Starting in April 2023, the Centre of Excellence for Resilient Infrastructure Analysis fosters research in the area of resilience in the natural and built environment. The Centre’s first research strand comes as part of the overarching UKRI programme ‘Building a Secure and Resilient World’ (BSRW)’, a 5-year programme which sought to tap the UK’s research and innovation system to tackle large-scale, complex challenges for the UK. The many strands of research needed to create more resilient cities and societies were brought together with several BSRW funding projects, which are here explored further to their completion.
Resilience for Integrated Water Systems (RIWS)
The RIWS project focused on performance-based resilience assessment for integrated water systems. Modelling was completed using the Water Systems Integration Modelling framework (WSIMOD), concentrating on a small urban catchment area in Luton, north of London. Two aspects of the work were developed – framework for IWS resilience assessment and use of the approach to develop a decision making under deep uncertainty (DMDU) approach to flood resilience.
In the first part of the work, eight challenges around performance-based resilience for integrated water systems were identified via literature review, primarily focusing on deriving consistent results across subsystems and on interpreting resilience results for actionable management. The key was trying to find the metrics so that resilience can be calculated in a unified way and then to use this metric to compare the resilience across the components of the water system.
Stakeholders consulted included Mott MacDonald, Affinity Water, Thames Water, Rivers Trust, Greater London Authority and the Environment Agency.
Outputs uploaded onto DAFNI: WSIMOD for the selected case study.
In the second part of the work, the team decided to examine the past 50 years and see how far uncertain conditions explored in DMDU were from an idealised (known) solution. The results worked well and could prove to be a way of understanding the effectiveness of their new methods for resilience planning under deep uncertainties via a collaboration with fellow DAFNI BSRW project lead on the USARIS project (Uncertainty quantification and Sensitivity Analysis for Resilient Infrastructure Systems), Associate Professor Dr Francesca Pianosi. This also offers insights into identifying interventions and highlights the importance of accurately characterising climate uncertainties.
Dissemination efforts include the publication of one paper, with a further two under review, one which covers the definitions of the resilience of integrated water systems, the second applying this concept of resilience for deep uncertainty.
“There were so many interesting functionalities to use,” PI, Professor Ana Mijic of Imperial College London, enthused. “The current WSIMOD model is publicly available and free to use via the DAFNI platform. It enables us to automatically override any existing parameters which really helps with the customisation of the model to explore a range of water systems configuration. We can change the behaviour of the existing components, and add new ones, which is really important to be able to do any kind of intervention.”
STORMS: Strategies and Tools for Resilience of Buried Infrastructure to Meteorological Shocks
The STORMS project focused on developing risk assessment models for buried infrastructure, such as water systems and telecommunications. A climate risk model assessed the risk from both the surface water and ground movements by feeding in the data projections of rainfall and soil moisture.
For the surface water, a hydrodynamic model was used to calculate both the soil erosion and water accumulation, which helps to calculate the erosion or deposition due to flooding. They examined the National Gas transmission network running through the River Eden catchment area which was damaged by Storm Desmond in 2015. Based on these results, the team could then calculate the risks for erosion and water accumulation in terms of potential damage for the pipework.
The team also examined rainfall scenarios by looking at extreme weather events in the UK and risk assessment from ground movement modelling, compared the results from five different regions within a gas network, and analysed the correlation between the risk level and the soil moisture level.
Stakeholders consulted include Line Search Before You Dig (LSBUD), Cadent Gas, National Grid, Humber Industrial Cluster, DSIT, NUAR and UKRI’s JASMIN.
Outputs uploaded onto DAFNI: Workflows for assessing the surface water and ground movement risks, together with the supporting datasets and models.
Dissemination efforts include three papers under preparation, extensive industry engagement via workshops and, as an indirect benefit, career progression of several team members. The Chartered Insurance Institute attended one workshop and will submit this research as evidence in their latest report.
Future plans include investigating new use cases through further data sharing and collaborations, taking into account new modelling approaches, as well as a comprehensive assessment of adaptation and Net Zero transition.
