Demand-side response is falling when net zero plans need it to rise

In 2022, National Grid ESO (now NESO, the National Energy System Operator) launched the Demand Flexibility Service (“DFS”) as part of its winter contingency against a backdrop of concerns over security of supply in the wake of the Ukraine war. In its Winter Outlook for 2024/25, NESO has determined that the spare margin is higher than in previous years, reducing the need for a contingency. Instead, it proposed to use the DFS as an in-merit margin management tool, and as a result, removed availability payments and reduced the overall economics of the scheme.

Changes to DFS make it less valuable to consumers

This year, NESO has determined that it does not require a contingency for security of supply and requested changes to the operation of the DFS, which Ofgem approved in late November. Since there isn’t a need for the DFS as an enhanced action, NESO will operate the DFS as an in-merit margin management tool which will only be dispatched when it is competitive against alternative actions, ie, the total cost of activating DFS volumes will be considered against the total costs of interconnector or Balancing Mechanism (“BM”) actions, be based on NESO’s forecast of its margin requirement within day.

Moving in to an in-merit service means that the Guaranteed Acceptance Price (“GAP”) has been removed. The GAP previously ensured providers received at least £3,000 /MWh during most events – without it, payments to customers under the DFS are expected to be as much as 90% lower than last year. While customers were paid £1-4 /kWh saved last year, they are likely to only receive around 10-30p /kWh this winter.

In Winter 2022, 1.6 million households and businesses took part in the DFS, rising to 2.2 million the following year. So far this year, 650 MW of DFS has been procured, but at no time has the amount procured matched the requirement, suggesting alternatives such as interconnectors and the BM have been more cost effective. This year there are 14 domestic providers registered, and 18 non-domestic, which aggregate demand across their customers.

So far there is no information about how many end users are signed up, but anecdotally, some 90% are apparently customers of Octopus Energy, as was the case in previous years, possibly because of the enhanced economics the company offered on top of the payments from NG ESO. Octopus expected participation this year to be much lower following the removal of the GAP. It surveyed 2,500 of its customers, 1,700 of whom responded, with 80% of the responding participants saying they were unlikely to engage with the DFS if the payments were cut.

However, it is difficult to justify paying DFS participants amounts which could be 10 times higher than the prices of alternatives such as interconnector and BM trading when the costs of the scheme are socialised across all electricity consumers. This would be deeply regressive, since the consumers most able to engage in demand-side response are those with high value assets such as electric cars, which are unaffordable to the poorest consumers.

The French experience of DSR is also poor

Countries with higher levels of electric heating rely more on demand-side flexibility than Britain does, for example, France where electric heating has been widespread for many years. The French grid operator Le Réseau de Transport d’Électricité (“RTE”) has developed sophisticated demand flexibility tools, with large industrial consumers typically representing the principle source of demand response. In the 1990s, the potential of demand response that could be mobilised was estimated to be around 5-6 GW, but this potential had fallen to 2.5 GW by 2017. France has set national objectives for demand side response of 4.5 GW in 2023 and 6.5 GW in 2028. In 2020, the available demand side response capacity was 3.2 GW, falling short of ambitions at the time.

The existing demand-side response schemes in France are managed through day-ahead forecasting and real-time adjustments, considering factors such as weather data and consumer behaviour, and include:

• The “Demand Response Block Exchange Notification” (“NEBEF”), where consumers reduce their demand in response to price signals or requests from the grid operator. This mechanism treats demand reductions similarly to energy generation, with remuneration provided to participants, and is open to consumers with demand of at least 1 MW (or aggregators who aggregate smaller loads which add up to at least this amount);
• Capacity Remuneration: this is similar to the Capacity Market in GB, where consumers and generators can receive payments for making capacity available during peak demand periods. In France, this includes both capacity (in MW) and actual energy reductions (in MWh), which are activated on request to help balance supply and demand;
• Policies and Targets: The French government has set ambitious targets for demand response as part of its Multiannual Energy Programme (“PPE”), aiming to enhance the flexibility of the electricity system in line with increasing renewable energy integration. The potential for demand response, especially from large industrial consumers and, increasingly, from residential and SME sectors, is seen as crucial for maintaining grid stability during extreme weather events.

In addition, Electricité de France (“EDF”), the main electricity supplier in France, operates a consumer tariff similar to the DFS. The “Tempo” tariff is designed to manage electricity demand and costs, incorporating a traffic light-like system with blue, white, and red days, each having different pricing. This system replaced the older “EJP” (Effacement Jour de Pointe) tariff, which had similar demand management features.

Under the Tempo system:

• Blue days are the cheapest, with users paying prices that are 30% lower at peak times (from 06:00 to 22:00) and 41% lower at off-peak times, compared to the regulatory price. This covers 300 days per year;
• White days are moderately priced, and occur on 43 days. Savings are still 10% at peak times and 34% at off-peak times;
• Red days are the most expensive with prices three times higher than the normal rate, and limited to 22 days during winter (1 November to 31 March). The price per kWh is three times higher than the normal rate in order to encourage consumers to reduce consumption during peak demand periods, in order to manage load on the electricity grid more efficiently. Even on red days, users can save up to 16% during off-peak hours if they cut down on usage.

Clients receive an alert by SMS or email notifying them that the next day will be a red day, or are notified through the EDF app and website. The colour also shows up on their electricity meter from 20:00 the night before.

