1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar

On the way to a smart, flexible GB energy system? Part 1 (overview and storage)

Things may be starting to move a bit faster in the world of GB energy policy after what you could be forgiven for thinking was a Brexit-induced slowdown. On 24 July 2017, the UK government’s Department for Business, Energy and Industrial Strategy (BEIS) and the energy regulator Ofgem published a number of documents that reveal their evolving thinking about the future of the GB electricity system. These publications followed on from some significant initiatives by Ofgem and National Grid. This is the first of series of posts assessing where all this activity may be leading.

The full holiday reading list from 24 July was as follows.

Other recent official publications that are relevant in this context and referred to below include:


The Response and the Plan cover a broad range of subjects; many of the other documents are rather more monothematic. We will follow the topic headings in the Response, referring to the other documents where they are relevant. However, it is helpful to start by framing some of the key themes underlying this area of policy by turning to the Pöyry / Imperial Report.

The CCC has recommended that in order to achieve the ultimate objective of the Climate Change Act 2008 (reducing UK greenhouse gas emissions by 80% by 2050), the carbon intensity of the power sector should fall from 350 gCO2/kWh to about 100 gCO2/kWh by 2030.  Pöyry / Imperial observe that in any future low carbon electricity system, “we should anticipate:

  • a much higher penetration of low-carbon generation with a significant increase in variable renewable sources including wind and solar and demand growth driven by electrification of segments of heat and transport sectors;
  • growth in the capacity of distribution connected flexibility resource;
  • an increased ‘flexibility’ requirement to ensure the system can efficiently maintain secure and stable operation in a lower carbon system;
  • opportunities to deploy energy storage facilities at both transmission and distribution levels; and
  • an expansion in the provision and use of demand-side response across all sectors of the economy.

System flexibility, by which we mean the ability to adjust generation or consumption in the presence of network constraints to maintain a secure system operation for reliable service to consumers, will be the key enabler of this transformation to a cost-effective low-carbon electricity system. There are several flexibility resource options available including highly flexible thermal generation, energy storage, demand side response and cross-border interconnection to other systems.”.

This explains why technologies and mechanisms that can increase system flexibility are a dominant theme in current GB electricity sector policy-making. But Pöyry / Imperial then go on to discuss the extent of the uncertainty that, based on their modelling, they consider exists about how much the main types of flexible resource may be needed on the way to achieving the CCC’s target. This is clearly shown in the table, reproduced below, setting out their assessment of “the required range of additional capacity of different flexible technologies to efficiently meet 2030 carbon intensity targets”.

With the exception of interconnectors, the table shows the amounts of each flexible technology in the low and high scenarios, at each of the three dates, varying by a factor of 5 or more. As regards interconnectors, an illustration of the potential uncertainties in the different scenarios modelled by National Grid in FES 2017 is provided by the two FES 2017 charts below.

Source: National Grid, FES 2017

Source: National Grid, FES 2017

The need for more flexible resources is clear, and Pöyry / Imperial calculate that integrating them successfully, as compared to the use of “conventional thermal generation based sources of flexibility”, could save between £3.2 billion and £4.7 billion a year in a system meeting the CCC’s 2030 target.  But it is also clear that there are many different possible pathways that could be followed to achieve this level of flexibility, and that even if we get to 100 gCO2/kWh by 2030 – which is by no means guaranteed – there will inevitably be, at least relatively speaking, “winners” and “losers” in terms of which flexible technologies, and which individual projects, end up taking a greater or lesser share of what could be loosely called the “flexibility market”.

What will determine who wins or loses out most in this competition will be the same factors as have driven changes in the generation mix in the UK and elsewhere in recent years – in particular, the interplay of regulation and technological change.  In 2016, as compared with 2010, the UK consumed 37% less power generated from fossil fuels and more than twice as much power generated from renewable sources: see the latest Digest of UK Energy Statistics. That shift is the result of subsidies for renewable generating capacity and reductions in the cost of wind and solar plants combined with other regulatory measures that have added to the costs of conventional generators. But whereas in the initial stages of decarbonising the generating mix, the relationship between regulatory cause and market impact has been relatively straightforward, making policy to encourage flexible resources is more complex: it is like a puzzle where each piece put in place changes the shapes of the others.

This is perhaps why the actions recommended by Pöyry / Imperial as having a high priority, summarised below, all sound difficult and technical, and require a large amount of collaboration.

Pöyry / Imperial recommended high priority actions for the flexibility roadmap (emphasis added)
Action Responsible Time frame
Publish a strategy for developing the longer-term roles and responsibilities of system operators (including transitional arrangements) that incentivises system operators to access all flexibility resource by making investments and operational decisions that maximise total system benefits. Ofgem in conjunction with industry 2018
Periodical review of existing system planning and operational standards for networks and generation, assessing whether they provide a level-playing field to all technologies including active network management and non-build solutions (e.g. storage and DSR), and revise these standards as appropriate. Industry codes governance and Ofgem Initial review by 2019
Review characteristics of current procurement processes (e.g. threshold capacity level to participate, contract terms / obligations) and the procurement route (e.g. open market, auctioning or competitive tendering) that enable more efficient procurement of services without unduly restricting the provision of multiple services by flexibility providers. Ofgem in conjunction with SO, TOs and DSOs By 2020
Assess the materiality of distortions to investment decisions in the current network charging methodology (e.g. lack of locational charging, double-charging for stored electricity), and reform charging methodology where appropriate. SO, DSOs and Ofgem By 2020
Assess the materiality of distortions to investment decisions in the absence of non-network system integration charging (i.e. back up capacity and ancillary services) and implement charging where appropriate SO, DSOs and Ofgem By 2020
Publish annual projections (in each year) of longer-term future procurement requirements across all flexibility services including indication of the level of uncertainty involved and where possible location specific requirements, to provide greater visibility over future demand of flexibility services SO and DSOs 2020 onwards


We looked at the current issues facing the UK energy storage sector and recent market developments in some detail in a recent post, so we will not dwell too much on the background here.

Storage – conceptually if not yet in practice – is the nearest thing there is to a “killer app” in the world of flexible resources.  It has the potential to be an important asset class on a standalone basis, but it can also be combined with other technologies (from solar to CCGT) to add value to them by enabling their output to match better the requirements of end users and the system operator.

