Build the model
Put the full chain on DAFNI: storm wind fields, the chance of each line failing, and what happens to the network afterwards, including how protection reacts.
DAFNI Fellow · The University of Manchester
Climate-Driven Cascading Failure Risk in Electricity Transmission Networks
Bringing weather-driven vulnerability assessment and sequential cascading failure simulation onto DAFNI, so that resilience evidence for the GB transmission system can be reproduced, examined and reused.
Explore the workshopFellowship updates
- Fellow
- Dr Yitian Dai
- Duration
- 2026 – 2027
- Demonstrator
- GB transmission network under windstorm scenarios
What the Fellowship is about
Storms are now the main cause of large power cuts in Great Britain. When wind brings down one line, the power it carried has to flow elsewhere, and other lines can overload and trip in turn. As more generation moves to wind farms at the edges of the network, the way these failures spread is changing. The Fellowship builds a model of this on the real GB transmission network, runs it with real and future storms, and asks the people who plan and operate the network whether the results are useful.
Map the risk
Use past storms and climate projections to work out which lines and corridors are most likely to fail and where a cascade is most likely to start, now and in future decades.
Link to other infrastructure
Pass the power outages to other models on DAFNI, so their effects on distribution networks, transport and communications can be studied.
Test it with practitioners
Put the results in front of network operators and regulators at the workshop, find out what they would need to act on them, and write that up as recommendations.
Updates and milestones
Last updated: 9 October 2026
- Fellowship begins
Fellows Plan agreed with DAFNI; data and model inventory started.
- GB network model onboarded
Open GB transmission dataset uploaded to DAFNI with documented metadata.
- Windstorm scenario workflow
First containerised run of the weather-to-outage pipeline on the platform.
- Workshop: Future Energy Network Resilience Under Climate Stress
Half-day workshop in Bristol, the day before the Supergen Energy Networks Hub Risk and Resilience Day on 23 March.
- White paper and guide published
Findings, DAFNI experience and technical recommendations released.
The model on DAFNI
Everything needed to rerun the analysis is kept on DAFNI, so that others can check it, change the inputs and use it for their own questions.
- The network: an open model of the GB transmission system (PyPSA-GB, based on ETYS 2023).
- The storms: hourly wind gusts from the ERA5 reanalysis, checked against Met Office weather stations, for named storms such as Arwen and Eunice and for every winter since 1979; UKCP18 climate projections for how storms may change.
- The model: for each hour of a storm, how likely each line is to fail, and then which other lines overload and trip as the power reroutes.
- The results: lists of failed lines and lost supply that other DAFNI models can take as input.
- The write-up: the code packaged to run on the platform, with worked examples and a short guide.
- 1LoadThe network and a storm
- 2RunLine failures, then the cascade
- 3SeeWhich lines fail and how much supply is lost
- 4Pass onThe outages, to other infrastructure models
- 5ReuseChange the storm or the network and run again
Future Energy Network Resilience Under Climate Stress
Half-day stakeholder workshop · 22 March 2027 · Bristol
The workshop will bring together transmission and distribution operators, the regulator, government analysts and researchers to test how weather-driven cascading failure evidence can support planning and operational decisions. It is co-organised with the Supergen Energy Networks Hub and held the day before the Hub's Risk and Resilience Day on 23 March 2027, also in Bristol, so that participants can attend both: the workshop for those who want to get into the modelling detail, and the Risk Day for the wider community.
Register via Risk and Resilience DayView the proposed agenda
Workshop aims
- Test whether practitioners find the demonstrator's evidence credible and usable: its assumptions, its representation of uncertainty and how results are presented.
- Establish what operators and regulators would need, in data, format and governance, before such evidence could inform an investment or operational decision, and what stands in the way.
The workshop's output is the participants' reaction to the evidence, which will shape the closing white paper.
Proposed programme (afternoon of Monday 22 March 2027)
- Welcome and framingWhy weather-driven cascading risk matters for a decarbonising GB system; the two questions the day is designed to answer (is the evidence credible and usable, and what would a decision need); how the afternoon runs and how participants' views will be captured for the white paper.
- The demonstrator in three casesResults only, no model walkthrough. Each case is presented in about ten minutes with the same GB map and the same set of figures.
- Case 1, a storm replayed. A recent named windstorm on the GB transmission model (PyPSA-GB, ETYS 2023): the gust footprint hour by hour, the circuits exposed, the outages and any overload chain the model produces, and a side-by-side comparison with what was recorded on the day. The question for the room: does the model reproduce what happened, and what does it get wrong?
- Case 2, the same storm moved and intensified. The footprint shifted along plausible alternative tracks and scaled to present-day and future return levels, giving a map of where outage and cascading risk concentrates across the network and which corridors carry it. The question: is this where you would expect the risk, and is the uncertainty shown in a way you could use?
- Case 3, one storm under two interventions. The base case rerun with (a) an operational measure, pre-storm reduction of flows on the exposed boundary, and (b) an investment measure, reinforcement or undergrounding of the specific circuits identified in Case 2, with the change in outage and customer impact for each. The question: would either result change a decision, and what is missing before it could?
- Live run on DAFNI (5 min). One parameter changed (storm offset or intensity), the workflow rerun on the platform, and the output map returned, to show the work is reproducible and open to others.
- Invited perspectivesFour short talks of about twelve minutes, each framed around a problem rather than an organisation.
- Transmission planning and operation. How weather-related risk to the transmission system is assessed today, in investment planning and in the control room ahead of a storm, and where current evidence falls short.
- Storm data and future climate. What storm footprint, reanalysis and climate projection data exist for GB, how reliable they are for engineering use, and what a network model should and should not ask of them.
- Regulation and policy. What evidence a resilience investment case or an operational standard needs to satisfy, how uncertainty is treated, and what would make model-based evidence acceptable.
- DAFNI. The platform, how workflows and data are shared and reproduced, and what other infrastructure models it connects to.
- Break
- Structured critique at tablesMixed tables of operators, regulators, data providers and researchers, each with a facilitator and a printed pack containing the three cases. Forty-five minutes of discussion guided by a one-page worksheet, then a short written summary per table.
- Credibility: which assumptions, inputs or results do you not believe, and why?
- Use: which decision, planning or operational, could each case inform, and in what form would the result need to be presented?
- Gaps: what would have to be added, changed or validated before you would act on it?
- Data: what data or experience does your organisation hold that would change these results, and could it be shared?
- Report back and synthesisEach table reports in three minutes; the facilitators consolidate the points into an agreed list of requirements, barriers and priorities that will structure the white paper, and identify follow-on case studies with named organisations.
- Close and next stepsWhat happens to the material gathered, the timetable for the white paper and technical guide, and how to access the DAFNI workflow.
Fellowship outputs
The primary output will be a white paper bringing together the demonstrator, stakeholder evidence, experience of DAFNI and prioritised technical recommendations.
DAFNI Fellowship poster
Climate-driven cascading failure risk in electricity transmission networks.
View poster (PDF)Practical guide
Running weather-driven cascading failure workflows on DAFNI.
Coming in 2027Fellowship white paper
Research findings, workshop outcomes and recommendations for DAFNI.
Coming in 2027Discuss the research or join the workshop
Expressions of interest are welcome from network operators, regulators, public bodies, researchers and DAFNI users.
Department of Electrical and Electronic Engineering
The University of Manchester
yitian.dai@manchester.ac.uk