- Showing results for:
- Reset all filters
Journal articleCollini H, Jackson MD, 2022,
Relationship between zeta potential and wettability in porous media: insights from a simple bundle of capillary tubes model, Journal of Colloid and Interface Science, Vol: 608, Pages: 605-621, ISSN: 0021-9797
Hypothesis & MotivationExperimental data suggest a relationship between the macroscopic zeta potential measured on intact rock samples and the sample wettability. However, there is no pore-scale model to quantify this relationship.MethodsWe consider the simplest representation of a rock pore space: a bundle of capillary tubes of varying size. Equations describing mass and charge transfer through a single capillary are derived and the macroscopic zeta potential and wettability determined by integrating over capillaries. Model predictions are tested against measured data yielding a good match.FindingsMixed- and oil-wet models return a macro-scale zeta potential that is a combination of the micro-scale zeta potential of mineral-brine and oil-brine interfaces and the relationship between macro-scale zeta potential and water saturation exhibits hysteresis. The model predicts a similar relationship between zeta potential and wettability to that observed in experimental data but does not provide a perfect match. Fitting the model to experimental data allows the oil-brine zeta potential to be estimated at conditions where it cannot be measured directly. Results suggest that positive values of the oil-brine zeta potential may be more common than previously thought with implications for surface complexation models and the design of controlled salinity waterflooding of oil reservoirs.
Journal articleSilva VLS, Salinas P, Jackson MD, et al., 2021,
A machine learning approach to accelerate convergence of the nonlinear solver in multiphase flow problems is presented here. The approach dynamically controls an acceleration method based on numerical relaxation. It is demonstrated in a Picard iterative solver but is applicable to other types of nonlinear solvers. The aim of the machine learning acceleration is to reduce the computational cost of the nonlinear solver by adjusting to the complexity/physics of the system. Using dimensionless parameters to train and control the machine learning enables the use of a simple two-dimensional layered reservoir for training, while also exploring a wide range of the parameter space. Hence, the training process is simplified and it does not need to be rerun when the machine learning acceleration is applied to other reservoir models. We show that the method can significantly reduce the number of nonlinear iterations without compromising the simulation results, including models that are considerably more complex than the training case.
Journal articleTitus Z, Heaney C, Jacquemyn C, et al., 2021,
Surface-based modelling provides a computationally efficient approach for generating geometrically realistic representations of heterogeneity in reservoir models. However, conditioning Surface-Based Geological Models (SBGMs) to well data can be challenging because it is an ill-posed inverse problem with spatially distributed parameters. To aid fast and efficient conditioning, we use here SBGMs that model geometries using parametric, grid-free surfaces that require few parameters to represent even realistic geological architectures. A neural network is trained to learn the underlying process of generating SBGMs by learning the relationship between the parametrized SBGM inputs and the resulting facies identified at well locations. To condition the SBGM to these observed data, inverse modelling of the SBGM inputs is achieved by replacing the forward model with the pre-trained neural network and optimizing the network inputs using the back-propagation technique applied in training the neural network. An analysis of the uncertainties associated with the conditioned realisations demonstrates the applicability of the approach for evaluating spatial variations in geological heterogeneity away from control data in reservoir modelling. This approach for generating geologically plausible models that are calibrated with observed well data could also be extended to other geological modelling techniques such as object- and process-based modelling.
