Infra investors’ cautious love affair with UK batteries
Paul Mason, a director of Harmony Energy Advisors, a UK-focused battery storage investor, recalls back in 2016 trying to attract institutional investors to an unlisted fund through which he could invest in batteries. “They told us to come back next year,” he recalls.
Even as recently as three years ago he saw little investor appetite for batteries owned by developers. “You virtually couldn’t give the things away,” Mason recalls.
In the end he decided to list in November 2021 an investment company, Harmony Energy Income Trust, on the London Stock Exchange. It has preferential rights to acquire projects from a privately-owned battery developer, Harmony Energy, which was founded back in 2010.
Today, Mason says, investors are willing to almost “throw money at the sector”, attracted in part by the sector’s fundamental allure – batteries’ capacity for buying electricity cheap when demand is lowest and selling when it is most in demand and prices are highest.
Driven by the recent spike in energy prices caused by the Ukraine war, the hot summer and outages in French nuclear power stations, many report eye-watering returns.
Alan Smallwood, a director at Anesco, a Reading, UK-based company that manages batteries on behalf of their owners, recalls buying electricity on 23 January 2022 at GBP 145 per MWh.
Later that day he signed a contract to sell this electricity on the energy online auction, EPEX, at GBP 1,000/MWh between 5pm-6pm on the following day, January 24 – one of his most profitable days last year.
Smallwood’s spread – the difference between the cost of buying and selling electricity – was GBP 835 per MWh. Other battery operators report similar stories.
In addition to such trading spreads, the speculation has also been fuelled by dynamic containment (DC) – the latest version of frequency balancing in which battery projects are called upon by the grid operator, National Grid ESO, to maintain the grid at around 50 hertz.
Smallwood at Anesco, which also operates as an EPC and O&M contractor, reports agreeing a DC “low” contract (when the frequency falls below 50 hertz) last September with National Grid ESO, the grid’s operator, for a four-hour slot for one of his batteries at an impressive GBP 80 per MW/h. National Grid in 2020 when DC had just started was offering a fixed price of just GBP 17 an hour.
In recent days, Smallwood won a dynamic containment high contract (when the frequency goes above 50 hertz) at GBP 60.15 per MW/h for a part of the day, the highest price he has ever been seen in this market.
Even better, DC auctions are so-called “pay-as-clear” auctions, meaning the 10 other batteries managed by Anesco also benefitted from the same clearing price.
Four to five years ago battery owners earned up to between GBP 80,000 and GBP 120,000 per MW annually, says one source who works as an “optimiser”, which specialise in trading electricity on behalf of owners of battery storage projects.
This was also driven by DC’s predecessors, enhanced frequency response (EFR) and fast frequency response (FFR) as well as wholesale trading of power. But as well as being more cumbersome than DC with batteries being bid a month in advance, EFR and FFR were less remunerative with contracts of between GBP 8-12 per MW/h.
Following the introduction of DC in 2020 – an event that many say triggered a flood of capital into the sector – and soaring energy prices, earnings have shot up to around GBP 150,000 per MW a year, says the trader.
At times battery revenues have soared well above this. Last July, amidst spiking demand driven by warm weather, batteries’ earnings rose so high that if every month was the same then total earnings for the year would total GBP 270,000 per MW, according to Robyn Lucas, a director at UK energy data company Modo Energy. But in March this year average earnings dropped to as low as GBP 53,000 per MW.
This GBP 270,000 per MW figure translates into an IRR well in excess of 20%; at such rates some investors say they can recover the full cost of buying or developing their batteries within two to three years.
Infra funds pile into the sector
In response to this very merchant opportunity, over the past two years the largest infrastructure managers including Macquarie, BlackRock, Copenhagen Infrastructure Partners (CIP), Masdar and Brookfield, as well as pension funds such as Alberta Investment Management Corporation and Railpen have begun to make their mark in the sector.
Most of the managers have bought battery developers with ambitious portfolios of development-stage assets, largely as they have high return targets so developing, building and then selling assets makes better sense than buying more expensive operational or shovel-ready portfolios. CIP, for example, partnered a year ago with London-based Alcemi, which at the time had a 4 GW pipeline of batteries and promised to complete construction of its first one around about now.
An executive at one large infrastructure executive says he has not bought a stake in a developer, rather he has formed a partnership and bought a majority stake in its projects. So far he has not raised debt – “our preference will be to fund through to construction and not have a bank looking over my shoulder” – although, he says, he might do so when the projects reach commissioning stage. Other infrastructure funds have gone down a similar path.
