Flexibility findings report


Less flexible electricity supply requires more flexible demand. In CREDS, we take ‘flexibility’ to refer to the capacity to use energy in different locations, at different times of day or year (via storage or by changing the timing of activity, including whether it takes place at all); to switch fuels; to smooth, move or create peaks in demand or, in the case of mobility, to re-arrange destinations and journeys in ways that reduce energy demand and/or congestion.

We contend that flexibility is constituted and limited by the interaction of social and technological/infrastructural arrangements including systems of storage and generation alongside social and institutional rhythms.

Research in this theme is informed by three key ideas.

  1. It should be grounded in an understanding of the timing of energy demand (domestic, non-domestic and in relation to the mobility of things and people) as an outcome of the rhythm, sequencing and synchronisation of activities.
  2. Interventions to mitigate peaks and increase flexibility in the timing of energy demand encompass a variety of technologies and pricing mechanisms which should not be studied in isolation.
  3. The wider societal impacts of harnessing flexibility, including the associated digitalisation and control automation in homes and work places, introduces new timings into people’s lives which create new challenges for Demand Side Management (DSM).

How flexibility is conceptualised is important

The way the energy industry and energy research conceptualise flexibility ‘fix’ a particular interpretation of normality, supposing that particular needs exist and need to be met. We make no such assumption.

We explore the different ways in which we can think about what it means to be flexible, or to have flexibility, which opens up the space of possibilities for how flexibility could be unlocked and harnessed more effectively. Our work challenges the representations of time and society in the energy sector, and proposes innovative ways of Conceptualising flexibility. This work laid the foundation for further studying the Institutional rhythmsOpens in a new tab that shape energy demand and its flexibility. We started by looking at ‘organisational flexibility’ in large UK high schools, asking whether such organisations can change the timing of their demand, how they might do this and what lessons (not necessarily energy-related) could be learned. Next, we examined the idea that adaptation in an organisation depends on a mass of intersecting temporal rhythms. For example, people responsible for caring for children are all too familiar with the punctuating nature that the timings of the school run play in the planning and unfolding of everyday life. And from a wider point of view, it is easy to see how the timing of school life is embedded in city and regional rhythms as it is connected to the timing of work, home, and everyday life.

Our research argues that flexibility is constituted beyond any one organisation, as we discuss in our paper on What is energy for? Social practice and energy demandOpens in a new tab, and in later work that investigates the Opportunities for and limits to temporal coordination, pdfOpens in a new tab. This research demonstrates that flexibility is a feature of how multiple practices hang together and of the changing relations between them. Flexibility exists in the intersections between organisational and institutional processes and so these define the scope for adapting and modifying energy use in either time or space.

Flexibility is not a win-win

Advocates of flexible technologies and tariffs argue that flexibility will benefit the whole system, including users. This is not necessarily always the case.

As our research shows, people’s activities underpin peaks and drops in demand we observe. Thus, having the ability to make demand (more) flexible ultimately depends not only on consumers’ willingness to change how and when they do things, but also on their ability to do so. Not everyone is able to; not everyone should have to.

Identifying those who might benefit and, more importantly, those who might be adversely affected by measures looking to harness flexibility is a complex task. To overcome some of the associated challenges, we have developed novel data analysis algorithms that allow for clustering of households with distinctive consumption patterns, as well as studying the long-term dependencies in their consumption behaviour.

Our work on demand-based consumer classification has enabled the analysis of the Distributional effects of Time-of-Use (ToU) tariffs on different types of consumers, on different income groups. That is, how much consumers will benefit from, or have to pay for, new tariffs looking to incentivise flexible demand. Our work goes a step further as we try to identify the specific activities likely to be responsible for gains or losses associated with the introduction of ToU electricity tariffs for the residential sector.

This research offers clear evidence to challenge the proposition that flexibility is a win-win. The findings, among other things, reveal that time-poor people such as single parent workers, are more likely to be worse off on an ill-designed ToU tariff compared to a flat tariff. This research was cited by Ofgem’s final Impact Assessment, pdfOpens in a new tab on ‘Electricity Retail Market-wide Half-hourly Settlement’ and used as input for Ofgem’s Electricity Network Access and Forward-Looking Charging Review. It also fed into a Parliamentary Office of Science and Technology (POST) note, which was published on their official portalOpens in a new tab in May 2023. Such regulatory reforms of tariffs have been estimated to bring about bill savings for UK residential customers of between £2bn and £5bn up to year 2045. The research shows that policymakers, industry and consumer associations should not have naïve views of flexibility as a panacea for an electrified future.

Learning from (flexibility) history

Infrastructural legacies and previous methods of balancing supply and demand are layered over time and in ways that matter for contemporary connections between social practices, the timing of energy demand, and for flexibility.

This interdisciplinary work on the Histories of Flexibility brought historians and social scientists together to look at selected cases using different approaches and scales of analysis to demonstrate this argument. This resulted in a special issue of the Journal of Energy History, in which we discussed the Legacies and lessons from the past when it comes to energy systems’ flexibility, the history of balancing demand and supply in UK’s gas networksOpens in a new tab, among many other interesting accounts that take a longer view of the concepts associated with energy demand flexibility.