In GB, as in a number of other jurisdictions, there is intense interest in developing distributed storage projects based on battery technology (for the moment at least, predominantly of the lithium ion variety), and a strong focus on doing so in a way that allows projects to access multiple revenue streams. There is also a general feeling that the regulatory regime needs to do more to recognise storage as a distinct activity but at the same time to do less to discriminate against it in various ways.

So, what do the Response and the Plan tell us about the vision for storage?

  • The Response points to National Grid’s SNaPS work, “which specifically considers improving transparency and reducing the complexity of ancillary services”.
  • It also points to work that has been done and/or is ongoing to clarify how storage can be co-located with subsidised renewable electricity generating projects and to provide guidance on the process of connecting storage to the grid. BEIS / Ofgem note that they see no reason why a network operator should not “promote storage…in a connection queue if it has the objective of helping others…to connect more quickly or cheaply”, and point out that Ofgem can penalise DNOs who fail to provide evidence that they are engaging with and responding to the needs of connection stakeholders.
  • BEIS / Ofgem highlight the proposals in the TCR Consultation on reducing the burden faced by storage in terms of network charges, notably the removal of demand residual charges at transmission and distribution level, and reducing BSUoS charges, for storage. A response to that consultation is to be published “in the summer”.
  • In relation to behind the meter storage, BEIS / Ofgem observe that at present: “technology costs and the limited availability of Time of Use (ToU)/smart tariffs are greater barriers…than policy or regulatory issues”. This may invite the response from some readers that it is precisely a matter for policy and regulation to promote time of use / smart tariffs: the CEPA Report makes interesting reading in this context.
  • BEIS / Ofgem “agree with the view expressed by many respondents” that network operators should be prevented from directly owning and operating storage” whilst slightly fudging the extent to which this may already be the case as a result of existing EU-based rules on the unbundling of generation from network operation, but “noting” the current EU proposals in the November 2016 Clean Energy Package to prohibit ownership of storage by network operators except in very limited circumstances and with a derogation from the Member State.
  • Flexible connections “should be made available at both transmission and distribution level”.
  • BEIS / Ofgem agree that the lack of a legal definition or regulatory categorisation of storage is a barrier to its deployment. Legislation will be introduced to “define storage as a distinct subset of generation”. This will enable Ofgem to introduce a new licence for storage before the changes to primary legislation are made. The “subset of generation” approach will also “avoid unnecessary duplication of regulation while still allowing specific regulations to be determined for storage assets” – such as whether the threshold for requiring national rather than local planning consent should be the same for storage as for other forms of generation.
  • The prospect of storage facilities benefiting, as generation, from relief from the climate change levy is also noted – although since the principal such relief (for electricity generated from renewable sources) no longer applies, this may be of limited use to most projects.

What the Response says about storage is typical of its approach to most of the issues raised in the CFE. If one wanted to be critical, it could be said that although, on the whole, BEIS / Ofgem engage with all the points raised by stakeholders, there is rarely an immediate and decisive answer to them: there is always another workstream somewhere else that has not yet concluded that holds out the prospect of something better than they can offer at present. On the other hand, perhaps that just highlights the points implicit in the Pöyry / Imperial Report’s recommendations: no one body can by itself create all the conditions for flexibility to be delivered cost-effectively, and it will be difficult fully to judge the success of the agenda that BEIS and Ofgem are pursuing for another two or three years.

But wait a minute.  On the same day as it issued the Response and the Plan, BEIS also published the CM Consultation. The sections of the Response on storage say nothing about this document, but it is potentially the most significant regulatory development in relation to storage for some time.

  • The Capacity Market is meant to be “technology neutral”. Above a 2 MW threshold, any provider of capacity (on the generation or demand side) that is not in receipt of renewable or CCC subsidies can bid for a capacity agreement in a Capacity Auction that is held one year or four years ahead of when (if successful) they may be called on to provide capacity when National Grid declares a System Stress Event.
  • A key part of the calculations of any prospective bidder in the Capacity Market, particularly one considering a new build project, who is hoping that payments under a capacity agreement will partly fund its development expenses, is the de-rating factor that National Grid applies – the amount by which each MW of each bidding unit’s nameplate capacity is discounted when comparing the amount of capacity left in the auction at the end of each round against the total amount of capacity to be procured, represented by the demand curve. Some of the de-rating factors applied in the 2016 T-4 Auction are set out below.
Technology class Description De-rating Factor
Storage Conversion of imported electricity into a form of energy which can be stored, the storing of the energy which has been so converted and the re-conversion of the stored energy into electrical energy. Includes pumped storage hydro stations. 96.29%
OCGT / recip Gas turbines running in open cycle fired mode.
Reciprocating engines not used for autogeneration.
CCGT Combined Cycle Gas Turbine plants 90.00%
DSR Demand side response 86.88%
Hydro Generating Units driven by water, other than such units: (a) driven by tidal flows, waves, ocean currents or geothermal sources; or (b) which form part of a Storage Facility. 86.16%
Nuclear Nuclear plants generating electricity 84.36%
Interconnectors IFA, Eleclink, BritNED, NEMO, Moyle, EWIC, IFA2, NSL (project specific de-rating factors for each interconnector) 26.00% to 78.00%
  • In the table above, storage has, for example, a de-rating factor approximately 10 percentage points higher than DSR and hydro and, if successful at auction, would receive correspondingly higher remuneration per MW of nameplate capacity than those technologies.
  • The typical potential storage project competitor in the Capacity Market is now more likely to be a shed full of batteries than a pumped hydro station. This has prompted industry participants to question whether such a high de-rating factor is appropriate to all storage. Ofgem, in considering changes to the Capacity Market Rules proposed by stakeholders, declined to take a view on this, deferring to BEIS.
  • BEIS, in the CM Consultation, finds merit in the arguments that (i) System Stress Events may last longer than the period for which a battery is capable of discharging power without re-charging; (ii) batteries degrade over time, so that their performance is not constant; (iii) a battery that is seeking to maximise its revenues from other sources may not be fully charged at the start of a System Stress Event. It proposes to take these points into account when setting de-rating factors for the next Capacity Auction (scheduled to take place in January 2017, and for which pre-qualification is ongoing), and splitting storage into a series of different categories based on the length of time for which they can discharge without re-charging (bands measured in half-hourly increments from 30 minutes to 4 hours). Bidders will be invited in due course to “self-select” which duration-based band they fall into.
  • Of course, deterioration in performance over time is not unique to batteries – other technologies may also perform less well by the end of the 15 year period of a new build capacity agreement than they did at the start. And, as with other technologies, such effects can be mitigated: batteries can be replaced, and who knows by what cheaper and better products by the late 2020s. However, a fundamental difficulty with the CM Consultation is that it contains an outline description of a methodology, based around the concept of Equivalent Firm Capacity, but no indicative values for the new de-rating factors.
  • It may be that BEIS’s concerns about battery performance have been heightened by the fact that the parameters for the next Capacity Market auctions show that it is seeking to procure an additional 6 GW of capacity in the T-1 auction (i.e. for delivery in 2018). There is reason to suppose that battery projects could make a strong showing in this auction, given their relatively quick construction period and the number of projects in the market, some of which may already have other “stacked” revenues (see our earlier post). Clearly it would be undesirable if a significant tranche of the T-1 auction capacity agreements was awarded to battery storage projects which then failed to perform as required in a System Stress Event.
  • It is arguable that the three potential drawbacks of battery projects are not necessarily all best dealt with by de-rating. For example, the risk that a battery is not adequately charged at the start of a System Stress Event is ultimately one for the project’s operator to manage, given that it will face penalties for non-delivery. Nor is it only battery storage projects that access multiple revenue streams and may find themselves without sufficient charge to fulfil their Capacity Market obligations on occasion: pumped hydro projects do not operate only in the Capacity Market, and even though they may be able to generate power for well over four hours, they too cannot operate indefinitely without “recharging”.  Moreover, National Grid is meant to give 4 hours’ notice of a System Stress Event, which may provide battery projects with some opportunity to prepare themselves.
  • However, the real objection to the de-rating proposal is not that it is not addressing a potentially real problem, but that it is only doing so now – given that the issue was raised by stakeholders proposing Capacity Market Rules changes at least as long ago as November 2016 – and with no published numbers for consultees to comment on.
  • The de-rating proposal illustrates a fundamental feature of the flexible resources policy space: one technology’s problems provide an up-side for competing technologies. Self-evidently, what may be bad news for batteries is good news for other storage technologies to the extent that they are not perceived to have the same drawbacks.
  • Seen in this light, the CM Consultation appears to be the main (perhaps only) example of a policy measure that supports the “larger, grid-scale” storage projects (using e.g. pumped hydro or compressed air technology) about which the Response has relatively little to say. However, a few percentage points more or less on de-rating may not make up for the lack of e.g. the “cap and floor” regulated revenue stream advocated by some for such projects.