The CEO of one battery developer recalls receiving bids from no less than 42 investors for a stake in his company he was recently selling. As well as being from a broad class of investors including pension funds, infrastructure funds, private equity funds and corporates, the bidders also came from most corners of the world including North America, the Middle East and Asia as well as the UK and Europe. “It was the broadest thing I’ve ever run,” says the developer, a former investment banker.
Scale is also an issue, says Martin Krastev, head of M&A at London-based corporate advisory firm IDCM. "Many of our clients are interested in investing in batteries and are asking us for suitable opportunities as the deals to date have been either too small or too early stage developments for the infrastructure investment community," says Krastev. "Some of our clients are more interested in contracted assets, while other are more interested in capturing the upside of merchant batteries," he adds.
Developers are becoming more ambitious as demand for batteries grows. Carlyle-backed Canadian battery developer, Amp Energy, is not just selling a 2 GW utility-scale battery pipeline as well as three two-hour batteries in Scotland that are due to be commissioned next year, it is also selling the business itself with its staff and management.
Meanwhile, there is a class of lower-return investors that to date has largely bought shovel ready and built batteries. These includes the UK’s largest battery owner, the London-listed Gresham House Energy Storage Fund, as well as other listed funds such as Gore Street Energy Fund and SMS PLC.
But some such more conservative investors are now also becoming more ambitious, partly due to the rising costs of batteries driven by the surging demand and price of materials such as cobalt and lithium.
Foresight has four funds with mandates to invest in batteries, all of which have typically bought batteries at ready-to-build stage. But Ross Driver, a managing director at Foresight, says one of the funds, the London-listed Foresight Solar Fund, recently changed its mandate to allow it to invest in development-stage batteries. “We are having discussions with developers to get into development stage assets and get into them at earlier stage,” says Driver. He is also keen to form a joint venture or invest in a battery developer.
SUSI Partners has been investing in post-development stage batteries through a dedicated energy storage fund since 2016 in Canada, the US, Australia as well as the UK. But today the firm via its open-ended SUSI Energy Transition Fund is primarily looking to partner with battery developers with which it can develop and build projects, says Richard Braakenburg, its Head of Equity Investments.
Even conservative pension funds are willing to overlook the sector’s merchant risk – and one that its investors say with pride is the only unsubidised form of power that is active in the renewable energy market.
As well as GLIL which in 2021 bought London-based battery developer Flexion Energy, Railpen along with Alberta Investment Management Corporation last July agreed to pay some GBP 400m for most shares in UK battery platform Constantine Energy Storage.
A spokesperson for the pension fund for UK railways workers acknowledges battery storage assets are “more exposed to merchant risk than traditional core infrastructure assets”.
But, the spokesperson adds, “the flexibility of batteries allows for revenue maximisation if controlled appropriately”, a reference to batteries being used for ancillary services such as dynamic containment as well as trading. Railpen has also “negotiate[d] terms with counterparties which partially hedge merchant risk.
Finding the next source of revenues
Railpen can also take some comfort from the growing presence of battery companies in the capacity market auctions, the only long-term contracted element of battery investing.
Some 600 MW of battery projects won capacity in National Grid ESO’s T1 auction in February for batteries to start operating in a year’s time, double the amount in the previous auction.
Also, some 5 GW of batteries secured contacts in the T4 auction in March for projects coming online in four years’ time. This still represented just 3% of all awarded capacity with gas, interconnectors and pumped storage all taking sizeably larger shares. Batteries are expected to play a larger role in the capacity auction as other, less green technologies play less of a role. However, the largest winning project in the T4 auction was a 1.7 GW CCGT project owned by the Czech utility EPH.
Amidst rapid investment into batteries there is little wonder that the sector is set to explode in size. While currently just shy of 2.5 GW of batteries are built, Modo forecasts that some 10 GW of batteries will be operational come 2026. Investors say around 800 battery projects are at various stages of development.
On paper, financing all this looks achievable given the flood of deep-pocketed equity investors waiting in the wings – as well as the increasing comfort of lenders with the sector. The largest battery debt deal to date in the sector, a EUR 235m loan to Zenobe in February, involved CIBC, Siemens Bank and Rabobank lending for the first time to the sector, in addition to Santander and NatWest, which had done so previously.
The size of batteries is also growing with Infracapital-backed Zenobe, among the most ambitious of developers, last November planning to build several batteries ranging in size between 200 MW and 400 MW while the bulk of operating batteries today are mostly up to just shy of 50 MW of capacity. Also unlike most operating one-hour batteries, all of Zenobe’s planned batteries can run for two hours. Over 750 MW of four-hour batteries were awarded contracts in the latest T-4 capacity auction.
