Brine Mixing in Layered Reservoirs; Prediction and Impact on Scaling Risks in EOR

Abstract

Prediction of produced water composition will help improve the assessment of scale risks, especially for some methods of chemical EOR. Our aim is to predict the mixing of chemicals and brines within layered reservoirs including the effects of fluid viscosity/mobility ratio. We show the potential impact (ie. benefit) of such mixing, leading to reservoir stripping of scale-forming ions. A new analytical model is presented which predicts an effective (Fickian) dispersion coefficient, directly relating to fluid composition. The effects of layer thickness and fluid mobility ratio are included as well as other important factors (permeability/porosity contrast, flow rate, flow length, transverse dispersion coefficient). The produced water composition profiles are predicted relating to each streamline so that compositions at the well completion can be examined. This information is used to show how different water compositions are produced and mixed within the well so that scaling potentials can be realistically assessed. Our results show that mixing within the reservoir potentially has a great influence on produced water composition, and hence scaling potentials, if the in-situ brine and injected solution are incompatible. Examples include the Alkaline-Surfactant-Polymer flooding process, whereby a highly alkaline solution is injected and the presence of even moderate levels of Ca/Mg ions within the resident brines gives rise to very high scaling potentials. Here, scaling risks at the production well are reduced by fluid mixing within the reservoir, because lower divalent ion concentrations are available for reaction following precipitation within the formation. Reservoir heterogeneity in the form of layering on the fine scale (several cm to fractions of a metre thickness) has a great influence on mixing. Fluid mobility ratios are also important and so are included in the model, with the effect of viscous fluids (eg. polymer solutions) predicted. The calculation method, proven by laboratory flow tests within layered packs, can predict solution compositions flowing within the reservoir and produced at well completions. In combination with reservoir simulation and scaling predictions, modified waterflooding and chemical EOR schemes can be better designed, also production well scale management can be improved.