This repository aims to be a collaborative project to study scenarii around the possibility of running Europeans country electricity supply only on renewables
http://wattif.pythonanywhere.com
A flask/reactjs application for evaluating the impacts of electrical system setup for european countries
- Only onshore wind and solar are considered as production units
- Storage is modelled in a generic way but mose of assumptions and costs are derived from Lithium-Ion batteries
At this stage contributions would ideally take the form of notebooks An example/base jupyter notebook exploring the particular case of France in year 2015 can be found in https://github.com/nikkozzblu/100-percent-renewables/blob/master/notebooks/Impacts%20of%20100%25%20renewables.ipynb
See https://docs.anaconda.com/anaconda/install/
Activate conda virtual default environement (or create a specific one) and install Pulp linear optimizer
> conda activate
> pip install pulp
Clone this repository and download the following dataset in the data directory:
- Hourly Consumption Data (ENTSOE): https://docstore.entsoe.eu/Documents/Publications/Statistics/Monthly-hourly-load-values_2006-2015.xlsx
- Hourly Wind Capacity Factor (EUROPA-EMHIRES): http://setis.ec.europa.eu/sites/default/files/EMHIRES_DATA/EMHIRES_WIND_COUNTRY_June2019.zip
- Hourly Solar Capacity Factor (EUROPA-EMHIRES): https://setis.ec.europa.eu/sites/default/files/EMHIRES_DATA/Solar/EMHIRESPV_country_level.zip
Unzip them in the data directory
Start the jupyter notebook engine
> jupyter notebook
Open the "Load data" notebook and run all the cells. If everything goes smoothly you should now have 275 CSV files in your "data" directory combining the 3 dataset by country and by year
You can now play with the "Impacts of 100% renewables" notebook and try it for different countries/years
This is just an initial step for this project that I run on my free time. Contributions are really welcomed in the form of Notebooks at this stage to:
- Simulate other sources of renewables production units:
- Hydro
- Biomass
- Offshore wind
- Solar thermal
- Geothermal
- Tidal
- Waves
- Refine the grid cost estimate
- Cross borders exchanges
- EROI costs in terms of energy investment
- Introduce new possibilities of assumptions on consumption:
- Flexible loads
- Electric cars fleets
- Add more datasets...
The next step for me would be to package those functions and interface it in the form of a minimalist web app to allow public to try various scenarii based on different country/year/cost assumptions