Polymer flooding is a mature EOR technology, but several pore scale phenomena with possible large influence on the reservoir scale are poorly understood. This paper describes and analyses oil mobilization experiments of heavy oils by imaging instable displacement at adverse mobility ratio water and polymer floods. Two-dimensional flood experiments have been performed using Bentheimer outcrop slabs. X-ray imaging is utilized to visualize displacements and to determine the underlying flow mechanisms. Viscous fingering, water channel formation and oil displacement are described for a series of mobility ratios. Mechanistic understanding of development and propagation of viscous fingers at adverse mobility ratio may be used to improve reservoir simulations. Description of oil mobilization for various mobility ratios may give guidelines for choice of polymer concentration and slug size for polymer floods. Bentheimer slabs were drained using oils with 4 viscosities (5 − 616 mPa∙s). X-ray imaging revealed differences in water-finger formation, and width and growth of fingers with increasing mobility ratio. Lower mobility ratios showed formation of wide fingers or water channels. Oil recovery was dominated by propagation of these channels, but still showed poor sweep efficiency (water breakthrough 0.3 – 0.5 PV). At high mobility ratio, water breakthrough occurred very early at 0.08 – 0.15 PV. Here, the oil recovery mechanism was totally different. Oil was mobilized by polymer injection through cross-flow into the water channels. Polymer flood showed rapid change in oil cut and high total oil recovery efficiency. Through analysis of 2D x-ray images, mechanisms for fingering initiation and propagation and for oil mobilization by polymer is visualized and discussed as a function of mobility ratio. The results presented here may impact polymer flood design, in particular the choice of polymer injection strategy for heavy oil reservoirs. Data show that relatively low polymer concentrations are sufficient for mobilizing heavy oil.

This paper discusses miscible gas injection in fractured reservoirs, based on compositional reservoir modeling. Developed miscibility in conventional one-dimensional systems is dependent on fluid properties alone, and can be estimated by well known experimental and numerical procedures. This is not the case for heterogeneous and multi-dimensional systems such as fractured reservoirs. Based on a multi-cell algorithm a method is proposed to determine conditions of developed miscibility in fractured reservoirs. The proposed method makes it possible to evaluate gas injection processes, and developed miscibility in particular, in specific regions of the reservoir. A systematic step-by-step extension from a single one-dimensional matrix block to more complex matrix-fracture systems shows that the minimum miscibility pressure/enrichment (MMP/MME) level in a fractured reservoir is significantly higher than for a conventional one-dimensional single-porosity system. This is mainly due to multi-dimensional flow and molecular diffusion. On the other hand, even at pressures significantly below the fracture system MMP, substantial enhancements in the recovery rate and ultimate oil recovery can be expected by non-equilibrium gas injection. It is shown that reduction of the gas-oil interfacial tension and internal Darcy flow induced by interfacial tension gradients play a key role. This capillary driven process may lead to accelerated production and very high ultimate recoveries. Introduction Conventional recovery strategies in fractured reservoirs normally involve depletion and/or immiscible injection schemes (water or lean gas injection). An alternative to these conventional recovery methods is enriched gas injection, potentially leading to a miscible situation and high ultimate recoveries. Miscibility in conventional single porosity reservoirs has been studied extensively over the past years. Experimental and numerical procedures have been developed that give a definitive measure of the miscible development for a one-dimensional (1-D) flow system1–3. The 1-D miscible flow process has been shown to be dictated by phase behavior alone, and not at all by rock or other fluid properties. For fractured reservoirs, however, the physics behind developed miscibility is unknown. It is even uncertain whether miscibility can be obtained in a fractured system. We have not found references on this subject in the literature. Hence, the objective of this work was mainly to gain a better physical understanding of the potential for miscible displacement in fractured reservoirs and to develop numerical procedures for determination of miscibility in dual-porosity systems. With the introduction of fractures, the displacement process no longer depends on fluid properties alone. The fracture-matrix geometry, size and interaction, and other physical phenomena also play an important role. Instead of a well-defined displacement of oil by gas, the injection gas tends to flow in the highly permeable fractures, surrounding the oil in the matrix blocks. The gas composition inside the matrix blocks is usually different from the composition in the fractures. Provided the matrix is higher than the capillary entry height, gas enters at the top and may the upper sides of the matrix block. This means that the gas composition near the displacement front will be different than for purely vertical flow. For significant differences in fluid compositions between the matrix and fracture media, diffusion may play an important role in fractured reservoirs. All this suggests that the total composition, (and hence the miscible process), near the displacement front will be different than from a one-dimensional slimtube displacement. Away from the injectors in a fractured reservoir, the fracture gas composition may be influenced by the upstream matrix-fracture fluid exchange. Intuitively, the gas entering the matrix blocks some distance away from the injectors will be richer than the original injection gas. That is, one might expect more favorable conditions of developed miscibility and improved recoveries from the matrix blocks in the reservoir regions away from the injectors.

