The forecasted growth of heat pump and EV chargers exposes networks to increasing challenges associated with cold load pickup following a prolonged outage, i.e. the initial demand could be beyond design parameters. Having a method to control and gradually ramp up demand after these prolonged shutdowns will allow DNOs to maintain network stability and reliability.
Smart Restart will develop and simulate a methodology for mitigating cold load pickup through smart meter-based load control. Central to the approach is the random offset feature built into smart meters, which could delay the reactivation of devices (such as heat pumps and EV chargers). Staggering when devices switch back on could prevent a sudden demand surge that could otherwise lead to a partial or total network shutdown.
Benefits
This project is categorised as a Research project; however, an indicative benefit pathway has been considered based on the counterfactual scenario.
· Base Cost (Counterfactual): In the absence of the Smart Restart methodology, unmanaged reconnection following a prolonged outage can lead to excessive peak demand, which may exceed substation or feeder capacity. The counterfactual involves capital-intensive reinforcement of LV feeders, installation of additional monitoring/control assets, or reactive operational responses to secondary outages—all of which increase system costs and customer disruption.
· Method Cost (Smart Restart): The proposed approach leverages existing SMETS infrastructure to enable staggered reconnection via randomised offsets. This can be achieved without additional hardware installation, offering a significantly lower-cost alternative. It also has the potential to reduce engineering dispatch requirements during recovery, allowing resources to focus on more urgent or complex faults. Additionally, the result from Smart Restart can also help the thinking behind Regional Energy System planning.
A high-level calculation suggests that, if implemented effectively, Smart Restart could reduce the occurrence of secondary outages due to cold load pickup. Based on potential savings from avoided Customer Interruptions (CIs) and Customer Minutes Lost (CMLs), the solution could deliver annual benefits of approximately £1.5 million, assuming 90% of these secondary outages are prevented.
Recipients of the benefit include:
· Consumers, through improved supply reliability and lower network costs (which flow through to bills)
· Network operators, through reduced reinforcement, fewer site visits, more efficient use of existing assets, helping the thinking behind Regional Energy System Planning
· The wider system, through enhanced grid flexibility and readiness for increasing LCT uptake
Learnings
Outcomes
The Smart ReStart project investigated whether existing smart meter functionality could mitigate Cold Load Pick-Up (CLPU), an emerging risk linked to the electrification of heat and transport. The work combined literature review, regulatory assessment, technical evaluation, and operational analysis.
Overall, the project finds that CLPU is a credible and increasing potential risk but also demonstrates that existing smart meter capabilities are not sufficient to provide a clear or deliverable solution at present. While some technical options show partial promise, limitations across regulation, system design, operational feasibility, and device coverage mean that no defined pathway to implementation has been established.
Literature review of Cold Load Pick Up
The literature review indicates that CLPU is likely to become more pronounced as low-carbon technologies (LCTs), particularly heat pumps and electric vehicles (EVs), increase. Modelling suggests that under certain scenarios:
· Post-outage demand could rise significantly above normal levels
· LV feeders could face overloading under high electrification conditions
· Household demand could increase materially during recovery
However, these projections are subject to uncertainty. Outcomes depend heavily on assumptions around behaviour, weather, and technology performance, meaning the scale, timing and location of impacts remain uncertain.
CLPU is a whole-system effect, driven by combined demand across heating, background load and EV charging. Electrified heating appears particularly important due to sustained recovery behaviour, although further modelling may be needed to quantify this impact.
The persistence of the CLPU spike is also important. The Cold Start analysis indicates that elevated demand can last well beyond the initial reconnection period, meaning that short, manual interventions or simple reconnection sequencing are unlikely to be sufficient in an electrified system.
The literature clearly demonstrates that CLPU is a real and increasing network risk, driven by the accelerating uptake of heat pumps and EVs. Without intervention, or significant LV network upgrade, post‑outage demand could exceed network design limits in multiple GB regions by 2030.
Regulatory Review (WP2)
The regulatory framework does not explicitly prevent the use of smart meter functionality for resilience purposes, but it does create substantial barriers.
The most significant constraint is that DNOs lack authority to control smart meter functions, with permissions held by suppliers under the Smart Energy Code. There is no established mechanism for coordinated use during restoration.
