Python-Based Algorithm for Calculating Physical Properties of Aqueous Mixtures Composed of Substances Not Available...

Area of Science:

• Chemical Engineering
• Computational Chemistry
• Thermodynamics

Background:

• Accurate prediction of aqueous physical properties is crucial for chemical process design and simulation.
• Existing models in commercial software like AspenTech may lack parameters for all relevant substances.
• Open-source solutions can enhance the flexibility and accessibility of property prediction tools.

Purpose of the Study:

• To develop and validate a Python-based open-source algorithm for predicting aqueous physical properties.
• To ensure compatibility with AspenTech's electrolyte property models.
• To provide an alternative for estimating parameters not available in commercial databases.

Main Methods:

• Developed a Python algorithm utilizing pure component property predictions and NRTL parameters from UNIFAC simulations.
• Validated the algorithm against experimental data for 37 binary aqueous mixtures.
• Compared predicted properties (density, heat capacity, viscosity, thermal conductivity) with experimental values.

Main Results:

• Achieved Mean Absolute Percentage Errors (MAPE) of 2.88% for density, 0.355% for heat capacity, 12.1% for viscosity, and 10.1% for thermal conductivity.
• Identified limitations in accurately reflecting density and viscosity trends at high concentrations for certain substances.
• Demonstrated that inaccurate viscosity and thermal conductivity predictions impacted heat transfer coefficient calculations in a falling-film evaporator simulation.

Conclusions:

• The developed algorithm offers a valuable tool for estimating initial physical property parameters for models not present in commercial databases.
• Caution is advised when using the algorithm for high-concentration systems or for properties like viscosity and thermal conductivity due to potential significant errors.
• Further refinement may be needed to improve accuracy for specific challenging conditions and properties.