No abstract

This paper presents a practical workflow for evaluating standard MMP estimation methods and fluid characterization and lumping schemes for equation-of-state models with a high number of pseudo-components (typically 22). It has also been illustrated the validity of a minimum tuning procedure (mostly to saturation pressure, constant-mass expansion test at Tres, differential liberation experiment if available; adjusting only Tc, Pc), and the predictability of each of the applied approaches. A quick estimate of MMP with the various methods available is critical when the experimental data are not measured. Given an uncertainty range to calculated MMP based on the data comparison, one might make a valuable contribution to the project where multi-contact miscibility needs to be modelled properly. In the present study, the experimental slim tube MMP data were measured for reservoir fluids interacted with rich separator gases (17 cases), with pure CO2 (3 cases). The reference reservoir fluids were mostly collected from the Norwegian Continental Shelf. A minimum tuning procedure for estimating MMP was validated throughout the present study. The fluid characterization workflow proposed in the present study is an efficient starting point to compare estimated MMPs. The results of MMP estimation methods, the ‘tie-line’ approach (in PVTsim software), slim tube simulation approach (both in PVTsim and Eclipse 300) and the ‘mixing cell’ approach, agree reasonably well within laboratory accuracy range. The uncertainty range observed in the present study would be acceptable for most field application cases.

Block to block processes in a vertical stack of matrix blocks in a fractured reservoir have been extensively studied for gas-oil systems. Similar investigations have, however, not been reported for the case of water-oil systems. This study details the block to block processes and the effect of capillary continuity between the matrix blocks on oil recovery by water displacement in a fractured reservoir. Fine grid simulation studies under water flooding have been carried out on Eclipse 100 for varying degrees of matrix to matrix contact to illustrate the dependence of oil recovery on the position and the extent of the areal contact between the matrix blocks. Our results show that for the simplified system of mixed wet matrix blocks, the phenomenon of oil re-drainage from the downdip blocks to the updip blocks takes place both with or without matrix contacts between them. Such a phenomenon is not observed in case of strong water wet rock. For strong water wet rock, capillary imbibition controls the oil recovery and the effect of capillary continuity is not significant. However, for reservoirs of mixed wettability, such as Ekofisk, the oil recovery is significantly enhanced even with a moderate areal contact between the matrix blocks. The additional oil recovery, which is substantial as compared to the isolated single matrix block oil recovery, is due to gravity forces and depends on the thickness of the payzone. The oil recovery and the recovery time are also dependent on the tortuosity of the capillary continuity. The findings of the study have applications in the prediction of oil recovery by water displacement in a fractured reservoir. Introduction Flow between matrix block and the fracture is controlled mainly by the capillary and the gravity forces though other forces such as expansion, diffusion and viscous may also influence the recovery process depending upon the pressure, temperature and the composition of the fluids involved in the process. For oil-water system, the influence of diffusion may be neglected and if the water displacement takes place at constant pressure above bubble point pressure, the role of expansion can also be neglected. For gas-oil system, the capillary pressure opposes the fluid exchange between the matrix block and the fracture and the oil expulsion (in absence of diffusion) from the matrix block can only be possible if the height difference between the gas-oil contacts in the matrix and the fracture is greater than the capillary threshold height. If the height of the matrix block is less than the capillary threshold height, oil will not be recovered from the matrix blocks unless there is a capillary continuity between them. Block to block processes have been intensively studied by many authors1–3. Festoy and Van Golf-Racht3 showed that the matrix to matrix contact between the blocks in a vertical stack improves the oil recovery significantly. According to them, high oil recovery and sustained production rates from some fractured reservoirs flowing below bubble point pressure may be attributed to some degree of matrix to matrix contact between the blocks in a vertical stack.