Learnings

Outcomes

The main outcomes of the project are the determination of the engagement rate, and the resulting flexibility, arising from each of the four distinct experiments that were part of the customer dataset being analysed.  
Engagement rates varied significantly between the experiments. For example, big turn up had 63% participation from smart households, while big turn down had just 2% participation from non-smart households. This indicates there is significant potential to more effectively engage households, particularly for turn down events.  

A table summarizing flexibility performance can be seen in the Appendix (figure 1) for each of the four experiments, and segmented by EV ownership. The analysis showed that there is great potential for generating flexibility from EV owning households, especially turning up demand when needed. For turn down of demand, the response is lower, indicating that non-EV loads will be important to provision of flexibility for turn down.  

The project was successful in maturing the TRL level from 2 (invention and research/concept formulation) to 4 (bench scale research/lab validation). The project confirmed the latent technical potential for the residential sector to provide flexibility at levels of national significance. Further work is required to:  

• develop tools and techniques that reflect the stochastic nature of the asset
• explore the impact of automation; to prove the reliability of the asset class to provide flexibility by evaluating response across seasons and at time of peak system stress
• explore how flexibility can be encouraged from the electric heating sector 

A comprehensive project report is available with greater detail on the project outcomes here.  

Lessons Learnt

As the project analysed pre-existing trial data, risk level was low, there were no significant changes and the planned outcomes were delivered successfully. The process did generate insights that are being carried forward into the design of pre-commercial trials,through the CrowdFlex SIF Discovery project and these include:

• The need to develop statistically reliable baselining processes, in particular to deal with the challenge of the stochastic nature of the asset.
• EV ownership was a primary determinant of flexibility provision, as was expected in the scope. The analysis showed that EV demand can be encouraged to increase demand, however one-off big turn down events look to rely on other assets such as white goods. Future work needs to explore turn down in more detail.
• The analysis indicated that the impact of season on flexibility potential looks to be important, but a bespoke trial will be needed to explore this issue fully.  

The project was very effective in demonstrating the technical potential of residential flexibility. 

• The analysis undertaken in the project showed that residential flexibility can be encouraged through a number of mechanisms and the aggregated flexibility capacity can be of national significance. For example, the project estimated that turn up could deliver 37GW of demand turn up, or 6.8GW of demand turn down, by 2030.
• This confirms that this novel asset class has significant latent potential and the sector should continue to be explored to move towards commercially viable sources of reliable residential flexibility.
• The analysis showed that the enduring response to ToU tariffs was strong and sustained, at least over the 6 months of data analysed in the project. Analysing more than one year of ToU tariff data is a goal for future trials.
• The big turn up experiment was shown to be very effective, and heavily reliant on EV demand. Big turn down was less effective, and future work will need to explore how best to incentivize flexibility from non-EV demand.