Imaging-based Dispersion in CT Coreflooding Experiments

Monitoring core flooding experiments with medical CT scanners has become increasingly accessible since the introduction of medical CT scanners in the geosciences in the mid-1980s. While initially 3D imaging was mainly used to assess rock heterogeneity in a qualitative manner, over the past decades the technique has matured to acquiring time series of 3D images (“4D”) in-situ during core flooding experiments. The medical CT scanners commonly used for this purpose (but also other 3D imaging methods) have a finite scanning time even for very coarse spatial resolution. Depending on the injection rate, this scanning time may not be negligible compared to advection time of fluids in the core. As it occurs in photography of moving objects where a long shutter time leads a “motion blurring”, such effect also occurs in core flooding experiment because of the relative movement of the fluids and the imaging system. This “image dispersion” can then be mis-interpreted as physical (hydrodynamic) dispersion or leads to incorrect heterogeneity characterization. To demonstrate the impact, we consider the scenario of solute transport. For illustration purpose we start with a homogeneous case described with a 1D dispersive model. We show that the degree of image-dispersion on the concentration profiles comes as an additional smearing of the profiles as the relative velocity of the CT imaging to the fluid front increases. We then proceed with 3D heterogeneous case where it is demonstrated how imaging-based dispersion leads to the misinterpretation of flow in spatially resolved heterogeneity as hydrodynamic dispersion. The findings suggest that the imaging-based dispersion effect needs to be considered for practically all 4D imaging methods such as CT, PET, MRI of dynamic processes with respect to the interpretation of heterogeneity and its effects, e.g. hydrodynamic dispersion, capillary dispersion of a flow front.