Scenarios outputs of 'CO2InnO-H2-CHP-Demonstrator' Open Modelica TEA model

This repository contains datasets representing the numerical outputs of a TEA (Techno-Economic Assessment) OpenModelica model used for the study of decentralized energy systems (DES) based on hydrogen use (H₂) and using H₂ in an internal combustion engine for combined heat & power (ICE-CHP). 

The TEA OpenModelica model is hosted on Github under the name "CO2InnO-H2-CHP-Demonstrator".

The datasets in this repository come with the Jupyter Notebooks used to process them in order to extract relevant data for a LCA (Life-Cycle Assessment) study and interactively generate relevant visualizations. The results dataset of the LCA and the dedicated processing notebooks are hosted in a companion Zenodo repository.

The scientific article link will be added once it is accepted. 

Context

The global aim was to conduct a TEA-grounded LCA of a H₂-ICE-CHP based DES to fill a gap in the contemporary literature, and contribute to the development of open-source approaches for combined TEA-LCA. It is built on a case study of the Offenburg municipality (Baden-Wurttemberg, Germany), which provided real data regarding several buildings energy requirements, with hourly resolution over a year. This municipality is a partner of the CO2InnO project, representative of small to medium cities wishing to decarbonize their electricity and mostly their heat demand, but which will not be connected to the future hydrogen backbone before 2035 at least. A simulated energy system has been developed in OpenModelica (Fritzson et al., 2020) following a TEA approach (Beerlage et al. 2024).

System model boundaries

The investigated system is centered arround the ICE-CHP in a context of heat-lead functionning where electricity exchanges with the grid are possible. It comprises three interconnected subsystems: electrical, thermal, and hydrogen.

Electrical system

• PV - Pholtovoltaic
• WT - Wind turbine
• BAT - Battery

Heating system

• GSHP - Ground source heat pump
• TES - Thermal energy storage

Hydrogen system

• PEMEL - Proton exchange membrane electrolyzer
• H2C - Hydrogen compressor
• H2S - Hydrogen storage

Scenario dimensioning

As the central element, the heating system was dimensioned based on economic optimization. Several scenarios were designed to assess how variations in hydrogen storage (H2S) capacity affect the feasibility and efficiency of the H₂ ICE-CHP-based DES (Beerlage et al., 2024). H2S was selected as a key variable due to its practical and economic constraints, including large volume requirements and high capital costs, which present major barriers in real-world implementation. By varying storage volumes, the study sought to identify an optimal size that balances hydrogen production and consumption without increasing emissions or inefficiencies. Given the system’s heat-oriented design, the primary goal was to meet thermal demand; electricity production via photovoltaic (PV) and wind turbines (WT) was primarily intended to support hydrogen generation rather than grid supply. This focus on heat demand influenced both control strategies and the sizing of key components such as the CHP unit and hydrogen storage. Scenario variables are detailed in the Table below.

Data origin and uncertainties

As part of the CO2InnO project, the city of Offenburg (Germany) provided hourly heat and electricity demand data over one year for five buildings. These buildings were modeled as a single system with shared infrastructure, based on simplifying assumptions agreed upon by project partners. Weather inputs were sourced from the open-source CBA Clima tool (Betti et al., 2024), using data from nearby Strasbourg (France) due to the absence of a dedicated Offenburg dataset. Mass and energy flows derived from simulations depend on quality of the OpenModelica model. Since the system design centers on a ICE-CHP unit and a proton-exchange membrane electrolyzer (PEMEL), these components’ models were validated in the original TEA study (Beerlage et al., 2024). The average relative deviation was ~10% between model and lab measurements for the PEMEL, and 1-4% compared to manufacturer specifications or the Hofner model for the ICE-CHP unit.

Structure & How to use

The OM_xxx zip files contains the outputs of the OpenModelica TEA model in different versions, before and after processing for transparency. The OM_model_parameters xlsx file serves as synthetic explainer regarding the name of the variables present in the datasets and the numerical values of parameters used in the original OpenModelica simulations. The Jupyter-Notebooks zip file contains the different notebooks using Python code to process the files. These notebooks have in-depth internal documentation to help new users. The pickle files contains the different subdatasets used in the whole pipeline. It allows users to only execute some part of the notebooks according to their need.

1. First, the export by OpenModelica lead to anomalities like duplicate rows. The original outputs are thus processed by the om_output_cleaning Jupyter Notebook to produce the cleaned outputs, and then can also be used to produce the extracted outputs which are used in the second step.
2. The extracted outputs can be processed with the balance_of_flows notebook to derive the flows' allocation between subsystem components (indispensable for LCA study).
3. The sankey_diagrams_only notebook is an independant version of the end of the balance_of_flows notebook, dedicated to flow visualization purposes.

Note : You should extract all the zip files in the same working folder in order to keep the relative links between notebooks and data files.

Requirements

Python environment manager

We suggest you to use Miniconda:

Miniconda is a free, miniature installation of Anaconda Distribution that includes only conda, Python, the packages they both depend on, and a small number of other useful packages.

 Miniconda Installation Guide

Setting up the environment

Note: the notebooks were tested on Python 3.11. 

Within your Python environment manager console (like miniconda console, or anything that suits you):

Create a new Conda environment with all required libraries (in this example named `da` for "data analysis"):

conda create  -n da -c conda-forge os pandas numpy tdqm matplotlib seaborn jupyterlab

(Add any required library)

Or, create the Conda environment only:

conda create -n da -c conda-forge

And install the libraries afterwards, after activating the environment:

conda activate da

conda install os pandas numpy tdqm matplotlib seaborn jupyterlab

(Add any required library)

Some libraries may require you to use  the `pip` manager instead of conda. In that case, use the command:

pip install name_of_library

Use the Jupyter Notebooks

1.  Activate the environment:

conda activate da

2. Launch Jupyter Lab:

jupyter-lab

You can select the Notebook you want to use by navigating in the directory where you downloaded the Notebooks.

They are thoroughly documented to allow you to understand what does each bloc of code.