Phil Grunewald, Malcolm McCulloch, Nick Eyre

Aims and Objectives of ReNEW

ReNEW was devised in response to the COVID-19 pandemic. Repeated lockdown periods had a profound impact on UK households, restricting their freedom of movement and forcing many to adopt new practices of working from home.

Oxford had developed detailed activity diaries, which clearly showed the impact of these restrictions. What was less clear was

1. How new activity patterns affect energy use and
2. Whether these changes are sustained after the end of restrictions

To answer these questions, ReNEW developed new methods to access UK smart meter data via the Smart Energy Research Lab, which makes smart meter data from 13,000 UK households available to accredited researchers. In addition, an entirely new data access platform has been created to access higher resolution smart meter and contextual data from our own panel of currently 200 participants.

Participant recruitment and some software licences were supported with funding from EPSRC (EP/M024652/1). The software development and data analysis is supported by the Oxford Martin School in its entirety. Towards the end of the programme, two major funding sources have been realised to scale ReNEW into a national programme.

Outcomes

Both research questions could be answered with the new data sources. The hypothesis stated in the original proposal that the pandemic could result in a paradigm shift in energy use could not be supported. The changes in energy use patterns - even during lockdown - were modest and post-pandemic energy patterns do not show profound changes above and beyond those caused by energy price rises and the cost of living crisis.

ReNEW therefore expanded its research focus, in consultation with the Oxford Martin School, and explored more acute interventions and their impact on household energy demand. These measures have been highly successful, resulting in energy demand reductions during critical high demand periods, of between 23% and 26%. These findings are consistent with the National Grid ESO Demand Flexibility Service trials during the winter 2022/23 and can contribute valuable insights into the mechanisms by which households are able or unable to provide such system services. This report will focus on the findings of these ReNEW trials.

ReNEW Research

Participant recruitment for smart meter data access

Preliminary research leading to the creation of ReNEW relied on a self-selecting convenience sample of households. With funding from EPSRC, it was possible to collaborate with a professional recruitment partner (Walnut Ltd.) to ensure that a more nationally representative sample could be created. From a panel of 100,000 UK citizens, 200 stratified households could be signed up for this research. Each of them completed a detailed socio-demographic survey, to help with the stratification, and consented to sharing their smart meter data via our technology partner Hildebrand Ltd..

Secure servers, data base infrastructure and APIs to access, store and process the smart meter data have been developed with the support of the Oxford Martin School.

Of the 200 participants, 140 continue to feed live smart meter gas and electricity consumption and tariff data with 30 minute and 1 minute resolution. The collaboration with Walnut and Hildebrand has been extended to sustain and grow this sample.

Smart meter data analysis

From the two datasets (SERL and ReNEW's own) the patterns of energy use during lockdown and between lockdown periods could be compared. Somewhat unexpectedly, the differences were not as big as expected. While some households showed clear evidence of front-loading the day, as part of working from home, in aggregate the effect size is small. This finding was especially noticeable for heating. The research hypothesis that people working from home would require more energy to heat their homes was not borne out by the data. Thermostat set points and timing schedules for home heating systems appear to have been left largely unaltered during lockdown.

Further research into cognitive and physical barriers towards engagement with home heating systems will form part of follow up research.

The pre/post COVID19 comparison was complicated by unprecedented energy price rises and the cost of living crisis. ReNEW data revealed a reduction in heating energy demand in the 2022/23 heating season compared to the previous year, but during particular cold spells, such as in the middle of December 2022, demand still exceeded that of the previous year. Across the population, demand is more sensitive to needs than costs.

Dedicated intervention studies

In coordination with the Oxford Martin School, the ReNEW programme responded to the above findings by adding a new aim of understanding the impact of more targeted interventions. Specifically, ReNEW would use the established data infrastructure to trial the ability of households to reduce demand during peak-demand hours, typically 5pm-7pm.

Fortuitously, this work pre-empted the practical need for such responses in the winter 2022/23. For the first time since the creation of the National Grid, the Electricity System Operator offered end-users a reward for reducing demand at peak hours. The Demand Flexibility Service (DFS) offers £3,000 per MWh of avoided demand and is delivered via energy suppliers or third parties.

ReNEW was able to inform this initiative with evidence from our pre-trials. Unlike DSF, which can only make aggregate data available, due to privacy and commercial sensitivities, ReNEW can give insights into the distribution of responses and lead to a deeper understanding of how and why some households are able or unable to respond at certain times. These data are now available for Oxford researchers to perform further analysis on.