While there will never be more than five red days in a row, and red days cannot take place on weekends and public holidays, combinations of successive red and white days are possible, which make it difficult for consumers to manage their demand. For example, there is only so long a household can go without doing laundry. In addition, reducing heating levels for extended periods can have very detrimental health implications for vulnerable consumers for example by exacerbating respiratory or circulatory illnesses. 

EDF has 70% of the market for individual households and 55-60% of the market for business customers. However, despite the very long history of variable tariffs, only 500,000 or 1.6% of households are on the Tempo tariff, against a target of 5 million. The vast majority of households prefer to be on the regulated tariff. Although some studies have shown that up to 100% of households would be better off on Tempo than the regulated tariff, it suffers from low public awareness, and problems for households if a series of red or white days occur on consecutive days.

Due to the relatively low number of households using the Tempo tariff, the contribution to demand reduction from the residential sector remains low, despite this type of tariff having been a feature of the French market for decades. This suggests that realising the potential for demand-side flexibility from the residential sector could be challenging unless these schemes can be shown to deliver higher consumer benefits in terms of price, and are automated so that they rely on less on active demand reduction. Currently both Tempo in France and the DFS in Britain depend on consumers taking active steps to reduce consumption – grid operators hope that automated energy management systems will deliver larger results, but there are considerations to be resolved around customer autonomy and consent.

Business participation in demand reduction has collapsed

Winter 2023/24 was the first in which triads had been abolished. Triads were the three half-hours with highest system demand between November and February, separated by at least 10 business days. They were used in the determination of Transmission Network Use of System (“TNUoS”) charges – businesses were charged based on their demand during the triads.

This led businesses to engage in “triad avoidance” – a process whereby they paid consultants to advise them on the when the triads may take place, so they could reduce their consumption in those periods, thereby reducing their TNUoS charges. Eventually Ofgem decided to abolish triads on the basis that triad avoidance was largely a zero sum game – businesses participating in triad avoidance meant that those that did not or could not ended up paying more. There were some arguments that by reducing demand at peak times, grid expansion could be avoided, however, with the growth in renewable generation and electrification, grid expansion will be necessary, and it’s possible that the entire triad approach simply delayed rather than avoided network investment (possibly to the detriment of the market).

With the retirement of the triads, it was expected that businesses would sign up for the DFS instead, but so far this has not materialised in any significant way – the vast majority of participants in the DFS are households. The typical reduction in demand realised by triad avoidance was around 2 GW, which is far larger than the demand reduction achieved through the DFS, even when it was being used for security of supply reasons. This suggests that businesses do not see the same level of benefit from the DFS that they did in triad avoidance, where 25-30% savings on network costs could be realised.

There was a small increase in demand-side response participation in the Capacity Market. In the T-1 capacity auction for WIN-24, 59 units comprising 773 MW of DSR (de-rated) capacity were successful in securing contracts, of which 600 MW was from Unproven DSR units. The previous T-1 auction for WIN-23 secured 726 MW of DSR. This increase is still far below the volumes realised through triad avoidance.

DSR should be a no-brainer but is proving to be easier said than done

Demand-side flexibility should be a no-brainer – so-called “negawatts” being, in theory, the cheapest form of supply (although in the case of businesses, this can mean on-site generation, so not strictly “negawatts”). But in reality things are not so easy. Where DFS must compete with other sources of supply, it struggles to demonstrate compelling economics for users, in large part because under normal market conditions, the entities with the greatest ability to flex demand are those with higher affluence and therefore a lower need to achieve savings. This is particularly true in the domestic market and explains why, in the absence of automation, it is likely to be difficult to realise large volumes of domestic DSR.

It would therefore make more sense to seek response from industrial and commercial consumers, but for these users, the schemes must be both easy to adopt and avoid interfering with business operations. The success of triad avoidance indicates that this can be done – 2 GW is a significant amount of peak shaving which is larger than the current largest single source of supply or generation in the GB market. But for whatever reason, neither the DFS nor the Capacity Market is delivering a comparable appeal to these consumers. In the non-domestic segment, automation is already well established by aggregators, with issues of consent being less fraught.

That it is difficult to identify any country which is successfully harnessing DSR to deliver meaningful volumes (10% of peak demand or more), confirms that this is not a simple problem to solve. This report from ACER sets out some of the challenges, which are relatively straightforward to describe but harder to solve. Equipment upgrades such as smart meters, device automation (eg programmable appliances) and control software need to be deployed at scale. And market frameworks need to be upgraded to provide access to all aspects of the market and sufficient price signals for response.

Policymakers love the idea of flexibility. The Clean Power 2030 Plan calls for demand-side flexibility to increase between four and five times to 2030, from 2.5 GW today to 10-12 GW in 2030.

Most of the growth is expected to come from smart EV charging. But there are headwinds – growth in electric car sales is slowing, despite reports that manufacturers are rationing the sale of petrol cars to meet Government targets. Increasing insurance costs for electric cars are starting to dent their popularity. In addition, access to smart charging does not mean people will use it – a Government study in 2022 found that 26% of electric car owners and 49% of plug-in hybrid owners with private charging facilities opted to use a three-pin plug instead of flexible charge facility.

Which is to say that an awful lot needs to happen for the growth in demand-side flexibility required by CP2030 to be realised, and right now, things are moving in the opposite direction with the changes to the DFS and the removal of triads as a basis for network charging. There will need to be a significant change to deliver the demand-side participation policy-makers are hoping for.