In Part 2 of this series we will focus on the role of aggregators (featuring the analysis in the CRA Report on independent aggregators) and the demand-side more generally.

, , , , , , , , ,

On the way to a smart, flexible GB energy system? Part 1 (overview and storage)

Strong and stable, or storing up trouble? The outlook for energy storage projects in the UK

While strength and stability have taken rhetorical centre stage in the run-up to the UK’s snap General Election on 8 June, the GB energy system faces radical uncertainty on a number of fronts at a time when its stakeholders need it least. So far, the main election focus on energy has inevitably been price caps for household gas and electricity bills. But once the excitements of the campaign and polling day are over, the new government will need to make up for lost time on some less potentially vote-grabbing issues that are central to the continued health of the GB energy sector. None of these is more pressing than how to respond to the possibilities opened up by energy storage technology.

This post will summarise the benefits of energy storage as an enabler of system flexibility, look at the technology options and market factors in play and consider both some of the practical issues faced by developers and the regulatory challenges that – General Election and Brexit notwithstanding – urgently need to be addressed by the government and/or the sector regulator Ofgem.

Benefits of energy storage

The most widely cited benefit of energy storage is the ability to address the intermittency challenge of renewable sources. For more than 100 years, the general lack of bulk power storage in the GB electricity system (other than a small amount of pumped hydro capacity) did not matter. Fluctuations in demand could easily be met by adjusting the amount of power produced by centralised fossil fuel plant that generally had fairly high utilisation rates. But in a power industry transformed by the rise of wind and solar technology, things are different. As a greater proportion of the generating mix is made up of technologies that cannot be turned on and off at will, often in areas where grid capacity is limited, storage offers the possibility that large amounts of power could be consumed hours or days after it is generated, reducing the otherwise inevitable mismatch between consumers’ demands for electricity and the times when the sun is out, the wind is blowing or the waves are in motion.

In a world that increasingly wants to use low carbon sources of electricity which are inherently less easy to match to fluctuations in demand than fossil fuelled generation, storage reintroduces an important element of flexibility. More specific advantages of energy storage range across value chain.

  • For generators, power generated at times of low demand (or when system congestion makes export impossible) can be stored and sold (more) profitably when demand is high, exploiting opportunities for arbitrage in the wholesale market and potentially also earning higher revenues in balancing markets. But storage does not just help wind and solar power. It can also help plants using thermal technologies that work most efficiently operating as baseload (such as combined cycle gas turbines or nuclear plants), but which may not find it economic to sell all their power at the time it is generated. Even peaking plants can use storage to their advantage by avoiding the need to waste fuel in standby mode (using e.g. battery power to cover the period in which they start up in response to demand).
  • For transmission system operators and distribution network operators, energy storage can mitigate congestion, defer the need for investment in network reinforcement and help to maintain the system in balance and operating within its designated frequency parameters by providing a range of ancillary or balancing services such as frequency response.
  • For end users, particularly those with some capacity to generate their own power, and providers of demand-side response services who aggregate end users into “virtual power plants”, energy storage can increase household or business self-consumption rates. And in a world of tariffs differentiated by time of use (enabled by smart metering), storage opens up the possibility of retail-level arbitrage or peak shaving: buying power when it is cheaper (because not many people want it) and storing it for use it at times when it would be more expensive to get it from the grid (because everybody wants to use it).

