🧪 Distillation Column Simulator

Advanced multi-stage continuous distillation column simulator with NRTL thermodynamic model and optimized numerical solvers.

🎓 Academic Origin: Originally developed as a student project at ENSGTI (École Nationale Supérieure en Génie des Technologies Industrielles) in 2018. Refactored and modernized for production use with web interface and comprehensive documentation.

---

✨ Features

🔬 Rigorous Scientific Models

• MESH Equations: Material balance, Equilibrium, Summation, Heat balance
• NRTL Model: Non-Random Two-Liquid theory for non-ideal mixtures
• Vapor-Liquid Equilibrium: Activity coefficient calculations
• Enthalpy Calculations: Rigorous heat balances

🧮 Advanced Numerical Methods

• Newton-Raphson solver for non-linear systems
• Block Tridiagonal Matrix optimization (MRSL21 algorithm)
• Full Matrix solver (MRSL01) for comparison
• Analytical Jacobian with numerical perturbations

💻 Modern Architecture

• High-Performance Backend: Fortran 90 for computational efficiency
• Python Wrapper: Easy integration with modern workflows
• Interactive Web UI: Streamlit-based interface with real-time visualization
• YAML Configuration: Human-readable simulation setup

---

🚀 Quick Start

Prerequisites

# System requirements
- gfortran (GCC Fortran compiler)
- Python 3.8+
- uv (Python package manager)
- Make

# Check installations
gfortran --version
python3 --version
uv --version

📦 Using uv: This project uses uv for fast, reliable Python dependency management. Install with: curl -LsSf https://astral.sh/uv/install.sh | sh (see docs/UV_GUIDE.md for details)

Installation

# Clone repository
git clone https://github.com/yourusername/distillation-column-simulator.git
cd distillation-column-simulator

# Install Python dependencies with uv (recommended)
uv sync

# Or use pip (alternative)
pip install -r frontend/requirements.txt

# Compile Fortran backend
cd backend
make
cd ..

Run Web Interface

# With uv (recommended)
uv run streamlit run frontend/app.py

# Or activate venv first
source .venv/bin/activate  # Linux/macOS
streamlit run frontend/app.py

The application will open in your browser at http://localhost:8501

Run from Command Line

# Compile backend
cd backend && make

# Edit configuration
nano ../config/simulation_config.yaml

# Run simulation
./distillation

# Results saved to RESULTS.csv

---

📊 Example: Acetone-Benzene-Chloroform Separation

column:
  n_stages: 30
  pressure_atm: 1.0

feeds:
  - stage: 14
    flow_rate: 1.0
    composition:
      acetone: 0.6
      benzene: 0.3
      chloroform: 0.1

thermal:
  reboiler_duty: 21000.0    # cal/s
  condenser_duty: 15000.0   # cal/s

System Features:

• Ternary mixture with azeotrope (Acetone-Chloroform at 66/34%)
• Non-ideal behavior (NRTL model essential)
• 30 theoretical stages
• Typical convergence: ~1700 iterations, 50 seconds

---

📁 Project Structure

distillation-simulator/
├── backend/                  # Fortran computational engine
│   ├── src/
│   │   ├── modules/         # Global variable modules
│   │   ├── thermodynamics/  # NRTL, equilibrium, enthalpies
│   │   ├── solvers/         # MRSL01, MRSL21 linear solvers
│   │   └── utils/           # Utilities (norms, residuals)
│   └── Makefile             # Build system
│
├── python/                   # Python wrapper package
│   └── distillation_wrapper/
│       ├── fortran_interface.py
│       ├── config_manager.py
│       └── data_processor.py
│
├── frontend/                 # Streamlit web application
│   ├── app.py               # Main application
│   └── components/          # UI components
│
├── config/                   # Configuration files
│   ├── simulation_config.yaml
│   ├── components.yaml
│   └── examples/
│
├── docs/                     # Documentation
│   ├── scientific/          # Theory and equations
│   ├── technical/           # API and usage
│   └── original/            # Original 2018 project docs
│
└── results/                  # Simulation outputs

---

📚 Documentation

Scientific Background

• Thermodynamic Models - NRTL theory and implementation
• MESH Equations - Material and energy balances
• Numerical Methods - Newton-Raphson and matrix solvers

Technical Documentation

• Installation Guide
• Configuration Reference
• API Documentation
• Python Wrapper Usage

Original Project (2018)

• Project Statement
• Full Technical Report (39 pages, French)
• See branch archive/ensgti-2018 for original unmodified code

---

🔧 Development

Compile Backend

cd backend
make clean && make

Run Tests

# Python tests
pytest tests/python/

# Fortran validation
cd tests/fortran && ./run_validation.sh

Code Style

• Fortran: Fortran 90 standard, detailed comments
• Python: PEP 8, type hints, docstrings
• Git: Conventional commits

---

📈 Performance

Solver	Method	Avg. Time	Iterations	Memory
MRSL21	Block Tridiagonal	~20s	~1700	Low
MRSL01	Full Matrix	~26s	~1700	Higher

Benchmark: 30 stages, acetone-benzene-chloroform, convergence 10⁻⁶

---

🎓 Academic Background

Original Project (2018)

• Institution: ENSGTI Pau, France
• Course: Modélisation des opérations unitaires II
• Instructor: Pr. Frédéric MARIAS
• Student: Younes AJEDDIG

Modernization (2025)

• Refactored for production use
• Added web interface
• Comprehensive documentation
• CI/CD integration

---

🤝 Contributing

Contributions welcome! Please see CONTRIBUTING.md for guidelines.

1. Fork the repository
2. Create feature branch (git checkout -b feature/amazing-feature)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open Pull Request

---

📝 License

This project is licensed under the MIT License - see LICENSE file for details.

---

🙏 Acknowledgments

• ENSGTI for the original academic framework
• Frédéric MARIAS for scientific guidance (2018)
• Open-source community for tools and libraries

---

📧 Contact

Younes AJEDDIG

• GitHub: @yajeddig
• LinkedIn: Younes AJEDDIG
• Email: younes.ajeddig@gmail.com

---

---

Made with ❤️ for Process & Chemical Engineering