Journal articleKampitsis AE, Kostorz WJ, Muggeridge AH, et al., 2021,
Journal articleSalinas P, Regnier G, Jacquemyn C, et al., 2021,
Modelling geothermal reservoirs is challenging due to the large domain and wide range of length- and time-scales of interest. Attempting to represent all scales using a fixed computational mesh can be very computationally expensive. Application of dynamic mesh optimisation in other fields of computational fluid dynamics has revolutionised the accuracy and cost of numerical simulations. Here we present a new approach for modelling geothermal reservoirs based on unstructured meshes with dynamic mesh optimisation. The resolution of the mesh varies during a simulation, to minimize an error metric for solution fields of interest such as temperature and pressure. Efficient application of dynamic mesh optimisation in complex subsurface reservoirs requires a new approach to represent geologic heterogeneity and we use parametric spline surfaces to represent key geological features such as faults and lithology boundaries. The resulting 3D surface-based models are mesh free; a mesh is created only when required for numerical computations. Dynamic mesh optimisation preserves the surfaces and hence geologic heterogeneity. The governing equations are discretised using a double control volume finite element method that ensures heat and mass are conserved and provides robust solutions on distorted meshes. We apply the new method to a series of test cases that model sedimentary geothermal reservoirs. We demonstrate that dynamic mesh optimisation yields significant performance gains, reducing run times by up to 8 times whilst capturing flow and heat transport with the same accuracy as fixed meshes.
Journal articleJacquemyn C, Pataki MEH, Hampson GJ, et al., 2021,
Geological modelling is widely used to predict resource potential in subsurface reservoirs. However, modelling is often slow, requires use of mathematical methods that are unfamiliar to many geoscientists, and is implemented in expert software. We demonstrate here an alternative approach using sketch-based interface and modelling, which allows rapid creation of complex three-dimensional (3D) models from 2D sketches. Sketches, either on vertical cross-sections or in map-view, are converted to 3D surfaces that outline geological interpretations. We propose a suite of geological operators that handle interactions between the surfaces to form a geologically realistic 3D model. These operators deliver the flexibility to sketch a geological model in any order and provide an intuitive framework for geoscientists to rapidly create 3D models. Two case studies are presented, demonstrating scenarios in which different approaches to model sketching are used depending on the geological setting and available data. These case studies show the strengths of sketching with geological operators. Sketched 3D models can be queried visually or quantitatively to provide insights into heterogeneity distribution, facies connectivity or dynamic model behaviour; this information cannot be obtained by sketching in 2D or on paper.
Journal articleKostorz WJ, Muggeridge AH, Jackson MD, 2021,
Journal articleAlarouj M, Ijioma A, Graham MT, et al., 2021,
We report a new, open-source, MATLAB-based 3D code for numerically simulating the self-potential (SP) in subsurface reservoirs. The code works as a post-processor, using outputs from existing reservoir flow and transport simulators at a selected timestep to calculate the SP throughout the reservoir model. The material properties required to calculate the SP are user defined and may be constant or vary in each cell. The code solves the equations governing flow and transport of electrical charge and global charge conservation using a control-volume-finite-difference scheme. Electrical currents associated with the SP may spread beyond the reservoir model domain, and the code allows for the domain to be extended vertically and laterally to account for this. Here, we present the governing equations and the numerical method used and demonstrate application of the code using an example in which we predict the SP signals associated with oil production from a subsurface reservoir supported by water injection.
Conference paperCosta Sousa M, Silva J, Silva C, et al., 2020,
Several applications of Earth Science require geologically valid interpretation and visualization of complex physical structures in data-poor subsurface environments. Hand-drawn sketches and illustrations are standard practices used by domain experts for conceptualizing their observations and interpretations. These conceptual geo-sketches provide rich visual references for exploring uncertainties and helping users formulate ideas, suggest possible solutions, and make critical decisions affecting the various stages in geoscience studies and modelling workflows. In this paper, we present a sketch-based interfaces and modelling (SBIM) approach for the rapid conceptual construction of stratigraphic surfaces, which are common to most geologic modelling scales, studies, and workflows. Our SBIM approach mirrors the way domain users produce geo-sketches and uses them to construct 3D geologic models, enforcing algorithmic rules to ensure geologically-sound stratigraphic relationships are generated, and supporting different scales of geology being observed and interpreted. Results are presented for two case studies demonstrating the flexibility and broad applicability of our rule-based SBIM approach for conceptual stratigraphy.
