In order to design and analyse Alkaline Surfactant Polymer (ASP) pilots and generate reliable field forecasts, a robust scalable modeling workflow for the ASP process is required. Accurate modeling of an ASP flood requires detailed representation of the geochemistry and the saponification process, if natural acids are present. The objective of this study is to extend the existing models of ion exchange and surfactant partitioning between phases to improve the quality of the model.
Geochemistry and saponification affect the propagation of the injected chemicals. This in turn determine the chemical phase behaviour and hence the effectiveness of the ASP process. A starting point of such a workflow is to carry out ASP coreflood tests and history matching (HM) using numerical models. This allows validation of the models and generates a set of chemical flood parameters that can be used for forecasts. The next step is upscaling from lab to field. The presence of (geo)-chemistry in ASP model improves significantly the quality of core HM especially for produced chemicals, breakthrough time and their profiles shape.
The addition of surfactant partitioning between the oleic and the aqueous phases based on salinity of the system as well as propagated distance (time) improves understanding of the required surfactant concentration. The partitioning of surfactant is important for coreflood matching of native cores as they tend to have more clays and minerals that affect ASP phase behaviour. The upscaling of the HM coreflood was conducted in two steps. First step the coreflood was scaled up with the distance between injector–producer pair as the scaling parameter. Second step was the application of the scaled up injection rates, residual saturations, etc. to the full field model. Sensitivity study for parameters such as grid size, well distance, ASP slug size, and rate of surfactant partitioning was performed. It was found that grid size of 50ft was optimum for ASP modeling. The higher rate of surfactant partitioning resulted to lower recovery. The optimal well distance was determined as 700ft for optimization of oil recovery. The reduction of ASP slug size from 0.5PV to 0.3PV leads to the reduction in oil recovery by 2-3%.
Usually chemical reactions accompanied ASP process are left out of the model due to increase in complexity as well as longer computational time. However, their addition as well as presence of surfactant partitioning between the oleic and the aqueous phases makes ASP models more realistic and it results in significant improvement to coreflood HM quality and prediction of ASP process.
Karpan, Volodymyr (Shell Exploration & Production) | Farajzadeh, Rouhollah (Shell Intl E&P BV) | Zarubinska, Maria (Shell Exploration & Production) | Dijk, Harm (Shell Intl E&P Co) | Matsuura, Tsuyoshi (Shell Exploration & Production) | Stoll, Martin (Shell E&P International Ltd)
In order to design and analyze Alkaline Surfactant Polymer (ASP) pilots and to generate reliable ASP field forecasts a robust scalable modeling workflow for the ASP process is required. A starting point of such a workflow is to carry out ASP coreflood tests and history match those using numerical models. This allows validation of the models and generates a set of chemical flood parameters that can be used for field-scale simulation forecasts.
It is well established that lowering of interfacial tension due to maximum of in-situ generated soap with injected surfactant and improved mobility control due to the polymer play a crucial role in the ASP process. Therefore, all models for the ASP process take into account these mechanisms in one way or the other. However, ASP models can differ in the detail in which (geo-) chemical reactions and the phase behavior are addressed. Inclusion of the more details into the numerical model could result in better understanding and more accurate prediction, but it comes at a price, viz., it requires more measured input data and increases computational time. Thus, depending on the accuracy requirements, available experimental data and time the modeling of ASP flood can be performed using different simulation approaches.
This paper describes several modeling approaches for ASP. We start with a brief description of these methods and their input requirements. Then we compare the ASP coreflood simulation results demonstrating the advantages and disadvantages of presented approaches. We also demonstrate that both ASP models can be applied at the field level by simulating an ASP flood in a sector model. Finally we give some recommendations and guidelines on how and when the proposed models should be used.
Alkaline/surfactant/polymer (ASP) flooding is an enhanced oil recovery (EOR) technique that involves the injection of a solution of surfactant, alkaline and polymer into the oil reservoir to mobilize the remaining oil. In this process the injected surfactant and the petroleum soaps generated in situ reduce the oil-water interfacial tension (IFT), improving the microscopic sweep efficiency (Nelson et al., 1984). Moreover, the macroscopic sweep efficiency is enhanced through improvement of the mobility ratio due to the injected polymer. Another important benefit of the alkali is the reduction of surfactant retention on the rock surface, allowing for the injection of smaller amounts of surfactant. Indeed, in some cases where the crude oil does not react with the alkali, the injection of alkali is recommended to prevent surface retention of expensive surfactant. As a further improvement, the addition of a co-solvent may enhance the combined solubility of the surfactant and the polymer in the injected ASP solution and reduce the viscosity of (micro-) emulsions formed when the ASP solution contacts the crude oil.
The ASP process is usually applied to tertiary floods in the drive mode. Because of the considerable costs of the chemicals associated with the ASP flooding, an ASP slug (a fraction of the reservoir pore volume) is generally injected, and then followed by a solution of a water-soluble polymer. Typical estimated incremental recoveries for ASP flooding after water flood are of the order of 10 to 20% STOIIP (Pitts et al., 2006, Qu et al., 1998, Vargo et al., 2000).