Additional challenges include:
· No standard approach to customer consent or acceptable use
· Unclear liability arrangements
· Limited precedent for network-led control
· Lengthy and complex code modification processes
Although policy developments (e.g. Smart Secure Electricity Systems (SSES), EV smart charging regulations, RIIO-ED3) indicate a shift toward demand-side control, these are not yet sufficiently defined to enable implementation. Regulatory change remains a key dependency with uncertain timelines.
The regulatory review concludes that Smart ReStart is feasible within the existing GB smart‑metering architecture, but cannot be implemented today without changes to permissions, consent mechanisms, and industry governance.
Technical Review (WP1)
Smart Meter Capabilities Today
The GB smart metering system provides a secure and interoperable communications platform through the Data Communications Company (DCC). SMETS2 meters support a range of functionalities that, in principle, could influence electricity demand following the restoration of supply after an outage. Although these functionalities were primarily designed to support supplier-led activities such as billing, prepayment and tariff management, some may also have potential applications in mitigating Cold Load Pick-Up (CLPU).
However, the current smart metering architecture was not developed specifically for operational network restoration. Consequently, several characteristics of the existing framework affect the extent to which smart meter functionality can be used to manage CLPU events.
Trigger Behaviour
SMETS2 meters include a randomised offset capability that introduces a delay before a control action is executed. The purpose of this functionality is to avoid large numbers of devices responding simultaneously to scheduled events, such as tariff changes. From a CLPU perspective, a similar approach could potentially reduce the rate at which demand returns following restoration by staggering the reconnection of controllable loads. However, under current arrangements, randomised offsets are only applied to scheduled events and are not triggered by power restoration. This limits their applicability as a CLPU mitigation measure in their current form.
Device Availability and Diversity
A number of smart-meter-enabled control technologies exist, although their deployment is not universal.
Auxiliary Load Control Switches (ALCS) are meter-connected switching devices that provide remote on/off control of dedicated electrical circuits, such as storage heaters or immersion heaters. Home Area Network Connected Auxiliary Load Control Switches (HCALCS) provide a similar capability for devices connected through the smart meter Home Area Network (HAN), potentially including technologies such as EV chargers. These technologies could contribute to CLPU mitigation by delaying or sequencing the reconnection of selected loads following restoration. However, their deployment remains limited, and their operation is restricted to binary control, whereby devices are either energised or de-energised.
SMETS2 also supports Auxiliary Proportional Control (APC) and Smart Auxiliary Proportional Control (SAPC). Unlike ALCS and HCALCS, these functionalities enable compatible devices to operate at varying power levels rather than simply switching on or off. In principle, this could facilitate a more gradual recovery of demand following restoration. However, current deployment levels are low, and the functionality is not readily applicable to several technologies expected to contribute significantly to future CLPU, including many heat pump installations.
Communications Characteristics
Communications between authorised users and smart meters are managed through the DCC using processes defined within the DCC User Interface Specification (DUIS). DUIS specifies the service requests used to exchange information and issue commands to smart metering devices. Commands are not delivered instantaneously and may be affected by communications latency, batching processes and retry mechanisms. As a result, smart meter actions occur over a period of time rather than at a precisely defined moment. Any operational approach that relies on smart meter communications must therefore account for these delivery characteristics.
Taken together, these characteristics indicate that while several SMETS2 functionalities may have relevance for CLPU mitigation, their effectiveness is influenced by factors including deployment levels, communications architecture and the intended design of the existing smart metering framework.
Assessment of Relevant Functionalities
The technical review considered a number of SMETS2 functionalities that could potentially support the management of post-restoration demand.
Randomised Offset introduces a delay before a control action is carried out. If applied during restoration, it could reduce coincident demand by staggering the reconnection of controllable loads across a customer population. However, the functionality is not currently triggered by restoration events and is limited to a maximum delay of approximately 30 minutes. Given that CLPU effects may persist for several hours, particularly where electric heating and EV charging are present, the potential contribution of randomised offset is likely to be limited if deployed in isolation.
ALCS and HCALCS (Auxiliary Switching) provide remote switching of specific loads or devices. These functionalities could support restoration strategies by delaying the reconnection of selected appliances during the initial stages of demand recovery. Their effectiveness is constrained by relatively low deployment levels, dependence on compatible equipment and the binary nature of the control they provide.
APC and SAPC (Proportional Control) provide a more flexible form of control by enabling compatible devices to operate at reduced power levels. This could support a more gradual restoration of demand and potentially reduce peak loading following an outage. However, deployment remains limited and the application of proportional control to heat pumps may be restricted by technical operating requirements and manufacturer constraints.