Journal articleYekta A, Salinas P, Hajirezaie S, et al., 2020,
This paper presents a numerical simulator for solving compositional multiphase flow and reactive transport. The simulator was developed by effectively linking IC-FERST (Imperial College Finite Element Reservoir SimulaTor) with PHREEQCRM. IC-FERST is a next-generation three-dimensional reservoir simulator based on the double control volume finite element method and dynamic unstructured mesh optimization and is developed by the Imperial College London. PHREEQCRM is a state-of-the-art geochemical reaction package and is developed by the United States Geological Survey. We present a step-by-step framework on how the coupling is performed. The coupled code is called IC-FERST-REACT and is capable of simulating complex hydrogeological, biological, chemical, and mechanical processes occurring including processes occur during CO2 geological sequestration, CO2 enhanced oil recovery, and geothermal systems among others. In this paper, we present our preliminary work as well as examples related to CO2 geological sequestration. We performed the model coupling through developing an efficient application programming interface (API). IC-FERST-REACT inherits high-order methods and unstructured meshes with dynamic mesh optimization from IC-FERST. This reduces the computational cost by placing the mesh resolution where and when necessary and it can better capture flow instabilities if they occur. This can have a strong impact on reactive transport simulations which usually suffer from computational cost. From PHREEQCRM the code inherits the ability to efficiently model geochemical reactions. Benchmark examples are used to show the capability of IC-FERST-REACT in solving multiphase flow and reactive transport.
Journal articleLi S, Jackson MD, Agenet N, 2020,
Journal articleOsman H, Graham GH, Moncorge A, et al., 2020,
Reservoir models typically contain hundreds-of-thousands to millions of grid cells in which petrophysical properties such as porosity and permeability vary on a cell-to-cell basis. However, although the petrophysical properties of rocks do vary on a point-to-point basis, this variability is not equivalent to the cell-to-cell variations in models. We investigate the impact of removing cell-to-cell variability on predictions of fluid flow in reservoir models. We remove cell-to-cell variability from models containing hundreds of thousands of unique porosity and permeability values to yield models containing just a few tens of unique porosity and permeability values grouped into a few internally homogeneous domains. The flow behavior of the original model is used as a reference. We find that the impact of cell-to-cell variability on predicted flow is small. Cell-to-cell variability is not necessary to capture flow in reservoir models; rather, it is the spatially correlated variability in petrophysical properties that is important. Reservoir modelling effort should focus on capturing correlated geologic domains in the most realistic and computationally efficient manner.
Journal articleKostorz WJ, Muggeridge AH, Jackson MD, 2020,
Journal articleLei Q, Jackson MD, Muggeridge AH, et al., 2020,
Modelling the reservoir-to-tubing pressure drop imposed by multiple autonomous inflow control devices installed in a single completion joint in a horizontal well, Journal of Petroleum Science and Engineering, Vol: 189, Pages: 1-16, ISSN: 0920-4105
Autonomous inflow control devices (AICDs) are used to introduce an additional pressure drop between the reservoir and the tubing of a production well that depends on the fluid phase flowing into the device: a larger pressure drop is introduced when unwanted phases such as water or gas enter the AICD. The additional pressure drop is typically represented in reservoir simulation models using empirical relationships fitted to experimental data for a single AICD. This approach may not be correct if each completion joint is equipped with multiple AICDs as the flow at different AICDs may be different. We use high-resolution numerical modelling to determine the total additional pressure drop introduced by two AICDs installed in a single completion joint in a horizontal well. The model captures the multiphase flow of oil and water through the inner annulus into each AICD. We explore a number of relevant oil-water inflow scenarios with different flow rates and water cuts. Our results show that if only one AICD is installed, the additional pressure drop is consistent with the experimentalzly-derived empirical formulation. However, if two AICDs are present, there is a significant discrepancy between the additional pressure drop predicted by the simulator and the empirical relationship. This discrepancy occurs because each AICD has a different total and individual phase flow rate, and the final steady-state flow results from a self-organising mechanism emerging from the system. We report the discrepancy as a water cut-dependent correction to the empirical equation, which can be used in reservoir simulation models to better capture the pressure drop across a single completion containing two AICDs. Our findings highlight the importance of understanding how AICDs modify flow into production wells, and have important consequences for improving the representation of advanced wells in reservoir simulation models.