In theory the sector should keep expanding as renewable energy production grows, given the intermittency of solar and wind power drives the need for batteries. But herein lies an issue that might sooner rather than later keep batteries owners up at night.
Battery owners’ main source of revenue since 2020 has been dynamic containment. But owners agree that these revenues have already begun to taper off as more batteries come online but the amount of DC being procured by National Grid has stayed relatively flat. One senior battery investor says that while DC volumes have risen slightly they “will never get to 10 GW”.
SUSI’s Braakenburg says: “I’ve never built an investment case on one single ancillary service like DC as it is frankly a shallow market. There is only so much frequency response the market needs.”
Harmony Energy’s Paul Mason says revenues are in part being propped up at present by DC and “the fact there is just 2.4 GW of operating batteries” – a reference to the relatively small amount of operating batteries. But as more batteries come online then there will be far greater competition for DC contracts.
Given this and the likelihood that current high wholesale prices are not going to last forever, Mason concludes that “current revenues are not sustainable gong forwards”. He predicts that batteries connecting to the grid in 2026 “will be at the bottom of this current boom” and that “revenues will fall”.
In turn, the proportion of batteries’ revenues from trading is expected to have to rise to compensate for the decline in income from DC.
But the push to trading may cause issues for infrastructure investors that have limited experience of trading and which currently rely largely on third party optimisers to do this for them, some sources claim.
“The issue is that infrastructure investors don’t understand markets stuff as well as the actual physical investment,” says one optimiser. He even predicts that having had a “rush of people into this space, in two years we are going to get a lot of distressed assets in the sector”.
Location, location, location
Some developers admit they have focused largely to date on doing the basics to get a battery into operation. “Many have focused on securing land, grid connections and planning, and bob’s your uncle,” says SUSI’s Blankenberg. “But we now need to shift away from that more passive approach,” he says.
A more active approach involves locating batteries in areas of most constraint: pockets around the UK where there is either too much or too little electricity to ensure supply meets demand, and also where existing infrastructure is not able to cope with excess amounts of electricity. “There are locational benefits to storage that traditional developers hadn’t thought about,” says Blankenberg.
One area of constraint is Scotland, which besides having huge amount of power due to come online from the ScotWind auction also has a smallish population. Therefore much of the power generated in Scotland can be pushed south to England, where demand is far higher due to its far larger population.
But an issue is grid constraint – only two transmission cables sit between the Scotland and England on the so-called B6 boundary. There is therefore large demand for batteries to store electricity and then move it to England when there is sufficient availability of infrastructure to enable it to do so.
National Grid ESO currently in part manages constraint by paying generators in Scotland to not produce electricity in times of excess supply. “There is a certain business case around building a battery in Scotland so you can undercut the wind farms,” says Mason at Harmony. “So rather than paying GBP 70 per MWh, you could be paying GBP 65 per MW and take the power and charge up your battery. Getting paid GBP 65 per MWh to get power in your battery is a pretty good starting place for trading.”
Besides Scotland, another major area of constraint in the UK is in eastern England, around the Humber. Like Scotland, it has huge amount power coming online, including the next two stages of the Dogger Bank offshore wind farm.
Harmony’s first battery at Creykle Beck is located in this area, which says Harmony’s Mason, “could be a potential point of constraint”. He adds: “We don’t price locational into our business case from day one. You view it as strategically interesting place where there might be some upside in the future, but it's not something we rely on.”
There are other drivers behind placing batteries in such constrained decent locations. The UK government has introduced a draft law, Review of Electricity Market Arrangements (REMA), which seeks to introduce different wholesale prices of electricity – the amount paid by suppliers – across 14 areas of England. Currently there is just one price, irrespective of demand. Batteries in an area with high demand will potentially be paid higher than those in zones with lower demand.
Another benefit of locating batteries in areas of constraint is that they are also more likely to win contracts in Balancing Mechanism (BM) – National Grid ESO’s constant auctions to supply electricity in 30-minute slots to ensure power supply meets demand.
The rewards for doing so are lucrative. Anesco earned a spread of GBP 835 per MWh for its contract on one of its best day of trading last year. Yet two coal-fired power stations each managed to get paid a whopping GBP 4,000 per MWh for the same slot between 5pm and 6pm in the BM.
BM auctions rely on power producers to plug gaps in supply near to where they are located. Knowing where these gaps are is therefore key to winning these contracts.
Batteries on paper are well placed to take part in the BM, particularly as the auction requires power sources that can import and export electricity quickly, which batteries are good at doing.
