Developing High Spatial Resolution MRI Methods for Characterisation of Porous Materials

In this thesis, high-resolution, quantitative 3D magnetic resonance imaging (MRI) methods are demonstrated to study the microstructure of, and fluid transport processes in porous rocks. A particular motivation of this work is to provide pore-scale, quantitative, spatially-resolved structural and flow information of rocks to aid the development of Digital Rock (DR) technology – a tool based on pore-scale imaging and modelling that plays an increasingly important role in the oil and gas industry and the deployment of carbon capture and storage technologies. To be able to study pore-scale characteristics of rocks, the spatial resolution of 3D MRI was increased by 1–2 orders of magnitude (relative to routine MRI acquisitions), up to as high as 17.6 μm, using sensitive MRI equipment in combination with rapid and under-sampled MRI pulse sequences and compressed sensing data reconstruction techniques; 17.6 μm is the highest spatial resolution reported for MRI images of rocks. To this end, a novel MR data under-sampling approach was developed using input from X-ray micro-computed tomography (μCT) data to derive optimal sampling schemes for acquiring high-resolution 3D MRI images of rocks. This approach was used to speed up the acquisition of structural and flow MRI images. Quantitative, spatially-resolved under-sampled 3D flow MRI methods, namely velocity mapping and spatially-resolved propagators, were developed and applied to study structure flow correlations for a single-phase flow through a Ketton limestone rock. 3D velocity maps acquired at 35 μm spatial resolution revealed that the flow in Ketton is highly heterogeneous with ∼ 10 % of the pores carrying more than 50 % of the flow. Structure-flow correlations were found between the local pore velocity and the size and topology of the pores. Coregistration of MRI and μCT data was used to identify complex flow patterns in the rock. By analysing 3D spatially-resolved propagators, each containing 331,776 local propagators, as a function of observation time, pore-scale flow dispersion was observed. Single-phase fluid flow velocity fields in Ketton and Estaillades limestone core plugs were computed using pore-scale lattice Boltzmann method (LBM) simulations, performed directly on the μCT images of the pore space of rocks, and then benchmarked to 3D MRI velocity maps acquired at 35 μm spatial resolution for flow of water through the same rock samples. For Ketton rock, good quantitative and qualitative agreement was found between the simulated and MRI velocity fields. For Estaillades rock, which presents a more heterogeneous case with many microstructural features below the spatial resolution of the μCT image, many complex flow patterns were qualitatively reproduced by the simulation, although some local differences between the LBM and MRI velocity maps were observed. Novel, chemically-selective under-sampled 3D MRI techniques were demonstrated to acquire quantitative, high-resolution images of oil and water fluid phases in Estaillades core plugs at the end of spontaneous and forced imbibition experiments. The high spatial resolution (35 μm) and quantitative nature of the MRI images acquired enabled oil- and water-containing microstructures to be identified and local oil and water saturations to be quantified. Disconnected oil clusters were observed in some large pores at the end of forced imbibition. Using a novel, high-resolution, chemically-selective 3D velocity mapping method, the remaining oil was confirmed to be stagnant. The flow of water in the rock was highly localised and distant from where the remaining oil was located, thus limiting the ability to recover more oil.This work was funded by Shell