Spontaneous imbibition is well-known to be one of the most effective processes of oil recovery in fractured reservoirs. However, the detailed pore-scale mechanisms of the counter-current imbibition process and the effects of different fluid/rock parameters on this phenomenon have not yet been deeply addressed. This work presents the results of a new pore-level numerical study of counter-current spontaneous imbibition, using coupled Cahn–Hilliard phase field and Navier–Stokes equations, solved by a finite element method. A 2D fractured medium was constructed consisting of a nonhomogeneous porous matrix, in which the grains were represented by an equilateral triangular array of circles with different sizes and initially saturated with oil, and a fracture, adjacent to the matrix, initially saturated with water and supported by low rate water inflow. Through invasion of water into the matrix, oil drops were expelled one by one from the matrix to the fracture, and in the matrix, water progressed by forming capillary fingerings, with characteristics corresponding to the experimental observations. The effects of wettability, viscosity ratio, and interfacial tension were investigated. In strongly water-wet matrix, with grain contact angles of θ < π/8, different micro-scale mechanisms were successfully captured, including oil film thinning and rupture, fluids’ contact line movement, water bridging, and oil drop detachment. It was notified that there was a specific grain contact angle for this simulated model, θ = π/4, above it, matrix oil recovery was negligible by imbibition, while below it, the imbibition rate and oil recovery were significantly increased by decreasing the contact angle. In simulated mixed wet models, water, coming from the fracture, just invaded the neighboring water-wet grains; the water front was stopped moving as it met the oil-wet grains or wide pores/throats. Increasing water-oil interfacial tension, in the range of 0.005-0.05 N/m, resulted in both higher rate of imbibition and higher ultimate oil recovery. Changing the water-oil viscosity ratio (M), in the range of 0.1–10, had a negligible effect on the imbibition rate, while due to co-effects of capillary fingering and viscous mobility ratio, the model with M = 1 had relatively higher ultimate oil recovery.

The aim of the present work is to assess the effects of wettability, heterogeneity, and viscosity differences on water‐oil displacement process in micromodel porous media through numerical modelling. The two‐phase flow was simulated by Cahn‐Hilliard phase field method (PFM) using a finite element package. The micromodel was initially saturated with oil (wetting phase) and oil was produced through invasion of the displacing phase into the matrix. The computed oil and water saturations were in good agreement with those obtained by the visual flooding experiment. Using the validated model, sensitivity analysis was performed to investigate the effects of different wettability states, heterogeneity, and viscosity on the displacement process. The obtained results showed that the final oil saturation of the homogeneous pattern is 0.37 which is 13 % less than that of the heterogeneous one. For the highly oil‐wet medium (), the capillary forces prevented water to invade more pore bodies and resulted in 0.62 ultimate oil saturation; however, the final oil saturation in the neutral wet () was ∼0.5. Increasing the viscosity of displacing agent formed lower channelling and fingering which led to higher oil recovery due to the favourable mobility ratio. The present study demonstrates that PFM can be a reliable approach to capture micro‐ and macro‐scale mechanisms in the simulation of immiscible two‐phase flow in micromodel porous media with a reasonable computational time.

No abstract

Biotechnology has had a major effect on improving crude oil displacement to increase petroleum production. The role of biopolymers and bio cells for selective plugging of production zones through biofilm formation has been defined. The ability of microorganisms to improve the volumetric sweep efficiency and increase oil recovery by plugging off high-permeability layers and diverting injection fluid to lower-permeability was studied through experimental tests followed by multiple simulations. The main goal of this research was to examine the selective plugging effect of hydrophobic bacteria cell on secondary oil recovery performance. In the experimental section, water and aqua solution of purified Acinetobacter strain RAG-1 were injected into an oil-saturated heterogeneous micromodel porous media. Pure water injection could expel oil by 41%, while bacterial solution injection resulted in higher oil recovery efficiency; i.e., 59%. In the simulation section, a smaller part of the heterogeneous geometry was employed as a computational domain. A numerical model was developed using coupled Cahn–Hilliard phase-field method and Navier–Stokes equations, solved by a finite element solver. In the non-plugging model, approximately 50% of the matrix oil is recovered through water injection. Seven different models, which have different plugging distributions, were constructed to evaluate the influences of selective plugging mechanism on the flow patterns. Each plugging module represents a physical phenomenon which can resist the displacing phase flow in pores, throats, and walls during Microbial-Enhanced Oil Recovery (MEOR). After plugging of the main diameter route, displacing phase inevitably exit from sidelong routes located on the top and bottom of the matrix. Our results indicate that the number of plugs occurring in the medium could significantly affect the breakthrough time. It was also observed that increasing the number of plugging modules may not necessarily lead to higher ultimate oil recovery. Furthermore, it was shown that adjacent plugs to the inlet caused flow patterns similar to the non-plugging model, and higher oil recovery factor than the models with farther plugs from the inlet. The obtained results illustrated that the fluids distribution at the pore-scale and the ultimate oil recovery are strongly dependent on the plugging distribution.

No abstract