Search

Estimating the longitudinal dispersion coefficient DL in flow through heterogeneous porous media is paramount to many problems in geological formations. Moreover, although it is well‐known that DL is sensitive to the morphology of such formations, it has been very difficult to establish a firm link between the two. We describe a novel deep convolutional neural network (DCNN) for estimating DL. The inputs for training of the network are a large and diverse set of data consisting of three‐dimensional images of porous media, as well as their porosity, and the associated DL values computed by random‐walk particle‐tracking (RWPT) simulations. The trained network predicts DL very rapidly, and its predictions are in excellent agreement with the data not used in the training. Thus, a combination of the DCNN and RWPT simulation provides a powerful tool for studying many flow‐related phenomena in geological formations, and estimating their properties. Plain Language Summary: Measuring or computing the dispersion coefficient DL in flow through porous media, a fundamental characteristic of transport in geological formations and risk analysis, is a time‐consuming endeaver. Moreover, although DL is sensitive to the morphology of a pore space, a direct link between the two has been missing. We proposed a deep convolutional neural network for predicting DL, using 3D images of porous media and their porosities. The accuracy of the predictions for the actual data indicates that the ability of the network for estimating the important flow and transport properties of porous media for new input data. The present work was at the core scale, on the order of the physical sizes of the sandstones used in our study. The same approach may be used at the eld scale. In that case, one generates the input data by following the same procedure, except that the data should be generated for models in which the permeability and porosity of the formation vary spatially, and represent correlated fields. Work in this direction is in progress. Key Points: Dispersion coefficientConvolutional neural networkFlow through porous media [ABSTRACT FROM AUTHOR]

The rate of carbon dioxide (CO2) dissolution in saline aquifers is the least well‐constrained of the secondary trapping mechanisms enhancing the long‐term security of geological carbon storage. CO2 injected into a heterogeneous saline reservoir will preferentially travel along high permeability layers increasing the CO2‐water interfacial area which increases dissolution rates. We provide a conservative, first‐principles analysis of the quantity of CO2 dissolved and the rate at which free‐phase CO2 propagates in layered reservoirs. At early times, advection dominates the propagation of CO2. This transitions to diffusion dominated propagation as the interfacial area increases and diffusive loss slows propagation. As surrounding water‐filled layers become CO2 saturated, propagation becomes advection dominated. For reservoirs with finely bedded strata, ∼10% of the injected CO2 can dissolve in a year. The maximum fraction of CO2 that dissolves is determined by the volumetric ratio of water in low permeability layers and CO2 in high permeability layers. Plain Language Summary: To limit global warming to 2°C, it is likely that large amounts of carbon dioxide (CO2) will need to be stored underground. A significant fraction of the total possible storage space for CO2 is in salt water reservoirs, kilometers beneath the surface. It is important that once the CO2 has been injected underground, it is securely trapped, otherwise, there is a risk that it could leak back to the surface. After the CO2 is injected into these reservoirs, it can dissolve in the surrounding water, greatly reducing the risk of leakage, although complete dissolution of all the injected CO2 may take millions of years. However, preferential flow of CO2 along more permeable layers in geological formations creates a complex front between the water and CO2 which increases the surface area available for dissolution. This study calculates the minimum amount of injected CO2 that can dissolve in such a reservoir and how far it travels. Using injection of CO2 for enhanced oil recovery at the Salt Creek Field in Wyoming as an example, we find that in one year around 10% of the total injected CO2 can dissolve into the surrounding water by this process and become trapped. Key Points: Injection of carbon dioxide into finely bedded reservoirs leads to enhanced contact area with water and hence enhanced dissolution ratesPropagation rates of free‐phase carbon dioxide transition from advection to diffusion dominated as the contact area with water increasesFor injection into the Salt Creek reservoir, Wyoming, nearly 10% of the injected carbon dioxide is predicted to dissolve in one year [ABSTRACT FROM AUTHOR]

Estimating flow and transport properties of porous media that undergo deformation as a result of applying an external pressure or stress is important to a wide variety of processes, ranging from shale formations into which a fracking liquid is injected to CO2 sequestration in geological formations. We propose a novel model for estimating the effective permeability of such porous media. Given the pore‐size distribution (PSD) of a porous medium before deformation, as well as the Young's modulus of its grains, the model uses Hertz‐Mindlin theory of contact between grains to compute the new PSD that results from applying an external pressure P to the medium and utilizes the updated PSD in the effective‐medium approximation to estimate the effective permeability at P. Comparison between the predictions and experimental data for several types of sandstones indicates excellent agreement between the two. Key Point: This demonstrates the first application of the effective‐media approximation to estimating the permeability of deforming porous media [ABSTRACT FROM AUTHOR]