Energy Systems Catapult collaborated with Scottish and Southern Electricity Networks (SSEN) on phase two of their Demand Diversification Service (DDS) project for Load Managed Areas (LMAs).

The project explored whether market-based flexibility services can help keep local electricity networks running smoothly as more homes adopt electric vehicles (EVs), heat pumps, and modern electric heating.

The challenge

For decades, SSEN have relied on the Radio Teleswitch Service (RTS) to shift demand from northern Scottish homes into the nighttime by controlling storage heater operation via longwave radio signals. Geographic regions where high demand loads like storage heaters can be controlled are designated Load Managed Areas (LMAs).

SSEN have committed to phase out LMAs to enable consumers to have greater choice in their supplier and tariff. However, parts of the network still need the diversity that LMAs provide, so SSEN are considering modern alternatives that protect consumers while maintaining demand diversity.

Additionally, the shift to EVs and heat pumps is changing when people use electricity, often clustering demand in the evening and overnight. That can strain local substations if lots of devices switch on together. DDS tests a smarter approach: using flexibility to shift when devices run, without compromising comfort.

Our solution

We combined the Catapult’s Living Lab with the WESA (Whole Energy Systems Accelerator) – developed with PNDC (University of the Strathclyde) – to test two flexibility approaches in real homes: Allocated Capacity (AC) and Dynamic Congestion Response (DCR).

In simple terms, AC caps the total demand a group of houses can draw at busy times, while DCR asks the group to dial devices up or down in response to local network conditions.

Using WESA, we mapped participating homes onto a virtual local network in real time. That let us send price or control signals to devices, observe the impact on a simulated substation, and gather feedback from households – all safely, and at a scale that would be impossible to test on the live system, as the current penetration of low carbon technologies like heat pumps and EVs is not at sufficient levels.

Delivery partners included PNDC (University of Strathclyde) for network simulation, ev.energy for EV smart charging, Connected Response for storage heating, and Homely for heat pump scheduling, with SSEN as commissioning client throughout.

What we did

Created virtual communities of high LCT uptake

We recruited 165 households with EVs, heat pumps and storage heating and aggregated their energy consumption to simulate demand as if they lived on the same street or feeder. This let us test what happens when lots of low-carbon devices operate in a small area at the same time, without any risk to the real network. It also allowed us to test a hypothetical scenario that does not currently exist anywhere on the network, but one that the grid is likely to face in future as LCT uptake increases.

Linked real homes to a safely simulated local network (WESA)

Using the Whole Energy Systems Accelerator (WESA), we connected live data from those homes to PNDC’s real-time network simulator. In practice:

• Homes ran their devices as normal, with smart controllers and apps in the loop.
• We sent simple pricing or control signals to the homes; devices automatically shifted when they charged or heated (for example, delaying EV charging or moving heat pump runtime).
• The homes’ real-time electricity use flowed to PNDC, where engineers measured the impact on a simulated local substation and feeders (including asset loading and voltage).
• If the simulated network looked too busy, the next signals were adjusted (e.g., day-ahead schedules or “turndown” windows) to spread demand and keep the virtual assets within limits.

This closed feedback loop let us trial flexibility safely, quickly and with real people –something that’s hard to do on the live system.

Ran targeted sprints to compare approaches

The project was structured into five live trial sprints, each lasting between one and five weeks, with a sixth sprint dedicated to deeper analysis of the data collected. Sprints were the core unit of the trial – defined periods during which specific combinations of asset types and flexibility mechanisms were actively tested, with control signals sent to participants’ devices throughout.

The sprints were sequenced to build up complexity progressively. Sprint 1 (December 2024, one week) started with EV participants only, to establish a baseline before adding further asset types. Sprints 2 and 3 (January-March 2025) brought in storage heater and heat pump households, building up to the full participant cohort of 165 homes. Sprint 4 (April-May 2025) switched to testing the Dynamic Congestion Response mechanism across all asset types. Sprint 5 (June 2025) returned to Allocated Capacity but introduced a deliberate experiment: a subset of EV participants agreed to switch their scheduler to a flat tariff, removing the strong overnight off-peak window incentive and revealing the technical headroom available when retail tariff constraints are lifted.

Between sprints, participants’ devices returned to normal scheduling with no signals sent – these “between-sprint” periods formed the counterfactual baseline used in the final analysis to measure what difference the DDS signals actually made.

Consumer Insight (CI)

We gathered feedback through surveys and interviews, and monitored app interactions to understand comfort, trust, ease-of-use and motivations. This helped us learn what makes automated flexibility acceptable and what gets in the way (for example, confusing app design).

What we found

• Flexibility works, when incentives and usability align:
  • Most households noticed little or no change in comfort.
  • Automation is valued when it’s reliable and transparent; poor app experiences erode trust.
• Financial incentives were the strongest motivator.
• Unconstrained storage heaters demonstrated the strongest absolute peak reduction (around one-third) within 24 hours of controls being applied, suggesting significant potential once legacy tariff constraints are addressed.
• EVs showed the largest headroom: removing strict off-peak windows (simulated flat tariff) enabled up to ~50% reduction in peak EV charging load in aggregate.
• Heat pumps responded reliably to day-ahead scheduling but contributed smaller volumes of flexibility per home.
• Storage heating was most constrained by legacy hardware and rigid tariffs; where unconstrained, meaningful peak reductions were demonstrated.
• Product design matters: near-real-time DCR was operationally complex and, as a consequence, day-ahead control is recommended for business-as-usual. AC is simpler and delivered predictable outcomes.

The single biggest barrier to local flexibility: national retail tariffs

One of the most important findings from the trial was that the financial incentives built into national retail electricity tariffs – particularly time-of-use tariffs that offer heavily discounted rates overnight (such as ~9p/kWh off-peak versus ~33p/kWh at peak times) – fundamentally overwhelm the signals that a local flexibility service like DDS can send. When a household’s energy bill rewards them so strongly for charging an EV or running a storage heater at midnight, a comparatively modest local flexibility incentive simply cannot compete. The result is that DDS signals are largely overridden, not because the technology doesn’t work, but because the economics make it irrational for households or their smart schedulers to respond outside their already-off-peak periods.

This was demonstrated most clearly in Sprint 5, when a cohort of EV participants agreed to switch to a flat tariff for the trial period, removing the overnight off-peak window entirely. The effect was striking: peak EV charging loads fell by up to ~50%, with charging spread much more evenly across the evening and night. The technology and the control mechanisms worked; it was the tariff structure that had been preventing them from doing so. This highlights a fundamental issue: less expensive retail tariff periods align with national wholesale markets, and these increasingly conflict with the needs of local distribution networks.

This finding has significant implications for the future of DDS and local flexibility more broadly. If demand diversification services are to succeed as a replacement for legacy RTS control – or to provide meaningful local flexibility as EV and heat pump uptake grows – national energy suppliers and the tariffs they offer will need to be part of the solution. This could mean suppliers designing tariffs whose off-peak windows are aligned with local network conditions, rather than a fixed national overnight window; or DDS incentive payments being structured to be genuinely competitive with the savings consumers currently receive from time-of-use tariffs. Without this alignment, local networks and national retail markets will continue to pull in opposite directions.

Why this matters

The trial shows how market-based flexibility can support local network resilience as LMAs are phased out and low-carbon technologies scale. Insights will inform SSEN’s pathway to adopt DDS in business-as-usual (with day-ahead operation) and help shape future product, tariff and consumer experience design that delivers benefits for networks and households.