“Although there is still a lot of validation and improvement to be done, DAFNI has really helped us to open new opportunities for looking at other issues”, explained PI Dr Xilin Xia, Assistant Professor in Resilience Engineering, University of Birmingham. “DAFNI enabled sharing and collaboration, so we could improve the methodologies and complete more comprehensive assessments. We will continue to engage with the wider utility sector and government as we have received so much interest!”
NIRD: National Infrastructure Resilience Demonstrator
The NIRD project focused on delivering an open-source modelling framework for stress-testing network resilience against extreme weather, such as flood and storm events. A very detailed road transport flow was created, which includes approximately 12 million journey-to-work travel patterns on UK roads. The model also included stress-testing quantifying road direct physical damages and indirect flow disruptions losses in monetary values for 18 historic flood events.
The NIRD showed that the estimated costs of indirect impacts proved to be far greater than the direct impacts for several flood historical events. While overall the model results were close to reported Government reviews of past historical floods, such as the summer of 2007, observed data for such events was found be generally lacking. Moreover the analysis showed that the ratios of direct and indirect impacts from flooding differed across flood events, which challenges the existing approach in Government reviews of allocating direct and indirect impacts equally to flood events.
The team presented and discussed their work with different stakeholders including the Ministry of Housing, Communities and Local Government (MHCLG).
Outputs uploaded onto DAFNI:
- Ordnance Survey Open Roads derived road datasets with bridges tagged
- Origin-Destination Matrix and flow and failure datasets
- NIRD model code
Dissemination efforts include one paper currently under review, a recent conference presentation at the European Geosciences Union General Assembly and the creation of several Jupyter notebooks, which have been uploaded onto both DAFNI and GitHub. As an indirect benefit, PI Dr Raghav Pant, Senior Research Associate at the Environmental Change Institute, University of Oxford now sits on the Climate Resilience Expert Panel for Ofgem, offering advice on how to implement resilience and consider stress-testing into Ofgem’s processes.
Since working with DAFNI, the team has put together an Impact Acceleration project with the UK Government’s Ministry of Housing, Communities and Local Government (MHCLG), which has received funding from the University of Oxford. For this project, the team are essentially helping the MHCLG understand road conditions and planning under the context of future housing development. They are also working with OpenCLIM on climate scenarios, producing flow patterns and failure outcomes for vulnerability and adaptation assessment for the Maximising Adaptation to Climate Change (MACC) Hub project. The team is expanding the UK model created with DAFNI to a European scale, thanks to a collaboration with EU Horizon MIRACA (Multi-hazard Risk Assessment for Climate Adaptation) on passenger and freight flows.
“This project has created a national-scale road network stress-testing model that requires large-scale computational resources, for which the DAFNI platform is ideal,” says Raghav. “The value of DAFNI has been immense. firstly in terms of providing funding for novel research, secondly by supporting the storage of data and implementation of the model at large scale, and thirdly in helping disseminate the project outcomes to a wider audience. Through such support we have been able to get follow-on funding to develop this work along different and interesting avenues.”
Pywr-WREW – A Water Resources Model for England and Wales to enable strategic analysis of the drought resilience of water supply infrastructure
Traditional water resource management, focused on individual companies, is no longer sufficient to address the complex issues surrounding England’s interconnected water supplies, increasing water demand, climate change and heightened ecological needs.
The Universities of Oxford and Newcastle, in collaboration with the Environment Agency and Ofwat, initiated the National System Simulate Modelling project. As part of this, they developed a comprehensive water resource model for England and Wales, built using the unique Python water system simulation software, Pywr, developed by the University of Manchester. Stakeholders consulted include the Environment Agency, University of Bristol and University of Sheffield.
The Pywr-WREW model integrates various water usage sectors and future scenarios to assess potential water shortages and solutions. The model, datasets and processing instructions were compiled and added as a workflow on DAFNI, that aimed to enhance the model’s accessibility and usability for researchers and practitioners. This increased the flexibility of the model and ease to update, rather than having to scroll through hundreds, if not thousands, of lines of code.
