Effect of KCl/NaCl Composition on Fines Migration Initiation: Experimental Study and DLVO Analysis

Abstract

Fines migration is a major cause of formation damage in sandstone reservoirs. During water injection, changes in brine chemistry can destabilize naturally occurring fines such as kaolinite, clay fragments, and fine quartz. As a result fines detach, migrate with the fluid, and accumulate in pore throats. This accumulation leads to reductions in permeability, increase in pressure drop, and decline in injectivity and hydrocarbon recovery. Since sandstone reservoirs account for a large share of global hydrocarbon production, controlling fines migration remains an important reservoir engineering challenge. The onset of fines migration depends on the physicochemical stability of the sand-fine-brine (SFB) system. Grain surfaces and fines are usually negatively charged in quartz-rich sandstone, and their interaction is governed by the balance between attractive van der Waals and repulsive electrical double-layer (EDL) forces. Fines tend to detach from the rock surface when repulsion becomes dominant over attraction forces. This behavior is described by DLVO theory which connects this interaction energy to salinity, brine composition, zeta potential, and particle size. Among these factors, salinity and brine composition are especially important as they control EDL thickness. Lower salinity increases repulsion, and below the critical salt concentration (CSC), fines migration starts. This study investigates the effect of NaCl/KCl brine composition on the critical salt concentration (CSC) for fines migration initiation in Berea sandstone. Approach used combines zeta potential measurements, machine learning prediction, DLVO modelling, and coreflooding validation. Zeta potential was measured for mixed NaCl/KCl brines using a Malvern Zetasizer, while the 100% KCl case was predicted using machine learning approach. TabM model was chosen to predict zeta potential values for KCl as it has the best performance among nine tested models (test R2 = 0.6983, test RMSE = 7.79 mV). These values were used as inputs to a low- rate DLVO model, which predicted a CSC of 0.03 M for KCl brines. Coreflooding experiments on a Berea sandstone core samples using KCl brines of decreasing concentration (0.20 M-0.01 M) were carried out to validate the DLVO prediction for 100% KCl brines. UV-Vis absorbance analysis of the effluent and pressure drop analysis indicated that a significant increase in fines concentration occurred between 0.03 M and 0.04 M KCl, placing the experimentally observed CSC in the range between 0.03 and 0.04 M. This result is in close agreement with the DLVO-predicted value of 0.03 M that confirms the reliability of the modeling approach. Since 100 % KCl case validated DLVO Model predictions, further models with different compositions of NaCl/KCl brines were constructed. DLVO modeling predicted CSCs of 0.05 M for 80/20 NaCl/KCl, 0.045 M for 50/50, and 0.04 M for 20/80, revealing a clear decrease in CSC with increasing KCl content. The findings show that brine ion composition, even within monovalent systems, has a strong effect on fines stability. Systems dominated by KCl are more likely to experience fines mobilization than systems dominated by NaCl at the same salinity, which has practical importance for injection water design and formation damage control in sandstone reservoirs.