No abstract

Summary Chemical enhanced oil recovery (EOR) leads to substantial incremental costs over waterflooding of oil reservoirs. Reservoirs containing oil with a high total acid number (TAN) could be produced by the injection of alkali. Alkali might lead to the generation of soaps and emulsify the oil. However, the generated emulsions are not always stable. Phase experiments are used to determine the initial amount of emulsions generated and their stability if measured over time. On the basis of the phase experiments, the minimum concentration of alkali can be determined and the concentration of alkali above which no significant increase in the formation of initial emulsions is observed. Micromodel experiments are performed to investigate the effects on the pore scale. For the injection of alkali into high-TAN oils, the mobilization of residual oil after waterflooding is seen. The oil mobilization results from the breaking up of oil ganglia or the movement of elongated ganglia through the porous medium. As the oil is depleting in surface-active components, residual oil saturation is left behind either as isolated ganglia or in the down gradient side of grains. Simultaneous injection of alkali and polymers leads to a higher incremental oil production in the micromodels owing to larger pressure drops over the oil ganglia and more-effective mobilization accordingly. Coreflood tests confirm the micromodel experiments, and additional data are derived from these tests. Alkali/cosolvent/polymer (ACP) injection leads to the highest incremental oil recovery of the chemical agents, which is difficult to differentiate in micromodel experiments. The polymer adsorption is substantially reduced if alkali is injected with polymers compared with polymer injection only. The reason is the effect of the pH on the polymers. As in the micromodels, the incremental oil recovery is also higher for alkali/polymer (AP) injection than with alkali injection only. To evaluate the incremental operating costs of the chemical agents, equivalent utility factors (EqUFs) are calculated. The EqUF takes the costs of the various chemicals into account. The lowest EqUF and, hence, the lowest chemical incremental operating expenditures are incurred by the injection of Na2CO3; however, the highest incremental recovery factor is seen with ACP injection. It should be noted that the incremental oil recovery owing to macroscopic-sweep-efficiency improvement by the polymer needs to be accounted for to assess the efficiency of the chemical agents.

In this work, an attempt to close gaps between micromodels and reservoir rocks was performed by constructing chips based on the X-ray micro-computed tomography (μCT) images of a Bentheimer core plug.

This paper investigates the additional oil recovery associated to viscoelastic flow instabilities encountered during polymer flooding. Single and two-phase polymer EOR experiments were conducted in micromodels that resemble porous media. To set a benchmark for non-viscoelastic flooding processes, Polystyrene Oxide (PEO) experiments are presented as well. The experimental workflow consists of three main steps. First, saturation of the micromodel with a synthetic oil. Second, displacement of synthetic oil by an aqueous PEO solution. Third, displacement of the remaining oil by a viscoelastic polymer solution. For evaluation purposes, viscosity of the polymer and polystyrene oxide solution are approximately matched. Furthermore, tracer particles are attached to the aqueous phase to enable high quality streamline visualization. The streamline data is gathered using a highspeed camera mounted on an epifluorescence microscope. In this study we demostrate that viscoelastic flow instabilities are highly caused and influenced by polymer properties. It is also shown flow instabilities dependence on pore space geometry and Darcy's velocity. We have observed a dependency of elastic turbulence on mechanical degradation, polymer concentration and solvent salinity. Furthermore, two-phase flood experiments in complex pore-scale geometries have confirmed that elastic flow inconsistency provides a mechanism capable of increasing oil phase mobilization by the viscoelastic aqueous phase. Due to high resolution particle tracing in the micromodels, the main causes of enhanced mobilization can be described as: (1) Moffatt vortices, (2) crossing streamlines, especially near grain surfaces and (3) steadily changing flow directions of streamlines. Thus, by adding viscoelastic additives to injection fluids and considering a sufficient shear rate, even a creeping flow is able to further enhance the displacement process in porous media by its elastic instabilities. This work provides an adittional understanding of pore-scale polymer displacement processes, namely oil mobilization due to elastic turbulence/flow instabilities. Using the potential of state-of-the-art micromodels enables to conduct high quality streamline visualization which is the key to an improved polymer EOR screening. Thereby enables to understand which properties of viscoelastic solutions contribute to oil recovery. Moreover, this analysis can be used to modify subsequently the fluid characteristics in order to achieve an optimized process application.

Summary Alkali/polymer (AP) flooding of high-TAN oil is a promising enhanced oil recovery (EOR) method. Phase tests reveal that the generated emulsions are thermodynamically unstable macroemulsions rather than Winsor-type emulsions as observed in alkali/surfactant (AS) systems. We investigated the effect of gas on the phase behavior and displacement efficiency of AS systems. The reason is that the impact of gas in solution on the displacement efficiency in alkali projects is significant, neglecting the gas effects underestimates the incremental recovery factor by >15%. Experiments and analysis were performed to investigate the effects of alkali injection into a live and dead high-TAN oil. Viscosity measurements using a capillary rheometer and oscillating u-tube were done to ensure the same viscosity of the dead oil (adding cyclohexane) to live oil. Alkali phase behavior scans were used to determine the amount of emulsions formed initially and over time. The structure and characteristics of the emulsions were investigated using a high-resolution microscope. Micromodel experiments (dead oil only) were performed to elucidate the displacement efficiency effects on pore scale, while flooding experiments showed the displacement efficiency on core scale. Phase experiments showed that initially, a substantial amount of emulsions is formed. The volume of the emulsion is changing over time reaching zero for the live and dead oil. The microscope pictures show that in the initial stage, a “middle phase” macroemulsion is present. With time, the middle phase disappears supporting the results of thermodynamically unstable emulsions seen in the phase experiments. Micromodels show that oil is mobilized by AP injection on a local scale by elongating ganglia and reducing the size of trapped oil and only a limited amount of macroemulsions is formed at the oil/alkali/water interface. The increased oil recovery is thus an effect of the local capillary number and mobilization of ganglia. Here, no stable three-phase system consisting of oil/microemulsion/water as in AS system is generated. Live oil AP corefloods lead to recovery factors of 95% compared with 74% for dead oil. The gas in solution improves the local pore scale sweep efficiency and needs to be included in the evaluation of AP flooding to ensure that incremental oil production is not underestimated for high TAN number oils. The main findings are as follows: Phase experiments of alkali with dead and live high TAN oil show that initially a large amount of emulsions is generated. However, these emulsions are thermodynamically unstable macroemulsions. Micromodel investigations show that the local pore scale displacement efficiency is improved by injecting AP solutions. Gas in solution is substantially improving the local displacement efficiency and needs to be included to correctly determine incremental oil production from AP flooding.