Impact

ReNEW has closely engaged with many stakeholders and NGOs, including

• the Department for Business Energy and Industrial Strategy (BEIS), for whom ReNEW results provided evidence in our role as 'Expert Advisor',
• OfGEM with advice and commentary on flexibility services
• the Energy Systems Catapult, with whom ReNEW has developed research proposals and collaboration,
• Octopus Centre for Net Zero, with whom ReNEW regularly compares findings and approaches to demand side flexibility

With our insights into smart meter data access procedures and privacy issues arising from collecting, linking and sharing such data, ReNEW has informed the debate on data access via workshops and conference presentations.

Beyond academia and professional bodies, ReNEW has also engaged with outreach activities, including the production of public videos with MPLS's Oxford Sparks team and visits to schools and summer schools.

Outputs

ReNEW has gained real-world traction faster than we had anticipated at the outset of the programme. Demand side response had hitherto been considered an unreliable option for system balancing.

The acute pressures brought about by the invasion of Ukraine and resulting energy wholesale price rises, have brought the need to implement demand side measures forward by several years. Unlike continental Europe, where public campaigns were implemented immediately to reduce the reliance on Russian gas, the UK opted for market led responses, while also intervening in the market with government supported caps on unit prices.

ReNEW has shown that behavioural demand responses can contribute significantly to alleviating future system stresses. Information, as well as monetary incentives, have a role to play in this context. Going forward, the relationship between behavioural, semi-automated and fully autonomous responses needs to be better understood, to ensure basic energy needs can still be met, even during crisis events.

ReNEW insights can help to target demand response measures. Those who are most able to respond, without detrimental effects on energy service provision, can be identified and targeted with response measures, thereby delivering a more equitable transition towards net-zero.

Future plans

Two major funded relationships have emerged from ReNEW:

1. Shift-0

Funded by the MCS Charitable Foundation. MCS is responsible for the accreditation of microgeneration solutions in the UK, including photovoltaics and heat pumps. Heat pumps need to be deployed at an unprecedented rate and scale in the UK. Government ambition is to deploy 600,000 heat pumps annually from 2028. At present fewer than 2 in 1000 households have a heat pump and their operational performance in the field varies from coefficients of performance between one and five units of heat per unit electricity. Shift-0 seeks to better understand this performance variation from a technical and social perspective. We specifically target recent adopters of heat pumps. Our smart meter access provides up to 13 months of historic gas usage, which can be used as a benchmark to contrast a heat pump's electricity demand. Importantly, we combine these quantitative insights with qualitative user experience insights, using the tools developed under ReNEW.

Identifying the factors that result in successful heat pump operation can inform and accelerate the national roll out of heat pumps, while avoiding deployment in circumstances where occupants would be faced with higher or unaffordable heating costs.

2. EDOL

Tools developed under ReNEW form an important part of the new £8.7m Energy Demand Observatory and Laboratory (EDOL). Oxford was successful in securing this nationally important programme grant in part due to the pioneering work on contextualised smart meter data collection and analytical tools developed as part of ReNEW. This success has been shared on the Oxford Martin School website.

EDOL will develop ReNEW's approaches further in the following ways:

1. Scale: the ReNEW sample of n=200 allowed for valuable validation of methodologies. For national representativeness and smaller effect sizes, a larger sample is needed, which incurs considerable expense for engagement, management, data curation and instrumentation. EDOL will involve over 2000 UK homes.

2. Breadth: additional sensors and instruments, such as temperature and occupancy sensors, as well as IoT device data will be added to the existing ReNEW data streams

3. Depth: Resources are now available to follow up with in-person home visits, where observed demand patterns do not agree with model predictions. These home visits include ethnographic observations and can inform future observations on factors relating to building fabric and household practices that are currently not captured and may improve the predictive power of future models.

4. Longitudinality: Energy demand will undergo intended and unintended changes during the transition towards net zero. A longitudinal resource to track of these changes in near real-time and intervene as required, provides a powerful tool for change when these observations are inconsistent with a desirable pathway for decarbonisation.

5. Multi-disciplinarity: ReNEW already spans social, physical and data science. We currently explore opportunities to link energy-use and building performance data with health data. This could expose health benefits from retrofit measures and strengthen the case for fast action.

Multiple unsuccessful funding applications have also been submitted during the project period. These included collaborations with heating and cooling, domestic hot water, AI, ethics and heat pump demonstrators.

Solutions

A finding with immediate application is the realisation that demand responses can reliably support electricity system stability. Such evidence was needed in a cautious and conservative sector, tasked with high reliability of power provision.

Demand side responses can support and facilitate a faster and deeper deployment of renewable energy sources and reduce the need to highly polluting backup generators.

The deployment of residential heat pumps, electric vehicles, photovoltaics and batteries will enhance the ability to provide responsive services, especially if their state and user needs are well understood. ReNEW has added to this understanding and will continue to do so.