STRIDE

Sequential Temporal Resource Investment for Depot Electrification

Master's Thesis Project - Technical University of Munich (TUM) Author: Arno Claude Duration: October 2025 - April 2026

---

Overview

STRIDE is a multi-stage optimization framework for electric truck depot infrastructure planning. It extends the REVOL-E-TION model to optimize sequential investment decisions across multiple time horizons (2025-2055) for:

• PV solar panels (renewable energy generation)
• Battery energy storage systems (ESS)
• EV chargers (depot charging infrastructure)
• Grid connections (import/export capacity)

Unlike traditional single-stage optimization, STRIDE implements a myopic sequential approach that better reflects real-world investment decisions under uncertainty, where future technology costs and demand patterns are not perfectly known.

---

Key Features

• Multi-stage planning: Optimizes investments across 5-year intervals (2025, 2030, 2035, 2040, 2045, 2050)
• Technology cost decline: Incorporates projected cost reductions for PV, batteries, and chargers
• CO2 emission constraints: Optional carbon budget limits for decarbonization scenarios
• Financial metrics: NPV, CAPEX, OPEX, LCOE calculations over 25-year project lifetime
• Extensible architecture: Built on REVOL-E-TION's Pyomo/oemof-solph MILP framework

---

Installation

Prerequisites

• Python 3.9+ (tested with Python 3.9.6)
• Virtual environment (strongly recommended)

Setup

1. Clone the repository with submodules:

   git clone --recurse-submodules https://github.com/your-username/STRIDE.git
   cd STRIDE

2. Create and activate virtual environment:

   python3 -m venv venv
   source venv/bin/activate  # On macOS/Linux
   # OR
   venv\Scripts\activate  # On Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Install REVOL-E-TION submodule:

   cd revoletion
   pip install -e .
   cd ..

5. Verify installation:

   python -c "import revoletion; print('REVOL-E-TION installed successfully')"

Important: Virtual Environment

⚠️ All code must be run within the virtual environment. The project dependencies (including Pyomo, oemof-solph, and REVOL-E-TION) are isolated to the venv to avoid conflicts with system packages.

If using PyCharm/Jupyter, configure the project interpreter to use venv/bin/python3:

• PyCharm: Settings → Project → Python Interpreter → Add Interpreter → Add Local Interpreter → Select venv/bin/python3
• Jupyter: Ensure kernel uses venv Python (check with import sys; print(sys.executable))

---

Development Status

• ✅ REVOL-E-TION integration and understanding
• ✅ CO2 constraint implementation and validation
• ✅ Multi-stage optimizer development
• ✅ Visualization module
• 🚧 Case studies and sensitivity analysis (CURRENT)
• 📅 Thesis writing

---

Usage

Running Multi-Stage Optimization

Configs are chained: base.yaml → depot → scenario/sensitivity

# Activate virtual environment
source venv/bin/activate

# Schmid base run
python3 -m multi_stage.main \
    -c configs/base.yaml configs/depots/schmid.yaml \
    -s inputs/schmid/scenarios/base.csv \
    --name schmid_base

# With scenario (optimistic)
python3 -m multi_stage.main \
    -c configs/base.yaml configs/depots/schmid.yaml configs/scenarios/optimistic.yaml \
    -s inputs/schmid/scenarios/base.csv \
    --name schmid_optimistic

# With sensitivity (WACC high)
python3 -m multi_stage.main \
    -c configs/base.yaml configs/depots/schmid.yaml configs/sensitivity/wacc_high.yaml \
    -s inputs/schmid/scenarios/base.csv \
    --name schmid_wacc_high

# Metzger depot
python3 -m multi_stage.main \
    -c configs/base.yaml configs/depots/metzger.yaml \
    -s inputs/metzger/scenarios/base.csv \
    --name metzger_base

Each run creates a self-contained directory in runs/<name>/ with:

