelling and simulation workflows
Sebastian Geiger, Victoria Spooner, and Lesly Gutierrez Sosa
Fractured reservoirs are abundant in the subsurface and crucial to the provision of energy (e.g. oil, gas, geothermal) and groundwater, as well as for the storage of CO2. Yet, these reservoirs are very difficult to characterise, develop, and manage due to their complex and uncertain geology. Fractures can also open or close as the stress state in the reservoir changes during production, adding further complexities to the reservoir management. Modern reservoir simulation and uncertainty quantification workflows allow us to support the design of field development and management plans that account for geological uncertainties. However, simulating these reservoirs is not trivial and the time spent running full-physics reservoir simulations is valuable; it should be used for studying models that explore a realistic range of (geological) uncertainties and hence provide the most insight.
We have developed new flow diagnostics tools for naturally fractured reservoirs which sacrifice some physical detail in exchange for speed but still allow us to compute some of the essential dynamic behaviours in the reservoir, e.g. how effectively wells communicate with each other or which wells are at risk of early breakthrough. A particular innovative addition to our flow diagnostics is that we can also account for geomechanical effects in the reservoir; we can therefore quickly evaluate if and where fractures are likely to open or close, and how reservoir dynamics evolve as a consequence, due to production-induced stress changes.
Our new technology enables us to screen large numbers of geological models based on their approximate dynamic and geomechanical behaviours in a matter of minutes prior to commencing more time-consuming full-physics reservoir simulations. Using intuitive metrics we quantify the reservoir dynamics to cluster models and reduce the number of full-physics simulations required for robust reservoir forecasting without affecting the original range of geological and geomechanical uncertainties. Flow diagnostics hence offer a natural pre-processing step that complements modern coupled hydro-mechanical reservoir simulation, uncertainty quantification, and optimisation workflows, allowing us to spend valuable simulation time on the cases that yield a greatly improved understanding of reservoir performance and related uncertainties.
Professor Sebastian Geiger holds the Energi Simulation Chair and is the Director of the Institute of GeoEnergy Engineering at Heriot-Watt University. Previously, he was an assistant and associate professor at Heriot-Watt University, spent time as a visiting researcher at the Australia National University, Imperial College London, and Aramco Research Centre in Houston, and was a post-doctoral researcher at ETH Zurich. His research interests include the characterization, modelling, and simulation of naturally fractured reservoirs across all scales. He has authored or co-authored more than 170 technical papers. Sebastian holds a PhD degree from ETH Zurich and an MSc degree from Oregon State University. He is a member of Interpore, SPE, EAGE, AAPG, and AGU. He is an elected Fellow of the Energy Institute and received the 2017 Alfred Wegener Award from the EAGE for his pioneering research into carbonate reservoir modelling and simulation.
Imperial College London (UK)
Scale-dependence of Reaction Rates in Porous Media & Physical and Chemical Heterogeneity
Branko Bijeljic, Thomas Oliveira, Yousef Al-Khulaifi, Qingyang Lin, Martin Blunt
Reactive transport of solutes in porous media is encountered in many applications, such as contaminant transport and remediation in subsurface, acidization to enhance permeability in oil recovery, and packed bed reactors in chemical engineering. A principal outstanding problem in subsurface reactive transport is to determine the effective reaction rates from the pore-scale upwards. This is of key importance in highly heterogeneous natural porous media such as carbonate rock. Carbonates are known to have a significant portion of their pore space as micro-porosity, which may lead to a very wide distribution of local velocities, increasing transport heterogeneity that affects mixing and ultimately reaction. Hence, there is a need for a systematic methodology that can identify and quantify the impact of physical and chemical heterogeneity on the reaction rates. Moreover, in many problems additional complexities arising from coupling of multispecies transport and reaction reversibility need to be accurately addressed.
We develop a new methodology termed Screening Pore-Scale Imaging and Modelling (SPIM) that can be used to predict the fluid/solid reaction rates based on the systematic characterization of both physical and chemical heterogeneity in multi-mineral systems [1-3]. Physical heterogeneity of the rocks is classified in accordance with the velocity distributions obtained by numerical flow simulation on dry micro-CT images. Spatial distribution of chemical heterogeneity is also provided from the images. Performing and analyzing coreflooding CO2/brine/carbonate experiments, we show that mineral reaction rates are an order of magnitude lower than the corresponding batch rates due to mass transfer limitations. We introduce a new metrics quantifying coupled reactive transport behaviour, which describes proximity of reacted minerals to the fast channels and slow regions. Overall, a higher degree of physical (initial pore structure and associated velocity field) and/or chemical (intrinsic reaction rates and mineral distribution) heterogeneity promotes the preferential channelling effect, as opposed to uniform dissolution.
Furthermore, we simulate 3D multispecies fluid/fluid reversible reactive transport  in a micro-CT image of carbonate rock that entails spatially resolved information on connected micro-porosity. Direct numerical simulation of Darcy-Brinkman  and advection-diffusion transport equations are coupled to a general geochemical model . We demonstrate salient features of mixing and reaction arising as a result of intricate pore space heterogeneity. We show that evolution of rates of formation and consumption is species-dependent, and highly distinct in macro- and micro-porosity. Well-mixed regions result in asymptotic reaction rates. In contrast, incomplete mixing leads to transient and, for some species, even non-monotonic reaction rate behaviour.
Overall, we conclude that reactive behaviour is simultaneously influenced by pore space heterogeneity, multispecies reactive transport, and reaction reversibility. This means that for complex reversible reactions in heterogeneous porous media, species-specific behaviour needs to be examined for an accurate determination of reaction rates.
Dr. Branko Bijeljic is a Principal Research Fellow at the Department of Earth Science and Engineering, Imperial College London. His principal interest is in flow in porous media where his research focuses on multi-scale imaging and modeling of flow, transport and reaction in porous media, with application to geological carbon storage, contaminant transport and enhanced oil recovery. He has over 140 theoretical and experimental publications on various aspects of transport phenomena in porous media and porous rock. Branko is the recipient of the Procter & Gamble Award by the InterPore Society in 2016. He is a member of InterPore, AGU, SPE and EAGE. He is currently Associate Editor of Water Resources Research journal. Branko teaches Sustainable Energy Futures course at Imperial College.