In addition to these, more traditional academic publications, and with a view to sharing some of the most relevant findings and insights derived from this work as far and wide as possible, we produced a photo essayOpens in a new tab that takes us in a journey through time to explore the making of energy supply and demand.

Five main findings emerged from this work:

  • Flexibility is positioned at the intersection of supply and demand.
  • Flexibility is an outcome of political and institutional arrangements.
  • Flexibility is a feature of service provision: this changes over time.
  • Flexibility is a feature of how energy is distributed in space and time.
  • Flexibility is as an outcome of the changing relation between supply and demand, and of intersecting formations of legal, financial, social, governmental, and technical modalities.

Historical interventions in supply and demand – such as the transition from using gas to electricity for artificial lighting – have long(er) term consequences for flexibility. We should heed these historical lessons now in our present efforts to promote net-zero and to decarbonise: these too will have implications for resource consumption and of demand in future.

New flexibility metrics are needed – going beyond price elasticities

Pricing and tariffs are considered key tools to reduce energy demand and achieve net-zero. Price elasticity is meant to be a direct measurement of how increases in prices of the supplied good or service correspond to decreases in demand, and vice versa. Our research, however, argues that this is only part of the picture as it does not take into account how people’s flexibility varies based on the time of day, location, work arrangements and social commitments. This is especially true for energy because people are actually consuming the services it enables – such as cooking, lighting and heating – rather than energy directly, and generally pay little attention to the amount of energy they consume.

Our aim was to develop alternative elasticity metrics to enrich our understanding of changes in demand in relation to price by asking some novel questions around changes in price elasticity across the day or between days, and thinking about the wider impacts of dynamic pricing and metrics for non-price elasticity. Preliminary findings show that new metrics based on time of day disaggregation provide accurate measurements of price elasticity of energy demand. We also found alternatives to the market and non-market values of time and compared changes in demand associated with ToU tariffs with changes before / after Covid-19 lockdowns and before / after price surges. Our clustering analysis reveals the variability of the ‘response strength’ of different types of consumers.

Our work provided evidence for a new Demand Flexibility ServiceOpens in a new tab which was launched in November 2022 by National Grid ESO; the results from the initial Demand Flexibility Service trials will be analysed in future work. Findings on new elasticity metrics may be used by regulators to evaluate the effectiveness of pricing measures and will inform energy economists on new ways to measure price elasticity of demand. This work was also informed by a joint meeting with BEIS/DESNZ which led to further research on the potential trading of flexibility certificates and a podcast on the Price elasticity of energy demand featured in the CREDS in Conversation series.

Technology-based flexibility harnessing is only part of the puzzle

Both in industry and academia, it is commonly taken for granted that technology-based strategies for harnessing flexibility are the main – if not the only – solution to the problem. This means that estimations of the flexibility provision potential are typically based on idealised, theoretical systems where flexibility is achieved only through the use of storage assets, energy efficiency measures, and automated control systems. What this widespread assumption fails to realise is that there is much more to the accurate estimation of flexibility potential, as it is people’s everyday life activities what drives the demand for energy. This is problematic as it could, on the one hand, lead to over-engineered technology-based interventions that essentially limit the potential for demand flexibility provision to what can be achieved through such means; on the other hand, it could contribute to the preservation of narrow views of what flexibility provision entails, which  fail to take into account the social dimensions of flexibility – for example, that the demand for energy associated with artificial lighting depends on activity and location, not simply time-of-day and levels of daylight.

Our research started with a review of the literature which explored the following:

  • The translation of meaning of demand in relation to concepts of non-negotiable energy end use effects associated with technology efficacy/efficiency
  • Issues of new fixities resulting from intended technological demand flexibility
  • Addressing issues of determinism in accounts of new technology impacts.

Forthcoming outputs include the development and use of an agent-based model, adapted from Capel-Timms et al. (2020)Opens in a new tab. The model performs accurate simulations of spatial and temporal variability of anthropogenic heat emissions (as a proxy to energy use) across Greater London. It looked at the activity and movement of people (in a city context) and how capacity constraints, service scheduling and energy-demanding technologies inform the timing and location of energy demand. The model combines time-use survey, census, work-day population, travel routing, and urban meteorological data. Further model development has led to a more flexible approach that allows for easier modelling of different city contexts and is now set to run for Greater London and Berlin and is being set up for modelling of Paris and Bristol.

The key findings for this work were:

  • Introducing new technologies results in new interactions with it which, in turn, changes the pattern of energy demand.
  • Findings from the model show that it is able to simulate the variability of energy consumption in time and space across London better than some other models.
  • The model also explored the connection between transport and building-centric energy use, highlighting that choice costs energy (e.g. being able to take different types of transport at different times of day using different routes, or responding to different weather conditions) and creates pollution. This will be the subject of further work.
  • When generating power load profiles for occupant-driven appliances, an approach based on sequence of activities occurring at the same location enables capturing more variability and proves to be more scalable compared to a probabilistic occupant-based approach.

These findings are of relevance to electrical distribution systems and urban planners, policy makers, national and regional travel operators and planners, DSO network operation planners, and academics.

Banner photo credit: Pat Whelen on Unsplash