Journal articleZhang Z, Geiger S, Rood M, et al., 2020,
Subsurface reservoir models have a high degree of uncertainty regarding reservoir geometry and structure. A range of conceptual models should therefore be generated to explore how fluids-in-place, reservoir dynamics, and development decisions are affected by such uncertainty. The rapid reservoir modelling (RRM) workflow has been developed to prototype reservoir models across scales and test their dynamic behaviour. RRM complements existing workflows in that conceptual models can be prototyped, explored, compared, and ranked rapidly prior to detailed reservoir modelling. Reservoir geology is sketched in 2D with geological operators and translated in real-time into geologically correct 3D models. Flow diagnostics provide quantitative information for these reservoir model prototypes about their static and dynamic behaviours. A tracing algorithm is reviewed and implemented to compute time-of-flight and tracer concentrations efficiently on unstructured grids. Numerical well testing (NWT) is adopted in RRM to further interrogate the reservoir model. A new edge-based fast marching method is developed and implemented to solve the diffusive time-of-flight for approximating pressure transients efficiently on unstructured tetrahedral meshes. We demonstrate that an implementation of the workflow consisting of integrated sketch-based interface modelling, unstructured mesh generation, flow diagnostics, and numerical well testing is possible.
Journal articleCollini H, Li S, Jackson MD, et al., 2020,
Zeta potential in intact carbonates at reservoir conditions and its impact on oil recovery during controlled salinity waterflooding, Fuel, Vol: 266, Pages: 1-16, ISSN: 0016-2361
It is well known that oil recovery from carbonate reservoirs can be increased by modifying the injected brine composition in a process ‘controlled salinity water-flooding’ (CSW). However, the mineral- to pore- scale processes responsible for improved oil recovery (IOR) during CSW remain ambiguous and there is no method to predict the optimum CSW composition for a given crude-oil-brine rock system. Here we report the first integrated experimental measurements of zeta potential and IOR during CSW obtained at reservoir conditions. The zeta potential is a measure of the electrical potential at mineral-brine and oil-brine interfaces and controls the electrostatic forces acting between these interfaces.We find that the measured zeta potential in clean samples saturated with formation brine is typically positive and becomes more negative with brine dilution irrespective of temperature. After aging and wettability alteration, the zeta potential changes and we suggest a more positive zeta potential indicates a positive zeta potential at the oil-brine interface and vice-versa. Injecting low salinity brine yields IOR when the oil-brine zeta potential is identified to be negative, but no response when it is identified to be positive, consistent with the hypothesis that IOR during CSW is caused by an increase in the repulsive electrostatic force acting between mineral-brine and oil-brine interfaces. We suggest that the optimum brine composition for IOR during CSW should be chosen to yield the largest change in zeta potential at the mineral-brine interface with opposing polarity to the oil-brine interface and can be determined using the experimental method reported here.
Journal articleSalinas P, Pain C, Osman H, et al., 2020,
Vanishing artifficial diffusion as a mechanism to accelerate convergence for multiphase porous media flow, Computer Methods in Applied Mechanics and Engineering, Vol: 359, Pages: 1-15, ISSN: 0045-7825
Numerical solution of the equations governing multiphase porous media flow is challenging. A common approach to improve the performance of iterative non-linear solvers for these problems is to introduce artificial diffusion. Here, we present a mass conservative artificial diffusion that accelerates the non-linear solver but vanishes when the solution is converged. The vanishing artificial diffusion term is saturation dependent and is larger in regions of the solution domain where there are steep saturation gradients. The non-linear solver converges more slowly in these regions because of the highly non-linear nature of the solution. The new method provides accurate results while significantly reducing the number of iterations required by the non-linear solver. It is particularly valuable in reducing the computational cost of highly challenging numerical simulations, such as those where physical capillary pressure effects are dominant. Moreover, the method allows converged solutions to be obtained for Courant numbers that are at least two orders of magnitude larger than would otherwise be possible.
