Injection Rate Effects On Waterflooding Mechanisms And Injectivity In Cohesionless Sand

Golovin, E. (University of Illinois) | Chudnovsky, A. (University of Illinois) | Dudley, J.W. (Shell Exploration and Production Company) | Wong, G.K. (Shell Exploration and Production Company)


This paper presents hydraulic fracturing and flooding experiments with cohesionless sand designed to investigate the injectivity mechanisms. Tests are performed in 2-foot cubic sand packs in a stiff steel chamber and subjected to controlled 3-D confining stresses. Multiple injection stages of a heated low viscosity dyed gel-water solution that solidifies to preserve the invaded zone morphology are done. Post-test examination of the solidified invaded zones displays a multilayered structure of sand developed during the flooding and/or fracturing stages of injection. Under flooding conditions a uniform sand structure is observed with a smooth boundary between the layers corresponding to fluid injected on the earlier stage and outwardly displaced by the following injection. In contrast with that, a heterogeneous, turbulent invaded zone structure with clearly visible irregular higher porosity/permeability “channels” is formed under fracturing conditions. The fluid injected after a “fracturing stage” moves primarily through these channels (bypassing the previously injected fluid) with a significant injectivity increase (up to 35-40%). No discrete fractures are observed. A cyclic low frequency injection rate is also shown to be an alternative method to fracturing for improving injectivity. Improved injectivity in this case is achieved without creation of any noticeable disturbance to the sand structures, like channels, and maintains a more uniform fluid distribution.


Some of the unconsolidated sand formations such as those offshore in the Gulf of Maxico and offshore Western Canada contain high viscosity oil and might be considered for water or polymer flooding to improve recovery and enhance oil production [1, 2]. Flooding in unconsolidated sand could lead to fracture initiation and propagation [3]. Fracturing may be related to the presence of small quantities of impurities and solids in the injection fluid (discrete fracture formation) as well as to a high injection rate (diffusive fracture formation). Although fracturing enhances permeability and injectivity of sand reservoirs, fracture initiation and growth must be under control to avoid intercepting production wells and other undesirable consequences of fracture creation.

For unconsolidated sand, there are currently no sound theoretical predictions for fracture initiation and propagation. There are still a only limited number of published laboratory studies of the flooding/fracturing processes in unconsolidated sand [4-7,11] for relevant ranges of in-situ stress, injection rate and injection fluid rheology, including limited experiments dedicated to studying the effect of flow enhancement by injection rate and pressure pulsing [8, 9]. All these publications describe tests performed in isotropic or biaxial stress conditions. They do not replicate the typical fully 3-D field stress conditions, with different minimum, intermediate and maximum principal stresses. In order to better quantitatively characterize flooding/fracturing in cohesionless sand, the study presented in this paper was carried out on cubical sand packs with different confining pressures in the three principal directions. It is a modified and upgraded pressure chamber previously used in our studies of hydraulic fracture initiation and propagation [10]. In this paper we report the observation and characterization of the effect of constant and cyclic injection rate on the waterflooding mechanism and injectivity in oil-wetted air-saturated and water-saturated cohesionless sand.