Home
A computational framework for optimization of waste hear recovery technology that uses packed-bed direct-contact condensers.
This framework compiles and synthesizes experimental and analytical data on heat and water recovery from waste flue gases using randomly packed bed direct-contact condensers. This synthesis automatically detects gaps in existing data to guide future experimental or numerical studies. Ultimately, the processed data serves as a training dataset for a surrogate neural network model designed for engineering and optimizing waste heat recovery (WHR) units.
The framework is at the development stage now. Currently, only the hydraulics segment is available.
-
Pressure drop in randomly packed beds:
-
Cross-validation of existing correlations.
-
Detection of gaps in existing experimental data.
-
Surrogate neural network model.
-
-
Heat-and-mass transfer in randomly packed beds:
-
Cross-validation of existing correlations.
-
Detection of gaps in existing experimental data.
-
Surrogate neural network model.
-
Currently contribution is limited to pressure drop models or raw data. See the corresponding pages for the model implementation template and registration steps. Submit new models via the issue template.
This project is licensed under the MIT License — see the LICENSE file for details.
If you use this framework in your research, please cite as follows:
Nikitin, M. (2026). WHR Optimization Framework (Version milestone). Zenodo. DOI: 10.5281/zenodo.20756923