What could all that mean in practice? Estimates in National Grid’s Future Energy Scenarios 2016 suggest that over the next 25 years, deployment of storage in the UK could grow at least as rapidly as deployment of renewables has grown over the last 20 years. Also in 2016 the Carbon Trust and Imperial College London published a study that modelled the implementation of storage and other flexible technologies across the electricity system, and showed projected savings of between £17 billion and £40 billion between now and 2050. In a consultation published in May 2017, distribution network operator Western Power Distribution (WPD) invited comment on its proposed planning assumptions for the growth of storage in GB from its current capacity of 2.7 GW (all pumped hydro plants): these are a “low growth” scenario that anticipates 4-5 GW (6-15 GWh) by 2030 and a “high growth” scenario of 10-12 GW (24-44 GWh) by that date. Growth of storage at that higher rate would see it outstripping or close to matching current government estimates for the development of new gas-fired or nuclear generation, or new interconnection capacity over the same period. (Although it should be noted that the government’s own projections for the growth of storage are more in line with WPD’s low growth scenario: see this helpful analysis by Carbon Brief.)

Technology options

As is the case in Europe and the rest of the world, energy storage in the UK is currently mostly supplied by pumped hydropower plants, which account for almost all storage capacity and are connected to the transmission system. Until very recently, the much less frequently deployed technique of compressed air energy storage (CAES) was the only other commercially available technology for large-scale electricity storage. The two technologies are similar in that both use cheap electricity to put a readily available fluid (water or air) into a state (up a mountain or under pressure) from which it can be released so as to flow through a turbine and generate power. They differ in that pumped hydro requires a specific mountainous topography, whereas CAES can use a variety of geologies (including salt caverns, depleted oil and gas fields and underground aquifers).

But it is batteries that are currently attracting the keenest investor interest in storage. There are many different battery technologies competing for investment and market penetration. Those based on sodium nickel chloride or sodium sulphur have made advances, but most storage attention surrounds batteries based on lithium-ion structures, also the battery of choice for the electric car industry, where competition has driven down costs. Just before the General Election got under way, the Department of Business, Energy and Industrial Strategy (BEIS) announced £246 million of funding for the development and manufacture of batteries for electric vehicles. Electric car batteries need to be able to deliver a surge of power far more rapidly than those deployed in the wider power sector: in Germany, car manufacturers are already exploring the use of electric car batteries that no longer up to automotive specifications in grid-based applications. In the North East of England, distribution network company Northern Powergrid is collaborating with Nissan to look at how integration of electric vehicles can improve network capacity, rather than just placing increased demands on the grid.

The cost of batteries has come down because of improvements in both battery chemistry and manufacturing processes, as well as the economies of scale associated with higher manufacturing volumes such as with Tesla and Panasonic’s new battery Gigafactory in Nevada. Underlining rising global expectations about low cost and set-up time for battery production, in March 2017 Tesla’s Elon Musk offered to build a 100 MWh battery plant in Australia within 100 days, or to give the system away for free if delivery took any longer.

Batteries are ideally suited to many applications, but they also have some drawbacks. They are less good at providing sustained levels of power over long periods of discharge, and on a really large scale, than CAES or pumped hydro. The non-battery technologies also have other selling points. For example, CAES also has a unique ability, when combined with a combined cycle gas turbine, to reduce the amount of fuel it uses by at least a third. Given the likelihood that the UK power system will continue to need a significant amount of new large-scale gas fired plant, even as it decarbonises, and given the current slow development of carbon capture and storage technology, the potential reduction in both the costs and the carbon footprint of new gas-fired power that CAES offers is well worth consideration by both developers and government. Finally, as regards future alternative technology options, hydrogen storage and fuel cells are the subject of significant research efforts and funding. Most enticing from a decarbonisation perspective, is the prospect of electrolysing water with electricity generated from renewables to produce “green hydrogen”, which can then be used to generate clean power with the same level of flexibility as methane is at present.

Models and market factors

In the abstract, it might be thought that energy storage projects could be categorised into five basic business models:

  • integrated generator services: storage as a dedicated means of time-shifting the export of power generated from specific generating plants (renewable, nuclear or conventional), with which the storage facility may or may not be co-located, and so optimising the marketing of their power (and in some cases, where there are grid constraints, enabling more power to be generated, and ultimately exported, than would otherwise be the case);
  • system operator services: providing frequency response and other ancillary or balancing services to National Grid in its role as System Operator (and potentially, in the future, to a distribution system operator that is required to maintain balance at distribution level): a distinction can be made between “reserve” and “response” services, the latter involving very quick reaction to instructions designed to ensure frequency or voltage control;
  • network investment: enabling distribution networks to operate more efficiently and economically, for example by avoiding the need for conventional network reinforcement. This was notably successfully demonstrated by the 6 MW battery at Leighton Buzzard built by UK Power Networks (UKPN). The results of WPD’s Project FALCON were a little more equivocal, but it is trying again, using Tesla batteries to test a range of applications at sites in the South West, South Wales and the East Midlands);
  • merchant model: a standalone storage facility making the most of opportunities to buy power at low prices and sell it at high prices, with no tie to particular generators, and perhaps underpinned by Capacity Market payments (see further below);
  • “behind the meter”: enabling consumers to reduce their energy costs (retail level arbitrage or peak shaving, as noted above, as well as maximising use of on-site generation where this is cheaper than electricity from the grid).

These models are far from being mutually exclusive. Indeed, at present, they are best thought of as simply representing different categories of potential revenue streams: the majority of storage projects will need to access more than one of these streams in order to be viable. Some will opt to do so through contracts with an aggregator, for whom a relationship with generation or consumption sites with storage, particularly if they have a degree of operational control over the storage facility, offers an additional dimension of flexibility.

In the short term, the largest revenue opportunity may be the provision of grid services. The need for a fast response to control frequency variations is likely to increase in the future as a result of the loss of coal-fired plant from the system.

Growing interest in energy storage also owes much to the decline in the UK greenfield renewables market, with the push factor of the removal or drastic reduction of subsidies previously available for new renewable energy projects and the pull factor of the battery revolution. According to a report published in May 2017 by SmartestEnergy, an average of 275 solar, wind and other renewable projects were completed in each quarter between 2013 and the last quarter of 2016, when the figure plummeted to 38. Only 21 renewable projects were completed in the first quarter of 2017.

So why, when UKPN, for example, report that between September 2015 and December 2016 they processed connection applications from 600 prospective storage providers for 12 GW of capacity, is the amount of battery capacity so far connected only in the tens of MW?

