ACID Twin: A Physics-Informed Reduced-Complexity Digital Twin Prototype for Candidate Screening and HCl Concentration Evaluation in Carbonate Matrix Acidizing

Carbonate acid stimulation design requires an early evaluation of reservoir suitability, HCl concentration, operational risk, and expected productivity response before laboratory testing, high-fidelity simulation, or field execution. This study presents ACID Twin, an offline, scenario-driven, physics-informed reduced-complexity digital twin prototype for screening carbonate acid stimulation candidates. The framework integrates thermal correction, simplified acid–rock reaction kinetics, dimensionless acidizing descriptors, mineralogy-based suitability scoring, productivity index response, skin factor response, corrosion risk estimation, Treatment Success Indicator (TSI) evaluation, and screening-level techno-economic evaluation. A base-case high-calcite carbonate scenario and a candidate reservoir library were evaluated using the HCl concentration cases of 12%, 15%, 20%, and 28%. The results show that the high-permeability limestone candidate and the base-case scenario are the most favorable screening cases, with Acidizing Suitability Index values of approximately 0.60 and 0.59, respectively. The base-case scenario gives a productivity index ratio of 1.42, a Treatment Success Indicator (TSI) of 0.614, and a five-year discounted NPV proxy of 67.6 MMUSD. Increasing HCl concentration raises the simulated productivity index ratio from 1.34 to 1.72 and the NPV proxy from 53.9 to 117.1 MMUSD, but it also increases the corrosion risk indicator from 0.406 to 0.663. Sensitivity maps show that acid volume, injection rate, contact time, salinity, and clay fraction strongly influence screening outcomes. The proposed framework should be interpreted as an offline simulation-based R&D prototype for decision support, not as a continuously synchronized, field-connected, or field-certified forecasting system. The literature-derived benchmark suite and a five-case HCl core flood comparison provide laboratory-scale plausibility checking and calibration-oriented model assessment; however, no independent field validation is claimed. Calibration followed by blind validation against historical field acidizing treatments is therefore required before operational deployment.