Injectivity Decline From Produced-Water Reinjection: New Insights on In-Depth Particle-Deposition Mechanisms

Abstract

Summary Injectivity decline during produced-water reinjection (PWRI) originates not only from filter-cake buildup but also from in-depth deposition of oil droplets or solid particles. Physical modeling of particle-deposition mechanisms in porous media is thus of key interest for optimizing PWRI operations. The present work brings new insights on oil-droplet and solid-particle-deposition mechanisms in porous media. The experimental conditions were selected such that the ratio between pores and particle sizes is sufficiently large to ensure in-depth propagation. The parameters are the nature of the particles injected and a Peclet number calculated on the size of the collector grains (Pg) that encompasses in a nondimensional form the impact of both the flow rate and the particle size. The results are analyzed within the framework of the "colloidal approach." For oil droplets and solid particles, the collection efficiency (?) shows a transition from a behavior in which ? varies as a power law of Pg, with exponent values -? [(diffusion-limited deposition (DLD)] to -1 [reaction-limited deposition (RLD)] that are typical of the convection/diffusion regime, to a behavior characterized by an increase of ? vs. Pg, typical of the hydrodynamic deposition regime. In the case of oil droplets (slightly charged), the transition occurs at a critical Pg value, PgC ˜ PgCgeom/10, corresponding to a diffusion-layer thickness around the collector grain of the same order of magnitude as the droplet diameter. In the case of electrosterically stabilized solid particles, the transition takes place at PgC << PgCgeom for small particles and at PgC > PgCgeom for larger particles. Introduction As the environmental regulations for water discharge are becoming increasingly stringent, PWRI is now recognized as an important issue with respect to environmental protection and profitability. In many mature zones, PWRI is becoming an economically attractive option. It enables reduction in processing costs of the produced water before disposal and allows handling the ever-increasing volumes of produced water without increasing environmental risks associated with water discharges. PWRI also can be used for waterflooding or pressure maintenance in the initial stage of production. Therefore, most of the produced water has to be reinjected either in a suitable formation for disposal or in the producing formation for improved-oil-recovery purposes. However, great uncertainties still remain about the consequences of PWRI and the actual injectivity behavior. If injectivity decline is severe, sustainable injection rate for extended periods will become impossible, jeopardizing the entire reinjection operation. From field observations, it was acknowledged that:With produced water, plugging is important and injectivity is lower than expected. The solids and the oil do not behave separately.With matrix PWRI, a continuous loss of injectivity is obtained, even in high-permeability formation (soft rock) (Detienne and Po 2005).Successful PWRI is likely to require fracturing (Detienne and Po 2005; Raaen 2005; van den Hoek and Bjoerndal 2005; Sweeney 2005). Thus, the injectivity performance is viewed increasingly as being dictated by a dynamic coupling between fracture growth and plugging of fracture faces. This general scheme raises additional challenges related to the issues of fracture containment, sweep efficiency, and conformance, especially in soft-rock formations. Simulation of PWRI under fracturing conditions is therefore necessary to optimize the injectivity behavior and to establish a PWRI strategy (Detienne et al. 2005; Ochi et al. 1999; van den Hoek et al. 1996). However, until now, development of models with reliable prediction ability suffers from the lack of a clear insight into the actual physics of the damage process. Understanding the damaging mechanisms when injecting water containing solid particles and oil droplets and evaluating the impact of such damage on injector performance are, thus, important research needs. PWRI-induced damage results from in-depth particle penetration into the surrounding formation and from particle accumulation on formation/fracture face to form a heterogeneous and highly compressible filter cake. The filter-cake permeability is among the important parameters that have a great impact on injectivity. Hence, a representative estimation of this parameter is required to forecast the injection behavior (van den Hoek et al. 1996). More generally, the impact of particles, including mixtures of oil and solids, on injectivity has received increasing attention recently (van den Hoek et al. 1996; Al-Abduwani et al. 2003; Al-Riyamy and Sharma 2002; Bedrikovetsky et al. 2001; da Silva et al. 2004; Hofsaess and Kleinitz 2003; Sæby et al. 2005). Regarding oil-in-water-emulsion flow and deposition in porous media, despite significant experimental and theoretical work (Soo and Radke 1984, 1986; Soo et al. 1986), impact on injectivity is still not understood clearly, especially under severe injection conditions (high flow rate into low-permeability formation). The objective of this study was to bring new insights into the impact of flow rate on in-depth deposition mechanisms of the colloidal particles present in produced water. The present work was focused on two kinds of particles:uncharged oil droplets in dilute and stable oil-in-water emulsions andelectrosterically stabilized latex microspheres. For both kinds of particles, a broad range of velocities was investigated. In the next section, some useful results regarding the theoretical background of the colloidal approach are presented. In the section after that, materials and experimental procedures are described. Experimental results are presented and discussed in the last section.

Modified Particle Detachment Model for Colloidal Transport in Porous Media

“…The particle capture mechanisms by the grain-particle and particle-particle attraction are assumed, so other mechanisms like bridging, Brownian diffusion, gravity segregation and external cake formation are not captured by the proposed model (Nabzar et al 1996;Chauveteau et al 1998;Tufenkji and Elimelech 2004;Rousseau et al 2008). Low concentration suspensions with constant fluid density and viscosity are considered, so the buoyancy effect of the dense suspension (Ilina et al 2008) is neglected.…”

“…Equations 1 and 2 together with the micro-scale-modelling-based formula for coefficient λ are called the classical filtration theory in the above references. The advanced theory for the filtration coefficient dependency on particle-grain and particle-particle interactions, flow velocity, Brownian diffusion and gravitational sedimentation was developed (see Nabzar et al 1996;Chauveteau et al 1998;Tufenkji and Elimelech 2004;Rousseau et al 2008), while the detachment coefficient is an empirical constant usually determined by tuning with the experimental data.…”

Particle Detachment Under Velocity Alternation During Suspension Transport in Porous Media

“…The impedance behaviour can be interpreted as a result of two simultaneous particle capture mechanisms (see Nabzar et al 1996;Chauveteau et al 1998;Rousseau et al 2008;Guedes et al 2009;Gitis et al 2010). The bulk mechanism is highly velocity-dependent while the weak capture is independent of flow velocity.…”

“…The detachment rate is assumed to be proportional to the difference between the current and critical parameters like velocity, concentrations, etc. The detailed theory for dependence of the capture kinetics (filtration) coefficient on the pore scale parameters for particle attachment was developed (Nabzar et al 1996;Chauveteau et al 1998;Rousseau et al 2008, see also references in Tufenkji and Elimelech 2004), while the detachment kinetics coefficients are the empirical constants usually determined by its tuning with the breakthrough concentration (Tufenkji 2007).…”

Adsorption and Removal of Organic Dye at Quartz Sand-Water Interface

“…The Peclet number, P e , encompasses in a dimensionless form, the impact of both the flow rate and the particle size, on the MB adsorbed mass on the quartz sand surface [37]. Moreover, it was shown that the collection efficiency, g, defined for a spherical collector, as the ratio between, the flow of particles retained by the collector and the incident flux, is related to P e by a power law having an exponent of À2/3 or À1, depending on whether the deposit is limited, respectively, by diffusion (DLD) or by reaction (RLD) [37]. Thus, according to these power laws, the fraction of MB retained on the quartz surface decreases when the flow rate, Q, increases.…”