Tenders and auctions

It may help to begin by looking at another very specific factor that drove this extraordinary level of interest in a technology that had been so little deployed to date. This was National Grid’s first Enhanced Frequency Response (EFR) tender, which took place in August 2016. A survey by SmartestEnergy, carried out just before the results of the tender were announced, found that 70 percent of respondents intending to develop battery projects in the near future were anticipating that ancillary services would be their main source of revenue.

National Grid were aiming to procure 200 MW of very fast response services. Although “technology neutral”, the tender was presented as an opportunity for battery storage providers and as expected, storage, and specifically batteries, dominated. All but three of the 64 assets underlying the 223 bids from 37 providers were battery units. Perhaps less expected were the prices of the winning bids: some as low as £7/MWh and averaging £9.44/MWh. The weighted price of all bids was £20.20/MWh.

This highly competitive tender gave the UK energy storage market a £65 million boost. The pattern of bids suggested that alongside renewables developers and aggregators, some existing utilities are keen to establish themselves in the storage market, and are prepared to leverage their lower cost of capital and accept a low price in order to establish a first mover advantage.

Independent developers who regard storage as a key future market might also have been bullish in their calculations of long-term income while accepting lower revenues in the near term to compete in a crowded arena. For all bidders, one of the key attractions was the EFR contract’s four-year term, which makes a better fit with their expectations of how long it will take to recoup their initial investment than the shorter duration of most of National Grid’s other contracts for balancing / ancillary services.

Aspiring battery storage providers also responded enthusiastically to the regular four year ahead (T-4) Capacity Market (CM) auction when it took place for the third time in December 2016. To judge from the Register for the T-4 2016 auction, some 120 battery projects, with over 2 GW of capacity between them, were put forward for prequalification in this auction. (This assumes that all the new build capacity market units (CMUs) described as made up of “storage units” and not obviously forming part of pumped hydro facilities were battery-based.) Although almost two-thirds of these proposed CMUs are described on the relevant CM register as either “not prequalified” or “rejected”, of the remaining 33 battery projects, no fewer than 31 projects, representing over 500 MW of capacity between them, went on to win capacity agreements in the auction.

There are a number of points to be made in connection with these results.

  • Taking the CM and EFR together, the range of parties interested in batteries is noteworthy, as is the diversity of motivations they may have for their interest.  It includes grid system operators (UKPN), utilities (EDF Energy, Engie, E.ON, Centrica), renewables developers (RES, Element Power, Push Energy, Belectric), storage operators, aggregators / demand side response providers (KiWi Power, Limejump, Open Energi) and end-users, as well as new players who seem to be particularly focused on storage (Camborne Energy Storage, Statera Energy, Grid Battery Storage).
  • Developers of battery projects are evidently confident that the periods during which they may be called on to meet their obligations to provide capacity by National Grid will not exceed the length of time during which they can continuously discharge their batteries – in other words, that the technical parameters of their equipment do not put them at an unacceptable risk of incurring penalties for non-delivery under the CM Rules: a point that some had questioned.
  • The CM Rules are stricter than those of the EFR tender as regards requiring projects to have planning permission, grid connection and land rights in place as a condition of participating in the auction process. This is presumably one reason why fewer battery projects ended up qualifying to compete in the T-4 auction as compared with the EFR tender.
  • For batteries linked to renewable electricity generation schemes that benefit from renewables subsidy schemes such as the Renewables Obligation (RO), the EFR tender was an option, but the CM was not, since CM Rules prohibit the doubling up of CM and renewables support. So, for example, the 22 MW of batteries to be installed at Vattenfall’s 221 MW RO-accredited Pen-y-Cymoedd wind farm was successful in the EFR tender but would presumably not have been eligible to compete in the CM.
  • Accordingly, CM projects tend to be designed to operate quite independently of any renewable generating capacity with which they happen to share a grid connection. But some of these projects are located on farms that might have hosted large solar arrays when subsidies were readily available for them. Green Hedge, four of whose projects were successful in the T-4 2016 CM auction, has even developed a battery-based storage package called The Energy BarnTM. Others CM storage projects are located on the kind of industrial site that might otherwise be hosting a small gas-fired peaking plant. UK Power Reserve (as UK Energy Reserve), which has been very successful with such plants in all the T-4 auctions to date, won CM support for batteries at 12 such locations.
  • The Capacity Market may be less lucrative than EFR, measured on a per MW basis, but it offers the prospect of even longer contracts: up to 15 years for new build projects.
  • Batteries are still a fairly new technology. The clearing price of Capacity Market auctions has so far been set by small-scale gas- or diesel-fired generating units using well established technology. In a T-4 auction, the CMUs, by definition, do not have to be delivering capacity until four years later – although the Capacity Market Rules oblige successful bidders to enter into contracts for their equipment, and reach financial close, within 16 months of the auction results being announced. Other things being equal (which they may not be: see next bullet), it will clearly be advantageous to developers if they can arrange that the prices they pay for their batteries are closer to those prevailing in 2020 than in 2016. It has been pointed out that although internationally, battery prices may have fallen by up to 24 percent in 2016, the depreciation of Sterling over the same period means that the full benefit of these cost reductions may not yet be accessible to UK developers.
  • The proportion of prequalified battery-based CMUs that were successful in the T-4 2016 CM auction was remarkably high. But may not have been basing their financial models solely or even primarily on CM revenues. In addition to EFR and other National Grid ancillary services, such as Short Term Operating Reserve or Fast Reserve, and possible arbitrage revenues, it is likely that at least some projects were anticipating earning money by exporting power onto the distribution network during “Triad” periods. This “embedded benefit” would enable them to earn or share in the payments under the transmission charging regime that have been the main source of revenue for small-scale distributed generators bidding in the CM, enabling them to set the auction clearing price at a low level and prompting a re-evaluation of this aspect of transmission charges by Ofgem. From Ofgem’s March 2017 consultation on the subject, it looks as if these payments will be drastically scaled down over the period 2018 to 2020. This may give some developers a powerful incentive to deploy their batteries early (notwithstanding the potential cost savings of waiting until 2020 to do so) so as to benefit from this source of revenue while it lasts. Those who compete in subsequent CM auctions may find that the removal of this additional revenue leads to the CM auctions clearing at a higher price.
  • As with EFR, some developers may be out to buy first mover advantage, and most already have a portfolio of other assets and/or sources of revenue outside the CM. But what they are doing is not without risk, since the penalties for not delivering a CMU (£10,000, £15,000 or £35,000 / MW, depending on the circumstances) are substantial.
  • Meanwhile, a sure sign of the potential for batteries to disrupt the status quo can be seen in the fact that Scottish Power has proposed a change to the CM Rules that would apply a lower de-rating factor to batteries for CM purposes than to its own pumped hydro plant.