Outputs uploaded onto DAFNI: Model, datasets and processing instructions.
Dissemination efforts include a detailed user guide and a collaboration with Dr Francesca Pianosi of the University of Bristol and Saskia Salwey (who has since moved from Bristol to University of Utrecht), applying the SAFE toolbox to explore different parameters in the model and inputs in terms of how climate change driven river flows impacted results, as part of the USARIS (Uncertainty quantification and sensitivity analysis for resilient infrastructure systems) project.
“We developed a local version of the model that can be used on your own computer,” explained Dr Anna Murgatroyd, Principal Investigator and Lecturer in Hydrology at Newcastle University, formerly at the University of Oxford. “It quickly became really valuable to run the model on DAFNI as we wanted to include longer model runs, more climate data and more water demands. To facilitate learning, we created a user group on DAFNI which contains the workflows and data to run the model, which we could easily share with our collaborators at the Environment Agency.”
Additional uses include training in new water systems planning software, the employment of a post-graduate research assistant, connections with Pywr users worldwide, career development and an ongoing collaboration with the Environment Agency.
Looking ahead, the team is hoping to collaborate with West Country Water and the Environment Agency’s regional water resources team, and with the University of Sheffield on a potential PhD research project, to investigate water needs for green hydrogen. Dr Murgatroyd is also interested in looking at how efforts to achieve Net Zero will impact water supplies and how shortages may impact these efforts.
FIRM: Flood Resilience Simulation Model
The FIRM (Flood Infrastructure Resilience Model) simulates flooding and the human response to flood events. Representing the interactions between people and flooding, and how this impacts safety, are the key distinguishing features of FIRM.
The purpose of the project was to recode a coupled hydrodynamic and agent-based model that was developed a few years ago, using a modelling platform called Netlogo, to make it faster, enable it to take advantage of DAFNI’s capabilities and to make it publicly available.
This model can simulate human responses during a flood event. It can be used to make decisions about whether to evacuate or take some other form of action in response to a flood warning, and then to obtain a sense of how different resilience strategies lead to different outcomes in terms of impacts on people, properties at risk and infrastructure disruption.
This work has benefited from engagement with end-users from organisations such as the Environment Agency, Airbus and Local Resilience Fora. A new paper has been published in the Journal of Urban Science, demonstrating the flexibility of the model, and the work will be further developed into a PhD project.
Outputs uploaded onto DAFNI: Netlogo code now in Java, Model workflow, Standardised input data (*.json), Example dataset (Towyn, North Wales), Instructions.
“Working with DAFNI and being part of the BRSW programme has added scale and visibility in a way that otherwise wouldn’t have happened,” says PI Richard Dawson, Professor of Earth Systems Engineering at Newcastle University. “The software development has laid the foundations for many other potential future developments to this work, from better visualisation tools to the scope that was available for exploring use of different types of hazards.”
Making FIRM available on DAFNI enables it to be integrated into DAFNI’s ecosystem of other models. This has opened up exciting new avenues for future development, such as coupling with more detailed urban hydrodynamic models, other hazard models, or by applying sensitivity analysis tools.
SCQUAIR: Small Changes, QUANT and AI Resilience – Simulating the Resilience of Transport Infrastructures using QUANT
The SCQUAIR project simulates the pattern of land use and transportation for England, Scotland and Wales, running myriad simulations to identify the impact of new jobs, in terms of where people live and how they travel there.
The underlying spatial interaction model is called QUANT, developed by Professor Michael Batty of UCL, and runs very rapidly in a web-based environment. It is configured in terms of thousands of small zones and three modes of transport (bus, rail, road) which bind together employment at place of work and population at place of residence.
The SCQUAIR project involved teaching artificial intelligence (AI) to add scenarios to QUANT, which project the magnitude and direction of people commuting, the differences in kilometres travelled when new transport options are introduced, the impact of job swaps between regions on road kilometres driven, and the transport they use to commute.
