Abstract
Waterflooding is a common recovery method used to maintain reservoir pressure and improve reservoir oil sweep efficiency. However, injecting cold water into a reservoir alters the state of in-situ formation stress and can result in the formation fracturing. In other words, it can cause the initiation and growth of thermally induced fractures (TIFs), even when the original fracture propagation pressure is not exceeded. TIFs can cause non-uniform distribution of the fluid flow in wellbores, a reduction in sweep efficiency, and early water breakthrough in nearby production wells. Modelling and history matching workflows that consider the dynamic nature of the TIF problem are critical. These workflows improve and validate reservoir and geomechanical models, identify and confirm observed TIF onset and propagation periods, and provide a history-matched sector model with the rock mechanical and thermal properties and stress gradients that can be used with confidence for subsequent studies. Modelling and the underlining assumptions of fluid flow in the TIF and reservoir matrix, as well as geomechanical changes due to cooling of the reservoir during injection, are detailed below. A 3D reservoir simulator coupled with 2D finite element TIF and geomechanical models were used to manually history match an injector (NI6) in the N Field sector reservoir model in which a TIF was observed. In this study, history matching workflows were developed to consider the dynamic nature of TIF development during waterflooding. The reservoir and geomechanical models were improved and validated via the observed TIF onset and propagation periods. The history-matched models produced can be used with confidence in subsequent studies. The practical workflows and guidelines developed here can be used in waterflooding operations during the modelling, design, and planning stages. The novelty of this study is the coupling approach of different complex processes done in order to capture dynamic changes during waterflooding operations. A similar history matching study could not be found in the literature.