Imaging of fluid flow in porous rocks by neutron radiography

Dynamic neutron radiography (DNR) is a tool that has been used for some years for the non-destructive testing of mechanical equipment, especially where fluid flow is involved (Lindsay et al., 1989; Balasko and Svab, 1996). The method is particularly valuable as it can present a visualization of the fluid flow. However, applications of this method to petroleum exploration and production have not been reported extensively in the literature until recently (Middleton and Pazsit, 1998a, 1998b, 1999; Balasko et al., 1999). Because the technique is capable of producing 'real-time' images of fluid flow in rocks, quantities of geological interest, such as flow rate, fluid mixing and rock inhomogeneities, can be observed. Quantitative estimates of porosity, permeability, fluid saturation and capillary pressure can also be made from the neutron-attenuation images of a porous reservoir rock. Previous studies have shown that DNR can (i) highlight inhomogeneities in fluid flow patterns in porous rocks, and (ii) provide a good estimate of relative fluid saturation, if 'heavy water' is displaced by oil, in porous reservoir sandstones (Middleton and Pazsit, 1998a). The power of DNR over NMR (nuclear magnetic resonance) computer-aided-radiography/tomography is that the resultant image is not distorted by the presence of iron, or other magnetic minerals, in the sample. The aim of this paper is to demonstrate the ability of this method to delineate (1) 'effective porosity' (pore space involved in fluid flow), versus (2) total porosity (total pore space in the rock). The attenuation of thermal neutrons, which are low-energy neutrons of energy of about 0.025 eV, can be described well by the exponential rule: I = Io exp(-μ ρ h), (1), where I is the intensity of the transmitted neutron beam, Io is the intensity of the incident neutron beam, μ is the neutron attenuation coefficient, ρ is the density of the rock sample, and 'h' is the thickness of the sample. The attenuation coefficient, μ, used herein reflects the total attenuation, due to both the scattering and capture processes for thermal neutrons. For the types of experiments that we are doing in the present study, a dry sandstone is either fully saturated with water under a vacuum, or partially saturated due to the passage of water through the sample under low pressure. Based on equation (1), one can write simple expressions for the attenuated neutron intensity of a dry sample, a partially saturated sample and a fully saturated sample. The partially saturated experiment gave an average porosity of 15%, and the fully saturated sample yielded and average porosity of 21%