But few batteries have won in the BM auctions, in part due to size. Larger projects that can make up for large shortfalls due to power outages are better placed than smaller ones. Sector specialists say generators most often mixed in the BM auctions have capacity of between 600 MW and 900 MW. But there are exceptions: the first battery to win in the BM back in 2018 was a 10 MW plant in Derbyshire managed by Anesco.
But batteries are having more successes in the BM. Modo said earlier this month that three times more batteries made BM dispatches in January compared to the month previous, although this was only 4.4% of batteries’ total revenues. This was partly due to windy weather, so there was higher demand for batteries in the BM to take power from the grid.
Modo also put it down to “lower inertia” in the system, meaning there was a sudden change in generation which, in turn, meant there was less so-called rotating spinning mass. Such lower inertia changes grid frequency below the 50 hertz target. While often frequency control is handled via dynamic containment, it can also be dealt with via the balancing mechanism, says Robyn Lucas at Modo. Batteries are well equipped to manage frequency movements “as they are quick to respond and will be online for fairly short periods of time”, say Lucas.
Battery specialist sources also attributed it to improvements in National Grid ESO’s control room. One said that until recently operators had to dial in bids by phone. “You didn’t know if someone would pick up the phone on the other end,” said the person. Today it has now become more automated, the sources said.
The CEO of one battery developer said National Grid ESO today better understands batteries. “They previously didn’t understand the state of charge of a battery [i.e. how charged up a battery was at any given time]. Therefore they were hesitant to dispatch batteries for any more than short periods. Now the technology is at the point they can see through to the state of charge of the battery.”
Nonetheless, batteries still only represent around 0.3% of participants in the BM, with the rest made up of conventional power sources.
Inflation starts to bite
Meanwhile, locating batteries in the right places is not the only challenge facing owners. Prices demanded by EPC contractors to source and build a battery as well as pay for grid connections and project management fees have declined sharply from around GBP 1,000 per kilowatt in the mid-2010s, sources say.
Yet EPC costs are on the rise from around GBP 500 in late 2021 to around GBP 600 today, largely driven by rising lithium prices as well as “general inflation of components post-Covid, shipping and rising steel costs”, says a battery contractor.
Foresight’s Ross Driver says EPC costs will remain relatively frothy for the time being until there is more competition among battery manufacturers. “The two largest, China’s CATL and BYD, have market shares of about 70% of the market,” says Driver, although he adds costs are “beginning to come down driven by lower commodity prices, including lithium”.
As a result, cell manufacturers have “a lot more sway now” with EPCs. “It will be harder to get the attention [of EPCs] for smaller projects [which offer lower revenues per contract], says Driver. “You can get smaller projects away but just unlikely to be at good value.”
Foresight’s bargaining chip is that it manages four funds, so he can offer large deals to EPC contractors. “There is a real benefit to working in a coordinated way,” Driver adds.
Meanwhile, getting projects even off the ground is also getting more complicated. While previously DNOs and the National Grid worried about a small number of large connections with generators, today they face thousands of connections, many of them small. Delays getting grid connections are now commonplace as National Grid goes about adding infrastructure to enable the deluge of renewable energy projects to start operating. “The most pressing challenge facing the market is that of grid constraints and planning,” says Anesco’s CEO, Hildagarde McCarville.
McCarville points to a 50MW battery project in the Midlands that was originally offered a grid connection with a DNO for 2025. However, its connection date has now been pushed back to 2033 after National Grid ESO told the DNO that it could not accommodate the connection within the original timeline, even after Anesco had accepted the offer and made initial payments.
“This left us with no choice but to place this project on ice, a concrete project that otherwise we would now be in the midst of detailed engineering and ready to construct in the coming months,” says McCarville.
There are other issues. Phil Thompson, CEO of Liverpool-based battery developer Balance Power says “only half of battery projects get through planning”, partly due to planning committees favouring protecting the look of the countryside.
Thompson also complains that one of his batteries in St Austell in Cornwall will struggle to bid for ancillary services such as dynamic containment due to a technicality. The grid operator has restricted the battery so it cannot export power during peak renewable generation hours and also cannot import power during peak demand hours, meaning the battery is potentially “constrained". “It is very difficult to bid for ancillary services if you’re constrained,” says Thompson. Other battery operators report similar actions by the grid operator.
But wise use of grid connections can offer lucrative returns. An executive at a utility says much connections linked to his batteries are largely unused so he now shares it with “growth industries” such as data centres and electric vehicle charging stations. Investors must take a similarly innovative approach to the sector amidst declining revenues from DC.