3D printing-based microfluidics for geosciences

Abstract

Three-dimensional (3D) printing offers the potential to repeatably generate porous media for the investigation of pore-scale processes such as CO2 dissolution and species transport in multiscale porous media. However, there are concerns regarding dimensional fidelity, shape conformity and surface quality of the 3D printed products, and therefore, the printing quality and printer limitations must be benchmarked. Firstly, we investigate the ability to generate porous media with our 3D printing setup. We show that our 3D printing setup allows for cheap and fast fabrication of micromodel devices from simple to complicated multiscale geometries, which enable the ability to perform repeatable single-phase flow species transport experiments. We use Particle Image Velocimetry (PIV) and Direct Numerical simulations (DNS) to show how, with our 3D printing setup, we can generate custom-designed micromodels accurately and repeatably with minimum pore-throat sizes of 140 µm. To enhance the management of subsurface engineering processes in multiscale porous media, containing fractures and matrix, it is crucial to comprehend the interaction between the larger-scale features (fractures) and the smaller-scale features (matrix). While models describing fluid flow in multiscale-porous media exists, it has not been validated experimentally due to lack of a benchmark experimental dataset. In this work we use 3D printing to fabricate geometries that encompass both fractures and matrix. We conduct species transport experiments and generate a benchmark experimental dataset for single-phase flow species transport in multiscale geometries. Our findings demonstrate that 3D printed multiscale micromodels allow for the visualization of species transport propagation in such geometries, enabling the acquisition of a benchmark experimental dataset. Subsequently, we use the acquired experimental dataset and direct numerical simulations (DNS) to validate the multiscale species transport Darcy Brinkman Stokes (DBS) simulations in multiscale geometries containing both fractures and matrix, which have not been previously validated. Our research shows how DBS can accurately predict the temporal evolution of species propagation in these multiscale geometries. Finally, we use simple 3D printed geometries consisting of a single channel and dead-end pores to investigate the trapping and dissolution of CO2 bubbles. Dissolution of CO2 bubbles in the pore-space is an important trapping mechanism during CO2 storage operations, however, a benchmark experimental dataset of dissolution of CO2 bubbles that could validate direct numerical models does not exist. We show that repeatable experiments can be performed in simple geometries and a benchmark experimental dataset for multiphase flow processes can be obtained. As a result, we developed a benchmark experimental dataset for validating DNS models describing the dissolution of trapped CO2 bubbles which have not been before validated against experimental data. Finally, we use DNS simulations and show that while DNS can accurately capture dissolution of a CO2 bubble in simple geometries, such as a dead-end pore and a throat, the current computational requirements do not allow for simulating more complicated cases.