This thesis presents an experimental investigation of two-phase flow properties of porous rock samples having different scales ranging from micro-CT imaging to conventional core plug scales. Advances in micro-CT imaging of porous materials provide the opportunity to extract representative networks from the images. This improves the predictive capability of porescale network models to predict multiphase flow properties. However, all these predictions need to be validated with laboratory data. Micro-CT imaging is currently limited to small sample sizes, having bulk volumes of the order of 0.1 cm??. Conventional core plugs, however, have sizes several orders of magnitude larger than that (bulk volumes of 10 cm?? or larger). The aim of this thesis is to investigate the scale effect on laboratory data and to provide reliable experimental data which can be used to test the predictive value of microCT based network models. Berea and Bentheim sandstones and Mount Gambier carbonate were used in the experiments. The core samples were thoroughly cleaned in order to obtain strongly, uniform water-wet conditions. Simple well-characterized fluid systems were chosen in the experiments: Air-brine fluid-system for drainage capillary pressure, resistivity index and spontaneous imbibition experiments while oil-brine fluid-system for wettability and relative permeability measurements. Drainage capillary pressure, resistivity index, relative permeability and spontaneous imbibition measurements were made on the cores having bulk volumes ranging from 0.1 to 12 cm??. Previous studies have shown that experiments at this scale are still lacking. The wettability was tried to keep strongly water-wet for all experiments. The experimental results show that the measurements of drainage capillary pressure, and resistivity index and spontaneous imbibition on small core samples, having similar scales as micro-CT imaging can be made accurately in the laboratory. The measurement of relative permeability remains challenging. This thesis concludes that commonly used homogeneous rock types (Berea and Bentheim sandstones and Mt. Gambier carbonate) can be considered to be sufficiently homogeneous from the pore to core scale based on the two-phase flow properties examined in this study. Hence, laboratory data taken from these rocks using conventional core plugs can be used to calibrate micro-CT based network models for multiphase flow properties.

Understanding soil structure, in particular the void spaces through which water, gases and solutes flow and in which organisms exist, is vital to a sustainable future on earth. The investigation of the structural behaviour of soil under different influences is fundamental to understanding and protecting the soil. This study has investigated the impact of bacteria on the biophysics of water retention and flow, aiming to elucidate the effect of three key components produced by the model organism, Pseudomonas fluorescens SBW25. Cellulose is an extracellular polysaccharide involved in the formation of the matrix of the bacterial biofilm, lipopolysaccharide is a cell membrane component required for bacterial attachment, and viscosin is a biosurfactant released from the bacteria. Four isogenic strains mutated so as to heighten or suppress production of one of these key components were used in addition to the wild-type strain. Labfield sandy loam soil was sieved and packed into replicate experimental cores which were incubated with different bacterial treatments. Following sterilisation, the gravimetric water content (u g g-1) of the soil was determined at equilibrated matric potentials from -1 cm to -100 cm during two wet-dry cycles. Sorptivity (S, mm s-1/2) of the soil, indicative of water repellency, was determined using a mini-infiltrometer setup and has been reported as the rate of infiltration of water into the soil. Bacteria have been shown to increase water repellency of soil, decrease the total water content at saturation and increase the water retaining ability of the soil as it drains (p < 0.05). Three-dimensional analysis of core scale structure was carried out using micro X-ray computed tomography (µXCT) and of aggregate scale structure using synchrotron-µXCT. Volumetric analyses of the 3D structures has shown decreased pore connectivity and destabilisation of aggregates in soil systems treated with bacteria deficient in the production of a key extracellular component, cellulose, LPS or viscosin (p < 0.05). Analyses of cracking patterns in two types of sandy loam soil, Labfield and Bullionfield has highlighted the importance of taking into account the soil type and its composition when studying soils, as even within soil classification groups different behaviours are observed. This study has provided clear evidence of the ability of bacteria and their extracellular components to impact upon (i) the hydrodynamics of water retention and flow in soil and (ii) the structural organisation, aggregation and stabilisation of soil.