Abstract Mathematical models of industrial processes have long been used to better design and operate facilities. Traditional applications of such models have been limited in scope and usage due to the rapid rise in complexity, execution time, and difficulties encountered as model size increases to encompass sufficient fidelity and scope to address overall economic impact to business units. Simultaneous solution techniques, referred to as equation-oriented (EO) modeling, as opposed to sequential modular (SM) approaches, have addressed many of the issues that previously limited model fidelity and scope across multiple scales from delivering actionable, value-adding results in a frequent, day-to-day, or even hour-to-hour time frame. Models of alkanolamines-based carbon dioxide (CO 2) capture facilities exemplify the ability of EO modeling to include the best high fidelity, multiscale, mechanistic models along with sufficient scope to optimize operations, allowing economic trade-offs among plant throughput and solution regeneration costs, in the context of the larger process that the CO 2 capture system serves. These models span scales from physical and chemical properties of CO 2 and absorbent molecules, process equipment or units, process plants, to site complexes. A steady-state flowsheet model of the CO 2 absorption and solution regeneration system is illustrated and discussed, both in a parameter estimation mode, elucidating system performance from observed plant data, and in an optimization mode, honoring operating constraints, reflecting control system configuration, while maximizing operating profit. The model is also able to help identify and quantify debottlenecking alternatives. The topics of model robustness, accuracy, and execution speed are covered as well. This application illustrates that integrated, high fidelity, multiscale models from molecular level to site-wide complex can be deployed in nonideal online environments to deliver benefits and insight that cannot be elucidated with simpler, less rigorous, more empirical models.