Load Limiting (LL) enables a temporary cap to be placed on total household demand at the meter. Unlike device-specific control mechanisms, load limiting acts at the whole-property level and is therefore independent of the types of appliances connected within the home. As a result, it offers a potentially effective means of constraining post-restoration demand and reducing loading on feeders and transformers. Nevertheless, the functionality is currently associated primarily with prepayment arrangements, and wider use for network resilience purposes would require regulatory approval, governance arrangements and appropriate customer protection measures.
Telemetry and Readback Functions enable information to be retrieved from smart meters regarding supply status and device operation. While these functions do not directly reduce demand, they could improve visibility of restoration progress and support the coordination of staged restoration strategies. Their value for operational use is currently constrained by limitations on DNO access to detailed smart meter data.
Communications Constraints were also considered as an overarching factor affecting all smart-meter-based approaches. Latency and retry cycles within the DCC communications architecture mean that instructions are delivered progressively rather than simultaneously. While this limits the scope for highly time-sensitive control, it may also introduce some natural diversity in the timing of customer demand restoration.
Conclusion
The technical review indicates that SMETS2 incorporates a number of functionalities that could contribute to the mitigation of CLPU. However, each functionality is associated with technical, operational or regulatory limitations that restrict its effectiveness as a standalone solution. Load limiting appears to offer the most direct mechanism for constraining post-restoration demand at the household level, whilst randomised offset, auxiliary switching and proportional control could provide complementary benefits under appropriate circumstances.
Overall, the findings suggest that CLPU mitigation is unlikely to be achieved through any single smart meter functionality. Rather, the evidence supports the development of a broader portfolio of complementary measures, with the most appropriate combination likely to depend on network characteristics, customer technology uptake and the operational context of the restoration event.
Operational Review (WP3)
The operational review examined current low-voltage (LV) network restoration processes, the practical constraints associated with using smart meter functionality during outage events, and the operational changes that would be required to implement Smart ReStart. The review drew on stakeholder engagement with Distribution Network Operators (DNOs), suppliers, the Data Communications Company (DCC) and industry code bodies, alongside an assessment of existing restoration practices and smart metering operational processes.
Current LV restoration procedures are based on established engineering practices. Once a fault has been resolved and supply is restored, customers' appliances typically return automatically to their pre-outage operating state. There is currently no coordinated mechanism for controlling how individual devices reconnect, nor is there a defined operational process linking DNOs, suppliers and the DCC during restoration. As a result, heat pumps, EV chargers and other high-demand appliances may resume operation simultaneously, contributing to CLPU risk.
The review also identified a number of operational constraints that influence the feasibility of Smart ReStart. These include limited visibility of low-carbon technologies connected to individual properties, the absence of real-time control capabilities within current restoration processes, communications latency across the smart metering ecosystem and the fact that DNOs do not currently possess direct authority to control relevant smart meter functions. Consequently, any future Smart ReStart solution is likely to depend on pre-configured behaviours and agreed operational processes rather than real-time intervention during an outage event.
Based on the findings of the review, several areas for further development were identified:
· Regulatory and code development
· Operational procedures
· Customer engagement and protections
· Industry governance
· Modelling and validation
The findings indicate that Smart ReStart has the potential to form part of a future DNO operational toolkit for managing LV networks during restoration events. However, implementation would require a phased approach. This is likely to involve further modelling and analysis, followed by controlled trials under an appropriate regulatory framework, before progressing towards a standardised industry process and wider deployment.
Further detail on the operational review methodology, findings and recommendations is provided in the full WP3 report on the SP ENW website here.
https://www.enwl.co.uk/future-energy/innovation/smaller-projects/network-innovation-allowance/enwl041--smart-restart-nia/
Lessons Learnt
The Project identified a number of lessons for future projects. These include starting engagement with smart meter governance bodies earlier, so that any regulatory barriers are flagged at the start of the Project. This type engagement could possibly be considered pre-project to help define the required outputs for future projects. The project also discovered that smart meter functionality should not be assessed in isolation from operational and regulatory constraints. Future innovation projects should engage suppliers and device manufacturers from project inception to ensure that any specific limitations are considered. Future CLPU studies would also benefit from access to higher-resolution network and customer demand data. Network resilience solutions are likely to require a portfolio approach rather than reliance on a single technology, so any future projects will need to balance a number of elements.