Finally, one other tender process, that took place for the first time in 2016, could point the way to another income stream for future projects. National Grid and distribution network operator Western Power Distribution co-operated to procure a new ancillary service of Demand Turn Up (DTU).

The idea is to increase demand for power, or reduce generation, at times when there is excess generation – typically overnight (in relation to wind) and on Summer weekends (in relation to solar). DTU is one of the services National Grid use to ensure that at such times there is sufficient “footroom” or “negative reserve”, defined as the “continuous requirement to have resources available on the system which can reduce their power output or increase their demand from the grid at short notice”.

National Grid reports that over the summer of 2016, the service was used 323 times, with “10,800 MWh called with an average utilisation price of £61.41/MWh”. The procurement process can take account of factors other than the utilisation and availability fees bid, notably location. Successful tenders in the 2017 procurement had utilisation fees as high as £75/MWh.

At present, the procurement process for DTU does not appear to allow for new storage projects to compete in DTU tenders, but once they have become established, they should be well placed to do so, given their ability to provide demand as well as generation. They could be paid by National Grid to soak up cheap renewable power when there is little other demand for it. If National Grid felt able to procure DTU or similar services further in advance of when they were to be delivered, the tenders could have the potential to provide a more direct stimulus to new storage projects.

Battery bonanza?

Those who have been successful in the EFR or CM processes can begin to “stack” revenues from a number of income streams. And the more revenues you already have, the more aggressively you can bid in future tenders (for example for other ancillary services) to supplement them.

But even if all the projects that were successful in the EFR and CM processes go ahead, they will still represent only a small fraction of those that have been given connection offers. Moreover, it looks as if the merchant and ancillary services models are the only ones making significant headway at present.  Why are we not seeing more storage projects integrated with renewables coming forward, for example? Why, to quote Tim Barrs, head of energy storage sales for British Gas, has battery storage “yet to achieve the widespread ‘bankable status’ that we saw with large-scale solar PV”?

Technology tends to become bankable when it has been deployed more often than batteries coupled with renewables have so far in GB. But even to make a business case to an equity investor, a renewables project with storage needs to show that over a reasonable timeframe the additional revenues that the storage enables the project to capture exceed the additional costs of installing the storage. What are these costs, over and above the costs of the batteries and associated equipment?  What does it take to add storage to an existing renewable generating project, or one for which development rights have already been acquired and other contractual arrangements entered into?

  • The configuration and behaviour of any storage facility co-located with subsidised renewable generation must not put the generator’s accreditation for renewable subsidies at risk because of e.g. a battery’s ability to absorb and re-export power from the grid that has not been generated by its associated renewable generating station. The location of meters is crucial here. According to the Solar Trade Association, only recently has Ofgem for the first time re-accredited a project under the RO after storage was added to it. While an application for re-accreditation is being considered, the issue of ROCs is suspended. Guidance has been promised which may facilitate re-accreditation for other sites. Presumably in this as in other matters, the approach for Feed-in Tariff (FIT) sites would follow the pattern set by the RO. For projects with existing Contracts for Difference (CfDs), there is no provision on energy storage. For those hoping to win a CfD in the 2017 allocation round, the government has made some changes to the contractual provisions following a consultation, but, as the government response to consultation makes clear, a number of issues still remain to be resolved.
  • An existing renewables project is also likely to have to obtain additional planning permission. There may be resistance to battery projects in some quarters. RES recently had to go to appeal to get permission for a 20 MW storage facility by an existing substation at Lookabootye after its application was refused by West Lothian Council. It will also be necessary to re-negotiate existing lease arrangements (or at least the rent payable under them), and additional cable easements may be required.
  • Unless it is proposed that the battery will take all its power from the renewable generating station (which is unlikely), it will be necessary to seek an increase in the import capacity of the project’s grid connection from the distribution network operators. Even if the developer does not require to be able to export any more power at any one time from the development as a whole, in order to charge the battery at a reasonable speed from the grid it will need a much larger import capacity than is normal for an ordinary renewable generating facility. The ease and costs of achieving this will vary depending on the position of the project relative to the transmission network. There may be grid reinforcement costs to pay for: UKPN has noted that there are few places on the network with the capacity to connect a typical storage unit without some reinforcement. They will also treat the addition of storage as a material change to an existing connection request for a project that has not yet been built, prompting the need for redesign and resulting in the project losing its place in the queue of connection applications.
  • A power purchase agreement (PPA) for a project with storage will need to address metering. For the purposes of the offtaker, output will either need to be measured on the grid side of the storage facility (the same may not be true of metering for renewable subsidy purposes), or an agreed factor will need to be applied to reflect power lost in the storage process. Secondly, in order to maximise the opportunities for arbitrage by time-shifting the export of its power, a project with storage may want more exposure to fluctuations in the wholesale market price, and even to imbalance price risk, than a traditional intermittent renewables project. The detail of how embedded benefits revenues are to be shared between generator and offtaker may also require to be adjusted if the addition of storage makes it more likely they will be captured.

For the moment, most renewables projects probably fall into one of two categories with regard to integrated storage.

  • On the one hand, there are those that are already established and receiving renewable generation subsidies, or which have been planned without storage and now simply need to commission as quickly as possible in order to secure a subsidy (for example, under RO grace period rules for onshore wind projects). For them, introducing storage into an existing project may be more trouble than it is worth for some or all of the reasons noted above. They have little incentive to deploy storage unless it is an economic way of reducing their exposure to loss of revenue as a result of grid constraints or to imbalance costs: these have been increasing following the reforms introduced by Ofgem in 2015 and will increase further as the second stage of those reforms is implemented in 2018, but for many renewable generators are a risk that is assumed by their offtakers.
  • On the other hand, for projects with no prospect of receiving renewable subsidies, it would appear that the cost of storage is not yet low enough, or the pattern of wholesale market prices sufficiently favourable to a business model built on  time-shifting and arbitrage to encourage extensive development of renewables + storage merchant model projects. If it was generally possible easily to earn back the costs of installing storage through the higher wholesale market revenues captured by – for example – time-shifting the export of power from a solar farm to periods when wholesale prices are higher than they are during peak solar generating hours, the volume and profile of successful storage + renewable projects in the CM and elsewhere would be different from what it now is.

