Mineral Precipitation in Fractures: Using the Level‐Set Method to Quantify the Role of Mineral Heterogeneity on Transport Properties

Abstract

We used a depth‐averaged reactive transport model to simulate transport of a supersaturated fluid through fractures and considered two models of precipitation‐induced surface alterations: (1) a localized 1‐D alteration of the surface and (2) a 3‐D alteration of the surface using the level‐set method. Comparing simulation results to a simple analytical solution for precipitation in a homogeneous (aperture and mineralogy) fracture demonstrated both models predict the alteration process reasonably well. However, comparing simulation results from both models to results from a recent experimental study of precipitation in a mineralogically heterogeneous fracture (Jones & Detwiler, 2016, https://doi.org/10.1002/2016GL069598) demonstrated that the ability of the 3‐D model to predict mineral growth across the aperture and in the fracture plane leads to much better agreement with the experimental observations. To quantify the role of reaction kinetics on the precipitation process, we ran additional simulations using the 3‐D evolution model for reaction rate constants ranging over 3 orders of magnitude. When the kinetics were much faster than the advective flux through the fracture, precipitation was focused near the inlet, which led to a transition from preferential flow near the inlet to more uniform flow near the outlet. When reaction kinetics were slow relative to advection, reaction sites grew uniformly throughout the fracture leading to the eventual formation of a single preferential flow path from inlet to outlet. Regardless of reaction rate, localization of precipitation reactions led to significant increases (3 to 20 times) in the time scale required to seal the fracture relative to a mineralogically homogeneous fracture.

Permeability Impairment and Salt Precipitation Patterns During CO2 Injection Into Single Natural Brine‐Filled Fractures

“…In our model, the solubility limit of NaCl is set to 26.84% by weight at 50°C(Haynes, 2013). The coefficient r m is set to a high value to mimic quasi‐instantaneous precipitation (Jones & Detwiler, 2019; Roels, 2015). Therefore, salt precipitation occurs whenever the halite concentration in the brine reaches the solubility limit under the temperature and pressure conditions of the fluid.…”

“…These three zones are illustrated in Figure 1(Burton et al., 2008; Ghafoori et al., 2017; Nooraiepour et al., 2018). Conversely, homogeneous salt precipitation, or nonlocal precipitation, has also been reported during the injection of CO 2 when reaction kinetics are slow relative to advection rates (Jones & Detwiler, 2019; Kim et al., 2012; Roels et al., 2014). Nevertheless, the local or nonlocal attribute of salt precipitation and its impact on reservoir permeability seem to depend on the drying regime , which is ultimately associated with the scCO 2 injection rate (Carpita et al., 2006; Kim et al., 2012; Miri & Hellevang, 2016; Ott et al., 2013).…”

“…More commonly, however, studies of mineral precipitation/dissolution‐induced fracture surface alterations are performed using synthetic fracture geometries or hydraulic conductivity fields (Cheung & Rajaram, 2002; Chaudhuri et al., 2012; Nissan & Berkowitz, 2019; Wang & Cardenas, 2019). Another common approach in studies of mineral precipitation/dissolution‐induced fracture surface alterations is to use fracture geometries that have been mapped through techniques that disregard the effective normal stress that the fracture experiences under in situ conditions (Detwiler & Rajaram, 2007; Glassley et al., 2002; Jones & Detwiler, 2019; Yasuhara & Elsworth, 2006; Vogler et al., 2018). Such deficient characterizations of fracture aperture fields may lead to imprecise investigations of formation dry‐out processes in fractures.…”

“…This change in porosity is then extended to a change in permeability, which is frequently updated according to analytical formulations describing k ‐ ϕ relationships (Carman, 1939; Saar & Manga, 1999; Verma & Pruess, 1988; Walsh & Saar, 2008), as discussed in a previous paragraph. Besides the approach of solid phase formation, alternative approaches to altering the porosity, or the fracture space, due to mineral precipitation/dissolution include models of 1‐D and 3‐D surface alterations (Jones & Detwiler, 2019; Starchenko et al., 2016), models of alterations of the hydraulic radii, proportionally to the amount of precipitate (Liu, Zhang, et al., 2013), and level set methods that capture the geometric evolution of the solid surface (Li et al., 2010). These approaches frequently employ simplifying assumptions, such as uniform mineral deposition and absence of the effects of crystal nucleation or heterogeneous surface mineralogy.…”

“…Despite a wide range of previous research on mineral precipitation due to formation dry‐out, still very few numerical studies exist that investigate the effect of salt precipitation on the hydraulic properties of heterogeneous, rough‐walled, natural fractures under reservoir conditions, that is, with realistic effective stresses and temperatures. Studies have been performed employing simplistic models of fractures (Borgia et al., 2013; Jones & Detwiler, 2019; Xu & Pruess, 2004) or with random aperture fields (Oh et al., 2013), which might not properly represent some features of flow and transport, such as flow channeling within fractures. Additionally, these models may fail to represent some important processes occurring during formation dry‐out, such as the influence of the fluid flow rate on the location of mineral (e.g., salt) precipitation.…”