Journal articleKampitsis AE, Adam A, Salinas P, et al., 2020,
Journal articleOnyenanu G, Hampson G, Fitch P, et al., 2019,
Characterisation of effective permeability in heterolithic, distal lower shoreface sandstone reservoirs: Rannoch Formation, Brent Group, UK North Sea, Petroleum Geoscience, Vol: 25, Pages: 519-531, ISSN: 1354-0793
The reservoir properties of distal lower-shoreface and distal wave-dominated delta-front deposits, which consist of sandstone beds with locally scoured bases and mudstone interbeds, are poorly understood. The lower Rannoch Formation (Middle Jurassic Brent Group) forms an interval of such heterolithic sandstones in many North Sea reservoirs, and is used to illustrate a workflow for rapid estimation of reservoir properties and their sensitivity to key parameters. Mudstone-interbed thickness distributions in cored reservoir successions are compared to the thickness distribution of sandstone scour-fills in an outcrop analogue(s) in order to identify mudstones with the potential to form laterally extensive barriers to vertical flow. Effective kv/kh at the scale of several typical reservoir-model grid cells (200 × 100 × 20 m) is estimated in intervals bounded by these mudstone barriers via a simple analytical technique that is calibrated to previously documented reservoir-modelling experiments, using values of sandstone proportion measured in cored reservoir successions. Using data from the G2 parasequence (Grassy Member, Blackhawk Formation, east-central Utah, USA) outcrop analogue, mudstones bounding 3–8 m-thick, upwards-coarsening successions in the lower Rannoch Formation may define separate stratigraphic compartments in which grid-cell-scale effective kv/kh is estimated to be 0.0001–0.001 using a streamline-based analytical method.
Journal articleMacAllister DJ, Graham MT, Vinogradov J, et al., 2019,
Characterising the self-potential response to concentration gradients in heterogeneous sub-surface environments, Journal of Geophysical Research. Solid Earth, Vol: 124, Pages: 7918-7933, ISSN: 2169-9356
Self‐potential (SP) measurements can be used to characterise and monitor, in real‐time, fluid movement and behaviour in the sub‐surface. The electrochemical exclusion‐diffusion (EED) potential, one component of SP, arises when concentration gradients exist in porous media. Such concentration gradients are of concern in coastal and contaminated aquifers, and oil and gas reservoirs. It is essential that estimates of EED potential are made prior to conducting SP investigations in complex environments with heterogeneous geology and salinity contrasts, such as the UK Chalk coastal aquifer. Here, we report repeatable laboratory estimates of the EED potential of chalk and marls using natural groundwater (GW), seawater (SW), deionised (DI) water and 5 M NaCl. In all cases the EED potential of chalk was positive (using a GW/SW concentration gradient the EED potential was c.14 to 22 mV), with an increased deviation from the diffusion limit at the higher salinity contrast. Despite the relatively small pore size of chalk (c.1 μm), it is dominated by the diffusion potential and has a low exclusion‐efficiency, even at large salinity contrasts. Marl samples have a higher exclusion‐efficiency which is of sufficient magnitude to reverse the polarity of the EED potential (using a GW/SW concentration gradient the EED potential was c.‐7 to ‐12 mV) with respect to the chalk samples. Despite the complexity of the natural samples used, the method produced repeatable results. We also show that first order estimates of the exclusion‐efficiency can be made using SP logs, supporting the parameterisation of the model reported in Graham et al. (2018), and that derived values for marls are consistent with the laboratory experiments, while values derived for hardgrounds based on field data indicate a similarly high exclusion‐efficiency. While this method shows promise in the absence of laboratory measurements, more rigorous estimates should be made where possible and can be conducted following
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.