However, battery costs will continue to fall, and wholesale prices are becoming “spikier”. It may only be a matter of time before GB’s utility-scale renewables sector, whose successful players have so far built their businesses on the predictable streams produced by RO and FIT subsidies, can get comfortable with business cases that depend more fundamentally on the accuracy of predictions about how the market, rather than the weather, will behave. Moreover, there is nothing to stop a storage facility co-located with a renewables project that has no renewable subsidy from earning a steady additional stream of income in the form of CM payments.

Arguably, the UK has missed a trick in not having adopted pump-priming incentives for combining storage with renewables, such as setting aside a part of the CfD budget for projects with integrated storage. But with the door apparently generally closed for the time being on any form of subsidy for large-scale onshore wind or solar schemes in most of GB, it is probably unrealistic to hope for any such approach to be taken in the near future.

Regulatory challenges

There are undoubtedly already significant commercial opportunities for some GB storage projects, but it does not feel as if the full power of storage to revolutionise the electricity market is about to be unleashed quite yet. This is perhaps not surprising.

Almost as eagerly awaited among those interested in storage as the results of the EFR tender was a long-promised BEIS / Ofgem Call for Evidence on how to enable a “smart, flexible energy system”, which was eventually published in November 2016. This Call for Evidence, the first of its kind, represented a significant step forward for the regulation of storage in the UK, but although it pays particular attention to storage and the barriers that storage operators may face it is not just “about” storage. It ultimately opens up questions about how well the current regulatory architecture, designed for a world of centralised and despatchable / baseload power generation, can serve an increasingly “decarbonised, distributed, digital” power sector without major reform. (At an EU level, the European Commission’s Clean Energy Package of November 2016 tries to answer some of these questions, and there is generally no shortage of thoughtful suggestions for reforming power markets, such as the recent Power 2.0 paper from UK think tank Policy Exchange, or the “Six Design Principles for the Power Markets of the Future” published by Michael Liebreich of Bloomberg New Energy Finance.)

However, whilst it is important to take a “whole system” approach, it would be unfortunate if the breadth of the issues raised by the Call for Evidence were to mean that there was any unnecessary delay in addressing the regulatory issues of most immediate concern to storage operators. Government and regulators have to start somewhere, and it is not unreasonable to start by trying to facilitate the deployment of storage since it could facilitate so many other potentially positive developments in the industry.

On 25 April Ofgem revealed that it had received 240 responses to the Call for Evidence, with around 150 responses commenting on energy storage. Barriers to the development of storage identified by respondents include the need for a definition of energy storage, clarity on the regulatory treatment of storage, and options for licensing. The response from the Energy Storage Network (ESN) offers a good insight into many of the issues of most direct concern to storage operators. Some of the other respondents who commented on storage also demonstrated an appetite for fundamental reform of network charging (described by one as “probably not fit for purpose in its current form”) and for significant shifts in the role of distribution network operators.

Interest in a definition of energy storage is unsurprising. It is arguably hard to make any regulatory provision about something if you have not defined it. But at the same time, the Institution of Engineering and Technology may well be correct when it says in its response to the Call for Evidence: “lack of a definition is not a barrier in itself…as the measures are developed to address the barriers to storage, it will become clear whether a formal definition is required and at what level…agreeing a definition should be an output of regulatory reform, not an input.”. In other words, how you define something for regulatory purposes – particularly if that thing can take a number of different forms and operate in a number of different ways – will depend in part on what rules you want to make about it.

Under current rules, energy storage facilities end up being classified, somewhat by default, as a generation activity – even though their characteristic activity does not add to the total amount of power on the system. But because storage units also draw power from the grid, they find themselves having to pay two sets of network charges – on both the import and the export – even though they are only “warehousing” the power rather than using it. Both these features of the current regulatory framework are strongly argued against by a variety of respondents to the Call for Evidence.

Treating storage as generation complicates the position for distribution network operators wishing to own storage assets. Under the current unbundling rules (which are EU-law based, but fully reflect GB policy as well), generation and network activities must be kept in separate corporate compartments. These rules are designed to prevent network operators from favouring their own sources of generation (or retail activities). The issue is potentially more acute when you have a storage asset forming part of the network company’s infrastructure and regulated asset base, but having the ability to trade on the wholesale power and ancillary services markets in its own right as well as to affect the position of other network users (by mitigating or aggravating constraints). UKPN considers that the approach it has adopted with its large battery project could provide a way around this problem for others as well – essentially distinguishing the entity that owns the asset from the entity responsible for its trading activity on the market. However, such an arrangement is not without costs and complexity, both for those involved to set up and for the regulator to monitor. The ESN has also made proposals in its response to the Call for Evidence about the conditions under which distribution network operators should be permitted to operate storage facilities.

It may be that the most useful contribution that transmission and distribution network operators could make to the development of storage would be to determine as part of their multi-year rolling network planning processes where it would be most beneficial in system terms for new storage capacity of one kind or another to be located. But the underlying question is whether at least some storage projects should be treated more as network schemes with fixed OFTO or CATO-like rates of return rather than being regarded as part of the competitive sector of the market along with generation and supply. (Similar concerns about the status of US network-based storage projects, admittedly in a slightly different regulatory environment, have been addressed by the Federal Energy Regulatory Commission in a recent policy statement and notice of proposed rulemaking.)

If storage is not to be treated as generation or necessarily part of a network (and required to hold a generation licence where no relevant exemption applies), what is it? Should it be recognised as a new kind of function within the electricity market? In which case, the natural approach under the GB regulatory regime would be to require storage operators to be licensed as such (again, subject to any statutory exemptions). That would require primary legislation (i.e. an Act of Parliament) to achieve, at a time when Parliamentary time may be at a premium because of Brexit – and then there would need to be drafting of and consultation on licence conditions and no doubt also numerous consequential changes to the various industry-wide codes and agreements.

The ESN’s Call for Evidence response has some helpful suggestions as to what a licensing regime for storage might look like. But is the licensing model is a red herring in this context? After all, the parallel GB regulatory regime for downstream gas includes no requirement for those wishing to operate an onshore gas storage facility to hold a licence to do so under the Gas Act 1986. And it is entirely possible to trade electricity on the GB wholesale markets (a key activity for storage facilities), without holding a licence under the Electricity Act 1989 (or even engaging in an activity requiring such a licence but benefiting from an exemption from the requirement to hold a licence).

