demand response analysis framework (draf) is an analysis and decision support framework for local multi-energy hubs focusing on demand response.
It uses the power of (mixed integer) linear programming optimization, pandas, Plotly, Matplotlib, elmada, GSEE, Jupyter and more to help users along the energy system analysis process.
The software is described and demonstrated in the open-access draf demo paper.
draf runs on Windows, macOS, and Linux.
A draf-version supporting time series aggregation is provided in the dev-TSA branch.

Features

• Time series analysis tools:
  • DemandAnalyzer - Analyze energy demand profiles
  • PeakLoadAnalyzer - Analyze peak loads or run simple battery simulation
• Easily parameterizable component templates:
  • battery energy storage (BES), battery electric vehicle (BEV), combined heat and power (CHP), heat-only boiler (HOB), heat pump (HP), power-to-heat (P2H), photovoltaic (PV), wind turbine (WT), thermal energy storage (TES), fuel cell (FC), electrolyzer (Elc), hydrogen storage (H2S), production process (PP), product storage (PS), direct air capture (DAC), and more.
  • Sensible naming conventions for parameters and variables, see section Naming conventions.
• Parameter preparation tools:
  • TimeSeriesPrepper - For time series data
    • Electricity prices via elmada
    • Carbon emission factors via elmada
    • Standard load profiles from BDEW
    • PV profiles via GSEE (In Germany, using weather data from DWD)
  • DataBase - For scientific data such as cost or efficiency factors
• Scenario generation tools: Easily build individual scenarios or sensitivity analyses
• Multi-objective mathematical optimization with support of different model languages and solvers:
  • Pyomo - A free and open-source modeling language in Python that supports multiple solvers.
  • GurobiPy - The Python interface to Gurobi, the fastest MILP solver (see Mittelmann benchmark).
• Plotting tools: Convenient plotting of heatmaps, Sankeys, tables, pareto plots, etc. using Plotly, Matplotlib, and seaborn.
  • Support of meta data such as unit, doc, src, and dims
  • Automatic unit conversion
• Export tools:
  • CaseStudy objects including all parameters, meta data, and results can be saved to files.
  • Data can be exported to xarray format.

Quick start

1. Install miniconda or anaconda

2. Open a terminal in the directory where you want to place draf in.

3. Clone draf:

   git clone https://github.com/DrafProject/draf
   cd draf
   

4. Create and activate the draf conda environment (conda env create will create a conda environment based on environment.yml which will install the newest versions of the required packages including the full editable local version of draf.):

   conda env create
   conda activate draf
   

5. (OPTIONAL) If the draf environment caused issues, you could install an older but more specific conda environment, e.g.:

   conda env create --file environments/environment_py39all_mac.yml --force
   conda activate draf39
   

6. (OPTIONAL) To use Gurobi (fast optimization), install a valid Gurobi license (its free for academics).

7. Open Jupyter notebook:

   jupyter notebook
   

8. Check if the imports work:

   import draf
   import elmada
   

9. (OPTIONAL) To use the latest electricity prices and carbon emission factors from elmada, request an ENTSO-E API key and set it to elmada:

   # You have to run this Python code only once (it writes to a permanent file):
   import elmada
   elmada.set_api_keys(entsoe="YOUR_ENTSOE_KEY")
   

10. Start modeling. Have a look at the examples.
    Start with the minimal if you want to write your own component.
    Start with the PV example if you want to import existing components.
    For more advanced modeling look at the draf_demo_case_studies.
    Consider the DRAF CHEAT SHEET

Structure

A CaseStudy object can contain several Scenario instances:

Naming conventions

All parameter and variable names must satisfy the structure <Type>_<Component>_<Descriptor>_<Dims>.
E.g. in P_EG_buy_T, P is the entity type (here: electrical power), EG the component (here: Electricity Grid), buy the descriptor and T the dimension (here: time).
Dimensions are denoted with individual capital letters, so <Dims> is TE if the entity has the dimensions T and E.
See conventions.py for examples of types, components, and descriptors.

Contributing

Contributions in any form are welcome!
Please contact Markus Fleschutz.

Citing

If you use draf for academic work please cite the draf demo paper:

@article{Fleschutz2022,
  author = {Markus Fleschutz and Markus Bohlayer and Marco Braun and Michael D. Murphy},
  title = {Demand Response Analysis Framework ({DRAF}): An Open-Source Multi-Objective Decision Support Tool for Decarbonizing Local Multi-Energy Systems},
  publisher = {{MDPI} {AG}},
  journal = {Sustainability}
  year = {2022},
  volume = {14},
  number = {13},
  pages = {8025},
  url = {https://doi.org/10.3390/su14138025},
  doi = {10.3390/su14138025},
}

Publications using draf

• Fleschutz et al. (2023): Impact of Landing Interruptions on the Optimal Design and Operation of Green Hydrogen Hubs
• Fleschutz et al. (2023): From prosumer to flexumer: Case study on the value of flexibility in decarbonizing the multi-energy system of a manufacturing company (open access)
• Fleschutz et al. (2022): Industrial grid fees vs. demand response: A case study of a multi-use battery in a German chemical plant
• Fleschutz et al. (2022): Demand Response Analysis Framework (DRAF): An Open-Source Multi-Objective Decision Support Tool for Decarbonizing Local Multi-Energy Systems (open access)
• Fleschutz et al. (2021): The effect of price-based demand response on carbon emissions in European electricity markets: The importance of adequate carbon prices (open access)
• Fleschutz et al. (2017): Electricity Cost Reduction Potential of Industrial Processes using Real Time Pricing in a Production Planning Problem (open access)

License and status

Copyright (c) 2017-2024 Markus Fleschutz

License: LGPL v3

The development of draf was initiated by Markus Fleschutz in 2017 and continued in a cooperative PhD between the MeSSO Research Group of the Munster Technological University, Ireland and the Energy System Analysis Research Group of the Karlsruhe University of Applied Sciences, Germany.
This project was supported by the MTU Rísam PhD Scholarship scheme and by the Federal Ministry for Economic Affairs and Climate Action (BMWK) on the basis of a decision by the German Bundestag.

Thank you Dr. Markus Bohlayer, Dr. Ing. Adrian Bürger, Andre Leippi, Dr. Ing. Marco Braun, and Dr. Michael D. Murphy for your valuable feedback.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.