Wall Condensation Heat‐Transfer Prediction for Energy Systems: A Mechanism‐Based Critical Review of Correlations, CFD and Physics‐Informed Models

ABSTRACT

Wall‐condensation heat‐transfer prediction is essential for condenser design, yet model reliability depends on transport mechanisms, operating conditions, and validation quality. This structured critical review evaluates empirical correlations, computational fluid dynamics, and data‐driven and physics‐informed models across internal channels, external steam–gas surfaces, and plate heat exchangers. Published evidence is assessed using source‐specific prediction targets, error metrics, sample sizes, and validation procedures, avoiding unsupported cross‐study accuracy rankings. A convolutional neural network achieved a reported mean absolute relative deviation of 5.82% on 1213 test points, although source‐level independence remains unverified. For steam–non‐condensable‐gas condensation, physics‐constrained XGBoost reported 11.22% mean absolute percentage error versus 21.63% for a comparator on 199 external test points. An illustrative cylindrical thermal‐resistance calculation demonstrates how heat‐transfer‐coefficient variations influence condenser area while highlighting geometry and pressure‐drop constraints. The proposed mechanism‐ and validation‐aware framework supports evidence‐based model selection, identifies transferability limitations, and guides preliminary thermal design without assuming universal predictive superiority.