As for some of the current financial disadvantages facing storage, it is encouraging that in consulting on its Targeted Charging Review of various aspects of network charging in March 2017, Ofgem provisionally announced its view that some double charging of storage should be ended. It consulted on a number of changes that, taken together, should have the effect of ensuring that “storage is not an undue disadvantage relative to others providing the same or similar services”. However, although welcome, these Ofgem proposals so far only cover the treatment of the “residual” (larger) element of transmission network charges for demand (applicable to distribution-connected projects), in respect of storage units co-located with generation. It remains to be seen whether – and if so, what – action will be taken to deal with other problems in this area, such the payment of the “final consumption” levies that recover the costs of e.g. the RO and FIT schemes by both the storage provider and the consumer on the same electricity when a storage operator buys that electricity from a licensed supplier. Storage operators can at present only avoid this cost disadvantage if they acquire a generation licence, which does not seem a particularly rational basis for discriminating between them in this context.

Speaking in March, the head of smart energy policy at BEIS, Beth Chaudhary, said that ending the double counting of storage “might require primary legislation”, adding that Brexit has made the progress of such legislation “difficult at the moment”. The General Election has only added to concerns of momentum loss, a sense of “circling the landing strip” in the words of the Renewable Energy Association’s chief executive, Dr Nina Skorupska.

“The revolution will not be televised”…but it probably needs to be regulated

What is the storage revolution? Storage will not turn the electricity industry into a normal commodity market, like oil, overnight – or indeed ever. We will still have to balance the grid. As before, what is being exported onto the grid will need to match what is being imported from it at any given moment. It’s just that storage will provide an additional source of power to be exported onto the grid (which was generated at an earlier time) and it will also facilitate more balancing actions by those on the demand side where they have access to it. It is also likely that increased use of micro grids, with the ability to operate in “island mode” as well as interconnected with the public grid, will result in the public grid handling a smaller proportion of the power being generated and consumed at any given time.

Of course, one could look at this and say: “Fine, but what’s the hurry?”. The UK developed a renewables industry when it was still a relatively new and expensive thing to do. Thanks to the efforts made by the UK and others, renewables are now both “mainstream” and relatively cheap. Those countries that are only starting to develop sizeable renewable projects now are reaping the benefit of the cost reductions achieved by the early adopters. Would it be such a bad thing if a GB storage revolution was delayed for a year or two while other markets experiment with the technology and help it to scale up, reducing the costs that UK businesses and consumers will pay for its ultimate adoption in the UK?

After all, we have to be realistic about the number of large and difficult issues the UK government and regulators can be expected to focus on and take forward at once. Is it not more important, for example, to reach agreement with the rest of the EU on a satisfactory set of substitute arrangements for the legal mechanisms that currently govern the UK’s trade in electricity and gas with Continental Europe (and the Republic of Ireland)? In addition, the General Election manifestos of each party prioritise other contentious areas of energy policy for action, such as facilitating fracking and reducing the level of household energy bills.

We do not deny the importance of these other issues, and BEIS and Ofgem resources are, of course, finite, but we would argue that storage and the complex of “flexibility” issues to which it is central should be high on the policy agenda after 8 June in any event.

  • GB distribution network operators have already done lot of valuable work on storage, much of it funded by various Ofgem initiatives (notably the Innovation Funding Incentive, Network Innovation Allowance and Low Carbon Networks funding). This has generated a body of published learning on the subject which continues to be added to and which it would be a pity not to capitalise on as quickly as possible.
  • Depending (at least in part) on the outcome of Brexit, we may find ourselves either benefiting from significantly more interconnection with Continental European power markets, or becoming more of a “power island” compared with the rest of Europe. In either case, a strong storage sector will be an advantage. Storage can magnify the benefits of interconnection but it would also help us to optimise the use of our own generating resources if our ability to supplement them (or export their output) through physical links to other markets was limited.
  • The UK has in some respects led the world on power market reform.  We have complex, competitive markets and clever companies that have learnt how to operate in them. Looking at storage from an industrial strategy point of view, the UK is may not make its fortune after by the mass manufacture of batteries for the rest of the world, but the potential for export earnings from some of the higher value components of storage facilities, and the expertise to deploy them to maximum effect, should not be neglected.
  • On the other hand, if the UK wants to maintain its position as an attractive destination for investment in electricity projects, it needs to show that it has a coherent regulatory approach to storage, both because storage will increasingly become an asset class in its own right and because sophisticated investors in UK generation, networks or demand side assets will increasingly want to know that this is the case before committing to finance them.
  • As the Call for Evidence and the other attempts to address the challenges of future power markets referred to above make clear, everything is connected. There is, arguably, not very far that you can or should move forward on any aspect of generation or other electricity sector policy without forming a view on storage and how to facilitate it further.
  • Finally, because some of the policy and regulatory issues are hard and resources to address them are finite, this will all take time, so that with luck, the regulatory framework will have been optimised by about the same time as the price reductions stimulated by demand from the US and other forward-thinking jurisdictions have started to kick in.

Almost whatever problem you are looking at, whether as a regulator or a commercial operator in the GB power sector, it is worth considering carefully whether and how storage could help to solve it. Storage has the potential, as noted above, to change the ways that those at each level in the electricity value chain operate, and with the shift to more renewables and decentralised generation, it has a significant part to play in making future electricity markets “strong and stable”. The “trouble” alluded to in the title of this post is change either happening faster than politicians and regulators can keep pace with, or innovation being stifled by the lack of regulatory adaptation as they find it too difficult to address the challenges it poses when faced with other and apparently more urgent priorities. Because the ways in which generators, transmission and distribution network operators, retailers and end users interact with each other is so much a function of existing regulation of one kind or another, it is very hard to imagine storage reaching its full potential without significant regulatory change. These changes will take time to get right, but since ultimately an electricity sector that makes full use of the potential of storage should be cheaper, more secure and more environmentally sustainable than one that does not, there should be no delay in identifying and pursuing them.



, , , , , , , , , , , ,

Strong and stable, or storing up trouble? The outlook for energy storage projects in the UK