× This Challenge was posted 3 weeks ago

03

Challenge view

Back to Project

Predicting Control Energy Volumes for Grid Stability

⛶  Fullscreen ↓  Download

What

Develop a predictive model to forecast control energy activations for grid stability. Specifically, participants are tasked with estimating the daily activation volumes (in MWh) of automatic Frequency Restoration Reserve (aFRR) activations in Switzerland. This includes distinguishing between positive and negative activations. Additionally, participants should consider how future scenarios - such as the expansion of renewable energy, climate change, and the reduction of fossil fuels - might impact these activation volumes.

Why

Ensuring a stable electricity grid requires maintaining a perfect balance between electricity generation and consumption. As the energy landscape evolves rapidly with increasing renewable energy sources, predicting and ensuring grid stability becomes more challenging. Renewables like wind and solar power are variable and less predictable compared to traditional power plants, and the electrification of sectors like transport and heating further complicates power flow predictability. To manage grid stability, ancillary services such as aFRR are crucial, as they provide quick adjustments to electricity supply or demand. By predicting aFRR activations, we can better manage and schedule assets, identify potential market opportunities, and ensure grid reliability in the face of an evolving energy landscape. Predicting aFRR activation calls is particularly challenging, especially when simulating scenarios for the future up to 2050. This challenge aims to determine if the fundamental input and output data of a typical electricity price forecasting model can also be used to predict daily aFRR activation volumes.

How

Participants will

  • Data Access: Receive curated datasets containing historical records of aFRR activation calls (both positive and negative) in Switzerland for the years 2018 to 2023 and files containing potential input features. The datasets will be pre-processed and include key contextual factors relevant to the challenge.
  • Model Development: Develop predictive models using statistical and machine learning algorithms to predict daily aFRR activation volumes. Emphasize model simplicity and interpretability. Ideally, the model should provide not only a point estimate for daily aFRR call-off quantities but also a probability distribution for these quantities. Models will be evaluated on a test dataset from the year 2024.
  • Scenario Evaluation: Consider how future trends, such as increased renewable energy adoption, climate change, and reduced fossil fuel usage, might influence aFRR activation volumes. Visualize assumptions and findings.

Benefits

  • By participating in this challenge, you will gain hands-on experience in predictive modelling, data analysis, and the practical application of their work to real-world energy grid stability issues within a short timeframe.
  • You support grid operators and energy companies in forecasting future grid stability. In this way, you will help to better quantify the costs of expanding renewables and, at the same time, identify potential market opportunities for flexible power plants or storage systems (e.g. utility-scale batteries).
  • Establish contacts with BKW's flexibility and market analysis teams.

Contact

Jill Huber, Fabian Gottschlich, Tom Felder (BKW AG)

All attendees, sponsors, partners, volunteers and staff at our hackathon are required to agree with the Hack Code of Conduct. Organisers will enforce this code throughout the event. We expect cooperation from all participants to ensure a safe environment for everybody.

Creative Commons LicenceThe contents of this website, unless otherwise stated, are licensed under a Creative Commons Attribution 4.0 International License.