• manifest.yaml - Full traceability (inputs, git commit, parameters)
• config_*.yaml - Copies of all configs used
• scenario_template.csv - Copy of input scenario
• revoletion/ - REVOL-E-TION outputs (contained)
• multi_stage_results.json - Aggregated results

Generating Visualizations

# Generate PNG + PDF plots (default)
python3 -m multi_stage.visualize runs/2026-01-07_base_schmid_test_50d

# Generate only PNG
python3 -m multi_stage.visualize runs/<run_name> --png

# Generate only PDF (vector graphics for thesis)
python3 -m multi_stage.visualize runs/<run_name> --pdf

# Generate all formats
python3 -m multi_stage.visualize runs/<run_name> --all

Available plots:

• investment_timeline - Stacked bar chart of new investments per stage
• cumulative_capacity - Line chart of total PV/ESS capacity over time
• cost_breakdown - CAPEX vs NPV per stage
• npv_waterfall - Cumulative discounted NPV
• cost_decline - Technology cost assumptions
• co2_compliance - Emissions vs regulatory limits
• fleet_growth - Fleet size evolution

Architecture Diagram

A full-page TikZ diagram showing data flow between STRIDE and REVOL-E-TION is available:

# Compile with LaTeX
cd multi_stage
pdflatex architecture_diagram.tex

This diagram shows:

• Blue boxes: STRIDE wrapper code (new)
• Green boxes: REVOL-E-TION existing code
• Orange hatched boxes: Modifications to REVOL-E-TION (CO2 constraint, size_existing extensions)
• Arrows: Control flow and data flow between components
• Loop: Stage iteration with capacity carry-forward

---

Folder Structure

STRIDE/
├── configs/                    # YAML configuration files
│   ├── base.yaml              # Complete shared defaults (source of truth)
│   ├── depots/                # Depot-specific overrides
│   │   ├── schmid.yaml        # fleet=84, CO2 baseline
│   │   └── metzger.yaml       # fleet=18, CO2 baseline
│   ├── scenarios/             # Scenario overrides
│   │   ├── pessimistic.yaml   # No CO2 constraint
│   │   └── optimistic.yaml    # 1.5C pathway
│   └── sensitivity/           # Single-parameter overrides
│       ├── wacc_*.yaml
│       ├── pv_capex_*.yaml
│       └── ...
├── data/                       # Research data and source documents
│   ├── grid_co2/              # CO2 emission factors + sources
│   ├── pv_capex/              # PV cost projections + sources
│   ├── ess_capex/             # Battery cost projections + sources
│   ├── financial_params/      # WACC, discount rates + sources
│   ├── depot_Schmid/          # Case study data (Schmid depot)
│   └── .../                   # Other data categories
├── inputs/                     # Scenario input files for REVOL-E-TION
│   ├── schmid/                # Schmid depot
│   │   ├── settings.csv
│   │   ├── scenarios/base.csv
│   │   └── timeseries/
│   └── metzger/               # Metzger depot
├── runs/                       # All run outputs (self-contained)
│   └── <run_name>/
│       ├── manifest.yaml      # Traceability
│       ├── config_*.yaml      # Config copies
│       └── ...
├── multi_stage/               # STRIDE wrapper code
│   ├── main.py
│   ├── config_loader.py
│   ├── sequential_optimizer.py
│   └── ...
└── revoletion/                # REVOL-E-TION submodule

---

Contributing

This is a thesis project and is not currently accepting external contributions. For questions or collaboration inquiries, please contact the author.

---

License

This project uses REVOL-E-TION as a submodule, which is licensed under its own terms. STRIDE-specific code will be licensed separately upon thesis completion.

---

References

• REVOL-E-TION: https://github.com/TUMFTM/REVOL-E-TION
• oemof-solph: https://github.com/oemof/oemof-solph
• Pyomo: http://www.pyomo.org/

---

Contact

Arno Claude Master's Student, Technical University of Munich Thesis Supervisor: Anna Paper (anna.paper@tum.de) Email: arno.claude@googlemail.com