Keep the renewable power flowing: develop data-driven algorithms to anticipate congestion in the grid and to rank mitigation options.
IMPORTANT. This PhD can also be completed in Dutch.
Congestion headroom and flexibility options for short- and medium-term grid planning.
While the energy system is undergoing tremendous changes (e.g. rollout of solar panels and electric vehicles), the grid operator is responsible for providing grid capacity for energy consumers and producers. When insufficient capacity is available, this has significant societal, financial and reputational impacts. Traditional network reinforcement is costly and slow, so shorter-term flexibility markets that are currently being developed are a tempting but unproven alternative. How can we quantify the risks involved and rank the options at our disposal?
As part of your PhD project, you will develop a novel decision support methodology that (1) integrates network models and measurements to learn spatiotemporal models of network usage and available flexibility options, including network switching, which influences forecasts; (2) generates probabilistic forecasts of remaining capacity that are calibrated on the tails (low remaining capacity) and quantifies uncertainties due to lack of input data; and (3) provides a risk-based ranking of mitigation options, including time-to-decide, and scales to online use for large networks with thousands of potentially congested assets.
We are looking for a mathematically-minded researcher who takes a principled approach to understanding data-driven methods. You will collaborate within a multidisciplinary team to develop algorithms that work in the real world.
The AI for Energy Grids Lab is a partnership between the Distribution System Operator Alliander, TU Delft, Radboud University, University of Twente, and HAN University of Applied Sciences. In the AI for Energy Grids Lab, we combine ground-breaking Artificial Intelligence (AI) methods with the reliable theory of the physical energy system. The area of data-driven scientific computing promises to combine statistics, time-frequency analysis, low-dimensional model reductions, and other techniques to extract information from data. With AI, we make such information useful for the management and planning of complex energy systems. For example, it is possible to use neural networks to model the state of the grid so operators can act in full awareness of the operating situation. This new AI for Energy Grids Lab investigates explainable AI and data-driven scientific modellings such as reinforcement learning, and graph neural networks for their applicability for state estimation, decentralised control of energy flows, risk-based investments and operation.
Your academic supervisors are Dr. Simon Tindemans (TU Delft, daily supervisor), Prof. Eric Cator (Radboud University) and Prof. Han La Poutré (TU Delft, CWI). Within the AI for Energy Grids Lab, you will be part of a team of five newly hired PhD researchers. Together with this team, you will spend two days a week at Alliander (in Arnhem), supported by around 10 scientific staff and data scientists and power system experts. You will develop and execute your own ambitious research program within the broader research vision of the team with the support of Dr. Jochen Cremer (TU Delft, AI for Energy Grids lab manager).
You will be hired as a PhD research at Delft University of Technology (TU Delft). As the Netherlands’ oldest technical university, TU Delft has a long history of research and education in electrical power engineering. Recognising the transformational power of Artificial Intelligence, TU Delft has launched a large cross-faculty AI initiative where this lab will expand the Delft AI Energy Lab. You will be a member of the Intelligent Electrical Power Grids research group at TU Delft headed by Prof. Peter Palensky, along with supervisors Simon Tindemans and Han La Poutré. The third supervisor Eric Cator is professor of applied stochastics at Radboud University. His group has a long standing collaboration with Alliander and a lot of experience in applying statistical models and AI to power grids.
Alliander is a distribution system operator that maintains and operates the medium-low voltage grid and connects end-users with generators to the transmission high-voltage grid. With the goal to enhance the transport capability of the distribution grid, this lab aims to address distinct challenges arising in the Netherlands and the thousands of distribution grids worldwide. This lab focuses on two ways of impact towards extending the grid capability where one improves the grid operations and the other investigates targeted infrastructure expansions. This lab combines the expertise of Alliander data scientists and electrical engineers, the latest digitalization programs that make grid data available, with the academic expertise on methods from artificial intelligence enabling the lab's scientific success and broader global impact. Alliander is one of the frontrunners worldwide as a DSO in applying smart algorithms and AI, it is also one of the places where the limits of existing methods and available commercial software surface.
- The PhDs of The AI for Energy Grids Lab will be employees at the knowledge institutes
- The submitted documents should include: CV, motivation letter, grades of all higher education degree programs (BSc, MSc, …), MSc thesis or (if not yet completed) another sample of academic/technical writing
- The application documents will be shared with the partners involved in the AI for Energy Grids Lab, namely individuals involved in the supervision of the PhD students from Radboud University, University of Twente, TU Delft and HAN
- Expiry date of the applications 31 May 2022
- An MSc degree (completed at the start date) in a discipline that combines mathematics and algorithmic thinking (e.g. Machine Learning, Systems & Control, (Applied) Mathematics, Operations Research, Electrical Engineering).
- An excellent academic record (typical grades of 8+ (Dutch) or A).
- The minimum requirement of a TOEFL* score of 100 or IELTS of 7.0 per sub-skill (writing, reading, listening, speaking) applies to all candidates wanting to pursue a PhD or PDEng programme at TU Delft. Candidates do not need to present the test results as part of their application. These results will be requested at a later stage during the selection procedure. “
- A good intuition for probability and statistics and an ability to read and critically analyse papers in the mathematics/computer science domain.
- You enjoy programming and strive to write code to a high standard.
- An affinity with the energy transition and its technical underpinnings (operation and planning of energy infrastructure).
- You like to work in a multidisciplinary team.
- Knowledge of electrical transmission or distribution systems.
- Experience with machine learning, preferably using Python.
- Experience with collaborative software development (e.g. open-source)
Alliander screens all applicants. Depending on the position, the screening consists of the following steps: checking references, checking the authenticity of identity papers and diplomas, an integrity check and requesting a certificate of conduct (VOG).
Dr. Simon Tindemans, email@example.com