AlgaeGrowth Simulator

Microalgal growth simulation for defensible CO2 capture estimation under Surat, India's tropical climate.

Goal

The goal of this project is to develop a model of an open pond culture of the algae species Chlorella vulgaris that can be used to estimate carbon sequestration based on Monod's kinetic equations describing algal growth, Steele's equation describing the inhibition of photosynthesis as light intensity increases, and CTMI (Chapman Temperature Model for Inactivation) equations which describe how temperature affects algal growth. All of the model parameters were taken from peer-reviewed articles cited with DOIs to enable reproducibility of the model and verification against the carbon credit standards of Verra, Gold Standard, and India CCTS.

Features

• Monod + Steele growth model with 20-layer Beer-Lambert depth integration
• CTMI cardinal temperature model with day/night split for Surat's 8-40 C range
• Conservative 50% lab-to-field discount — outputs 6-10 g/m2/day, not lab-ideal 20+
• Three-season Surat climate — Dry (Oct-Feb), Hot (Mar-May), Monsoon (Jun-Sep) with monthly PAR, temperature, and cloud cover from WeatherSpark data
• CO2 accounting — species-specific 1.83 g CO2/g biomass conversion factor (Schediwy et al., 2019)
• Harvest cycling — automatic biomass reset at configurable threshold with event tracking
• Interactive Plotly charts with season background bands and harvest markers
• CSV/JSON export with embedded methodology disclaimers for verification workflows
• Full transparency — every parameter links to its DOI source; equations, assumptions, and "not modeled" factors are visible in-app

Architecture

Simulation Flow

Growth Model

Demo

Quick Start

# Clone
git clone https://github.com/vmehtacode/AlgaeGrowth-Simulator.git
cd AlgaeGrowth-Simulator

# Install dependencies (Python 3.12+)
pip install -e ".[dev]"

# Run
streamlit run src/ui/app.py

Opens at http://localhost:8501. Adjust parameters in the sidebar and click Run Simulation.

Project Structure

AlgaeGrowth-Simulator/
├── src/
│   ├── simulation/          # engine.py, growth.py, light.py
│   ├── climate/             # temperature.py, growth_modifier.py, loader.py, surat.yaml
│   ├── models/              # parameters.py (frozen dataclasses), results.py
│   ├── config/              # species_params.yaml (DOI-cited), loader.py
│   └── ui/                  # app.py, sidebar.py, results.py, charts.py, export.py
├── tests/                   # 172 tests across 9 files
├── docs/diagrams/           # SVG architecture diagrams (beautiful-mermaid)
├── .streamlit/              # Theme config (green eco theme)
├── pyproject.toml           # Python 3.12+, dependencies, tool config
└── LICENSE                  # MIT

Scientific Methodology

Growth Model

The simulator combines three peer-reviewed submodels evaluated at each daily timestep:

1. Steele Photoinhibition (light response)

$$f(I) = \frac{I}{I_{opt}} \cdot e^{1 - I/I_{opt}}$$

Peaks at 1.0 when irradiance equals I_opt (80 umol/m2/s for C. vulgaris), declining at higher intensities.

2. Monod CO2 Kinetics (substrate limitation)

$$f(CO_2) = \frac{CO_2}{K_s + CO_2}$$

Half-saturation K_s = 0.5 mg/L dissolved CO2.

3. CTMI Cardinal Temperature (thermal response)

$$\varphi(T) = \frac{(T - T_{max})(T - T_{min})^2}{(T_{opt} - T_{min}) \left[ (T_{opt} - T_{min})(T - T_{opt}) - (T_{opt} - T_{max})(T_{opt} + T_{min} - 2T) \right]}$$

Returns 0 outside [T_min, T_max], peaks at 1.0 at T_opt. For Surat: T_min=8 C, T_opt=28 C, T_max=40 C.

4. Beer-Lambert Depth Integration (light attenuation)

$$I(z) = I_0 \cdot e^{-(\sigma_x \cdot X + k_{bg}) \cdot z}$$

Pond divided into 20 layers; growth rate computed at each layer midpoint and averaged. Avoids 20-50% overestimation from using surface irradiance alone.

5. Combined Growth Rate

$$\mu_{net} = \mu_{max} \cdot f(I_{avg}) \cdot f(CO_2) \cdot \varphi(T_{day}) \cdot D_f \cdot \frac{h_{photo}}{24} - r_m \cdot \varphi(T_{night}) \cdot \frac{24 - h_{photo}}{24}$$

Where D_f = 0.5 (lab-to-field discount) and r_m = 0.01/d (maintenance respiration).

