EVALUATING NATURAL AMINO-ACIDS AS EFFECTIVECLAY STABILIZERS FOR WATER-BASED FLUIDS: EXPERIMENTAL AND COSMO-RS SIMULATION STUDIES.

The use of water-based fluids for the hydraulic fracturing of gas shales significantly boosted the production of clean gas. However, during the fracturing process, geochemical reactions often occur between the base fluid and the reactive clay minerals present in shales. This causes the swelling and dispersion of the clay minerals, eventually leading to the destabilizing of the formation. Clay stabilizers are often used to remedy the situation; however, current conventional stabilizers are challenged by cost, molecular weights, and toxicity. This necessitates empirical research for cost-effective, monomeric, and ‘green’ stabilizers. Herein, the clay stabilizing potentials of 4 amino acids (AAs), namely, Alanine, Arginine, Glycine, and Proline, were investigated. Seven experimental techniques were utilized for the evaluation. COSMO-RS simulation was also employed to analyze and model the inhibition effect of the AAs theoretically. Upon experimentation, all the AAs displayed varying clay stabilization abilities with Glycine showing the highest potential. The inhibition efficiency of the AA solutions was compared to the base fluid. Glycine reduced the swelling of the clay by 35%, followed by Proline (23%), Alanine (14%) and Arginine (11%). This was also compared to a KCl inhibiting solution which only showed an 18% reduction in swelling. Glycine and Proline solutions showed improved inhibition than that of KCl. The inhibition potential of AAs is mostly based on their ability to form hyper-stable hydrogen bonds with water and clay minerals. This action limits the uptake of water into the clay, thus reducing swelling. The COSMO-RS theoretical analysis further confirmed that the inhibition performance of AAs is also controlled by the following: (1) the activity coefficient (γ) between AAs with water, (2) the hydrogen bonding energy (EHB) between the AAs and water, and (3) the side-chain length (ps) of the AAs. A multiple regression model was also developed using the predicted COSMO-RS properties (γ, EHB, and ps). The model predictions strongly agreed with the experimental results with percentage errors of less than 0.5%.