What is the best way to deploy DFS next winter?

Developed in just four months, National Grid ESO’s Demand Flexibility Service (DFS) attracted widespread coverage. Energy UK senior policy analyst Naomi Baker looks at the key learnings from the winter just gone and what changes could be made to the service before next winter.

Key learnings

DFS proved that while manual turn down from individual households is not firm, aggregated domestic turn down can provide a reliable grid balancing tool.

The DFS delivered:

• 350MW capacity (similar to a small transmission-connected thermal plant)
• 31 providers
• 6 million reach
• 16 tests with 1.6GWh turn-down
• Two live events with 0.6GWh

Octopus Energy, (the only supplier to offer the service to all customers with a functioning smart meter) delivered over a third of the volume (with one in every four eligible customers participating).

Best way to deploy DFS for winter 2023/4

With continued risk of supply chain disruption and less coal back-up, the best application is again as a winter contingency service to preserve system margins. Providers recommend a similar design but with a focus on expanding volumes and adding an option for closer to real time dispatch (Octopus found that at least 300,000 customers were willing to deliver turn down [totalling 100MW] with four to six hour lead time).

The following design changes would support the service to expand:

• Implementing a trusted third-party administered switching mechanism.
• Commitment to a successor programme to support a shift from manual to API-based processes.
• Allowing stacking with the Capacity Market (CM) to unlock industrial and commercial demand.
• Allowing asset metering to improve alignment with ESO balancing services.
• Clearer dispatch processes for greater price discovery

With these changes, an expanded DFS could provide a system buffer at a lower cost than last winter’s arrangements.

Going forward

While DFS has functioned as a dispatchable capacity market, from spring 2024 this back-up is unlikely to be needed. This new capacity then could pivot to help with system balancing and network constraints, easing teething pains as the system transitions.

Despite the potential, demand response has largely been expected to fit into services and dispatch platforms designed for large centralised generators. DFS shows what can happen if the approach is flipped.

As the Ofgem consultation into the Future of Distributed Flexibility makes clear, new demand for vehicle charging and heating could support the system, smoothing peaks, or, act as parasitic loads, exacerbating peaks and requiring additional infrastructure.

It was assumed that change here would accelerate with the completion of the smart meter rollout and market wide half hourly settlement from 2027. However DFS momentum, a creaking grid and escalating system costs (balancing and constraints cost £4 billion and £2 billion respectively last year) provide both an opportunity and a need to move faster.

While there are challenges (how to account for ‘avoided costs’ of infrastructure not required?), the business case from the systems perspective is clear. Crowdflex suggests that domestic flexibility could reduce peak demand by 10%. With a winter peak of 49GW, 2022, this is a potential prize of 5GW. If demand can meet system needs at a lower overall cost than generation, it is a win-win for customers. Octopus estimates there is already 2GW of controllable demand in electric vehicles as well as large, untapped domestic potential. The key to creating value here rests with providers ability to aggregate and shape loads to deliver on system needs.

The energy sector excels at complex high-capex technologies. The challenge here instead requires customer engagement, software, data and business development, time to evolve and learning by doing. If there is not a solid business case for providers, this learning will be delayed. 2027 may also be too late. Unless we are confident that demand can plug the gap, additional infrastructure will be needed. By 2027 then, the future shape (and costs) of the energy system could be baked in.

Energy UK is asking the ESO to collaborate with industry to develop a successor scheme to run from April 2024. This ‘bridging’ service would transition demand response to a fully commercial service.