Outputs uploaded onto DAFNI: QUANT input data e.g. flows and network data, a generalised QUANT model, different network behaviours, maps of scenario impacts.
Dissemination efforts include a workshop with the Bartlett School of Planning, presenting to delegations from Korea, China and Brazil, a paper and book chapter under review, and a Masters student continuing the research to look into accessibility.
“Thanks to DAFNI, we managed to run 1 million road, bus and rail scenarios for the UK,” enthused PI Dr Richard Milton, Senior Research Fellow at University College London. “The results revealed something that we hadn’t predicted, which is the different nature of the networks, in terms of resilience and predictability. We started coding early on, which proved to be a tremendous advantage, as we were able to hand over a working model to the researchers. To begin with, the project was split over lots of different platforms, so it was great to have DAFNI bring it all together.”
There is now an incredibly rich source of data to begin evaluating, from city impact scenarios to nodes covering the flows of 20 million people going to work, whether by cycle, bus, road or rail. The model is computationally very intensive with a very fast run of five seconds. Future plans include detailed analysis of this data and the potential for creating a portfolio of examples, from flooding to transport network infrastructure where the model could offer critical support to urban planning and resilience in the UK.
USARIS – Uncertainty quantification and sensitivity analysis for resilient infrastructure systems
Computational modelling provides a vital tool to support infrastructure investment decisions. However, model outputs are conditional on a range of uncertain assumptions and input data. Overconfidence in model results and insufficient consideration of the breadth of possible futures are key obstacles to resilient infrastructure design.
The USARIS project, led by PI Dr Francesca Pianosi, Associate Professor in Water & Environmental Engineering at the University of Bristol, integrates a generic methodology into DAFNI , which analyses the propagation of uncertainties and enables better model construction, validation, and use for decision-making under uncertainty.
Ultimately the project aimed to promote best practices for responsible modelling and robust decision-making in the DAFNI user community.
Stakeholders consulted include the Water Industry Modelling Advisory Group (MAG) and the Environment Agency
Outputs uploaded onto DAFNI: Concept pilot applications in hydrology, water resources and energy sector. Rainfall run-off model (Hymod), Wind power model (with Hannah Bloomfield – BRINES) Pywr water systems resources model (with Anna Murgatroyd – Pywr-WREW)
Dissemination efforts include leading a DAFNI webinar, three papers in review, a poster presentation at the EPSRC Supergen Energy Networks Hub, invited and keynote talks, two workshops, a planned seminar with the Department of Energy Security and NetZero (DESNZ), a training course for the NERC Centre for Doctoral Training, and career advancement for Post Doctoral Research Associate Saskia Salwey who is continuing her career in water resources at the University of Utrecht.
The team wanted to acquire an understanding of the uncertainty and sensitivity of models, how to prioritise efforts to improve model accuracy, and to look at to what extent design options are effectively robust to a range of uncertainties. Uncertainty quantification (UA) and sensitivity analysis (SA) were the methodologies to help address these challenges, inspired by the XLRM (eXogenous uncertainties, policy Levers, Relationships, and Measures) framework for Decision-Making under Deep Uncertainty.
Three broad categories of questions that these methodologies would help us with are 1) to prioritise efforts for model improvement; 2) how the models are validated, and 3) the system behaviours. Once established, the team could examine the key drivers of the system performance in the future.
The team implemented DAFNI visualisations in the form of Jupyter notebooks, where essentially the calculation of output ranges and sensitivity indices is completed by relying on a free Python package called the SAFE Toolbox.
The team’s next plans are to examine the propagation of uncertainties when linking with different models and work on a more structured approach to analysis. Then, the value of methodologies can be realised. Model behaviour can be difficult to predict and lowering the barrier as to the uptake of these new methodologies could prove crucial in resilience success. There is also evidence of the need to advocate for more open access data, so the models could be used to feed into systems at a national scale.
“We are very grateful for all the work the DAFNI team did on essentially implementing new parameter suite functionality to run a model up to 1,000 times with different variations of input parameters,” said Dr Pianosi.