Parameter Sources

Parameter	Value	Unit	Source	DOI
mu_max	1.0	1/d	Schediwy et al. (2019)	10.1002/elsc.201900107
K_s (CO2)	0.5	mg/L	Schediwy et al. (2019)	10.1002/elsc.201900107
I_opt	80.0	umol/m2/s	Razzak et al. (2024)	10.1016/j.gce.2023.10.004
r_maintenance	0.01	1/d	General microalgae literature	—
sigma_x	0.2	m2/g	Schediwy et al. (2019)	10.1002/elsc.201900107
k_bg	0.5	1/m	General open pond literature	—
carbon_content	0.50	g_C/g_DW	Schediwy et al. (2019)	10.1002/elsc.201900107
CO2:biomass	1.83	g_CO2/g_DW	Derived: (44/12) x 0.50	10.1002/elsc.201900107
discount_factor	0.50	—	Schediwy et al. (2019)	10.1002/elsc.201900107
T_min	8.0	C	Conservative lower bound	—
T_opt	28.0	C	Converti et al. (2009)	10.1016/j.cep.2009.03.006
T_max	40.0	C	Converti et al. (2009)	10.1016/j.cep.2009.03.006

Assumptions

• Monod kinetics with single-substrate (CO2) limitation
• Steele photoinhibition (growth declines above I_opt)
• Day/night temperature split (avoids averaging artifacts at 35+ C)
• 50% lab-to-field discount factor (midpoint of 40-60% range)
• Forward Euler integration with 1-day timestep (~2-5% deviation from finer stepping)
• 30-day month approximation for seasonal mapping

Factors Not Modeled

• Nutrient limitation (N, P depletion)
• Contamination and zooplankton grazing
• Pond surface evaporation and water balance
• CO2 gas-liquid transfer kinetics
• Culture crash recovery dynamics
• Light spectrum variation (PAR only)

References

1. Schediwy, K. et al. (2019). Kinetic modeling of Chlorella vulgaris. Eng. Life Sci., 19(12), 935-943. 10.1002/elsc.201900107
2. Razzak, S.A. et al. (2024). Microalgae cultivation for CO2 capture. Green Chem. Eng. 10.1016/j.gce.2023.10.004
3. Converti, A. et al. (2009). Effect of temperature on Chlorella vulgaris growth rate. Chem. Eng. Process., 48(6), 1146-1151. 10.1016/j.cep.2009.03.006
4. Bernard, O. & Remond, B. (2012). Validation of a microalgae growth model. Bioresour. Technol., 111, 386-394. 10.1016/j.biortech.2012.07.022
5. Rosso, L. et al. (1993). An unexpected correlation between cardinal temperatures. Appl. Environ. Microbiol., 59(3), 744-754.
6. Steele, J.H. (1962). Environmental control of photosynthesis in the sea. Limnol. Oceanogr., 7(2), 137-150.

Testing

# Run all 172 tests
pytest

# With coverage
pytest --tb=short -q

# Specific module
pytest tests/test_growth.py -v

Test suite covers growth kinetics, light attenuation, temperature response, climate integration, simulation engine, harvest logic, CO2 accounting, parameter validation, and YAML loading.

Carbon Credit Disclaimer

This simulator estimates CO2 capture based on peer-reviewed growth models and Surat climate data. Results are intended to support carbon credit verification applications, not to represent verified carbon credits.

For actual carbon credit issuance, submit simulation exports alongside site-specific data to an accredited verification body:

• Verra (VCS) -- Largest voluntary carbon market; ISO 14065 verification
• Gold Standard -- WWF-founded; requires 3+ UN SDG contributions
• India CCTS -- Carbon Credit Trading Scheme under BEE

License

MIT License. Copyright (c) 2026 Vatsal Mehta. See LICENSE.

Citations

If you use this simulator in research or verification workflows, please cite:

@software{algaegrowth_simulator,
  title   = {AlgaeGrowth Simulator},
  author  = {Mehta, Vatsal},
  year    = {2026},
  url     = {https://github.com/vatsalmehta2001/AlgaeGrowth-Simulator},
  note    = {Monod + Steele + CTMI growth model for CO2 capture estimation}
}