Blog post

Are merchant storage markets overrated in the UK?

23 November 2022

A brief blog about the decline in spinning inertia, merchant revenue risks and the cost of capital in UK battery storage

Spinning inertia is not a phrase you hear every day. It sounds a bit like an obscure band who had a one-hit techno wonder in the 1980s. What it actually means, of course, is the inherent inertia, or as my son’s physics text book says “the tendency to continue existing in a state of motion” in the steam turbines of both conventional and nuclear fueled power stations.

 

Generating turbines are big chunky pieces of kit with massive kinetic energy. They don’t readily speed up or slow down. As an aside, engineers over estimating the level of spinning inertia in the turbines was the fundamental cause of the Chernobyl nuclear disaster, but that’s a separate blog about how nuclear engineering is probably not a good career choice for blind optimists. The good thing about spinning inertia is that it means that the generating output of the power station remains very constant over time. As long you don’t recklessly twiddle with the controls, nothing very much happens. In contrast, wind and solar generating capacity lacks this kinetic energy, so the generating output is much more variable as the weather changes. This in turn makes the national grid system more vulnerable to changes in frequency, so if the generating output drops, the frequency drops too. Conversely, if power supply exceeds demand, the frequency goes up. The core frequency of the grid is 50 Hertz, and it is imperative to maintain this core frequency for safe grid operation. But movements away from 50 Hertz are becoming larger and more frequent. At best, these frequency variations make your computer flicker, and at worst, they can cause industrial motors behave erratically and can ultimately trip out the whole electrical grid system.

 

As we retire conventional power stations and replace them with intermittent renewables, the spinning inertia of our electrical system drops and becomes more vulnerable to variations in frequency. To counter this, the network operating companies have been busily procuring long duration battery storage units to act as mini shock absorbers all over the country. If the frequency goes up, they absorb the trillions of surplus electrons, and if the frequency reduces, they then discharge those electrons back into the system. This “grid balancing” takes place every second and to date, it has kept the UK electrical system broadly in check. But, as we transition to carbon-free generation, the system needs a lot more of these little shock absorbers to be in place. There is currently about 1.6 GW of storage connected to the grid, and the Government predicts that this will need to rise to as much as 30 GW by 2030. If a typical 1 MW 2-hour battery costs GBP 650k, then the total investment required in the UK alone in the next eight years is in the order of GBP 18 billion.

 

The grid procures this storage from the private sector through a series of different daily markets that trade with the battery on a bid and offer basis. These cover the availability of the battery (capacity markets), trading between high and lower wholesale prices (wholesale arbitrage markets) and several frequency management markets under the collective banner of Grid Balancing.  These have grown in number and complexity over the years to the point where a typical grid connected battery can participate in at least five different markets. What underpins nearly all these markets (apart from the annual capacity auction) is their reliance on a constant series of small and short-lived electronic trades between the battery and the grid. These trades take place at a bewildering speed and frequency as different parts of the grid either call for additional capacity or seek to dump electrons out of the system by charging up the connected batteries. These trades happen constantly throughout the day and night, with storage providers bidding either to sell their stored electrons into the system or to be paid to absorb them back. It’s an astonishingly complex and clever system but fundamentally it is an unpredictable daily, or even half hourly, market. Therein, dear reader, lies the problem.

 

The one thing you can say with absolute certainty about project finance banks is that they don’t like uncertainty. The non-recourses nature of storage projects means that there is no upside for the bank if projects overperform. Unlike the Chernobyl engineers, banks take the view that things can only ever get worse, not better. This makes project finance banks very risk averse, and like skittish impala when it comes to uncertainty. Hence, their reluctance to fund battery projects, which rely significantly on the two main markets: grid balancing and arbitrage. Even though these revenue streams are basically backed by the National Grid, and banks understand the need for GBP 18 billion of grid storage investment, the merchant nature of the two main markets is unpalatable.

 

Finance, like nature, abhors a vacuum andthe big infrastructure funds have cheerfully stepped into this gap to fund projects at IRRs between 10% – 12%. Unlike the banks, they typically take equity in projects, so they are happy to take more risk, albeit at a cost of finance 5 – 7% higher than senior debt.

 

To our mind, this is a perverse outcome and an unnecessary cost to the grid, and ultimately the consumer. Battery technology is well-proven, the need for grid balancing is overwhelming, and revenue streams are backed by a state-owned entity, the National Grid. This is not the sort of investment that should be priced at double digit project IRRs. We estimate that this cost to the consumer of these artificially high IRRs is almost GBP 5 billion.

 

To be fair to the Government, they have belatedly recognized that a mainly merchant grid balancing market is artificially inflating the cost of capital for battery projects. Their response to a recent consultation paper has signaled a probable Cap and Floor system for longer duration storage assets. The catch is that they won’t introduce this mechanism until 2024 at the earliest, and it is also unclear as to how existing battery projects will be treated under the new regime. Furthermore, nobody is clear what the Cap and Floor prices will be. Predictably, a lot of the responders to the consultation would like more of an alfresco system, as they want the comforting certainty of a solid floor on downside risk but are less keen on a return-limiting Cap.

 

Overall, it’s a tricky balancing act for the Government. Caps and Floors suggest the existence of a trading market in between. Setting the parameters at a level that attracts investment but not at too high a cost of capital is the real skill here. This also begs the obvious question: what happens when the marginal price of a MWhr of long duration storage on a windy and sunny august afternoon hits the parameters of the Cap. Does the Grid live with the resultant frequency surge? If not, what does it do with those pesky surplus electrons?

 

A tried and trusted alternative pricing regime may be a Contracts for Difference (CFD) system. With this, the market determines the price of a MWhr of storage for a fixed period of say ten years and subsequently invests on the back of revenue certainty. This has worked a treat in offshore wind, so it should be a well-trodden path for long duration storage to follow. The risk for the Government is that they overpay for a CFD against falling battery prices, but that trend seems to have reversed in recent months so it may be more of a theoretical risk than a real one.  

 

In conclusion it is clear that it will be a few more years and pithy blogs to go before the Government settles on a solution. The decline of spinning inertia means that the UK grid needs lots of storage, but the Government is still nervously clambering about in the dark when it comes to deciding on the best way to procure it. Meanwhile, the industry is overpaying for financing what are fundamentally low-risk projects. In theory, constantly trading merchant markets sound like a good thing for discovering the optimum price per MWhr of grid balancing at any given hour of the day. But in reality, the uncertainty of these markets has artificially inflated the cost of capital. Risk and the cost of money are two sides of the same metaphysical coin, and the debt markets need visibly less theoretical risk if they are to step in to fund storage.

 

Having seen the success of CFD regimes in driving down the cost of financing wind power, we prefer this CFD approach to pricing, versus the proposed Cap and Floor approach. But, regardless of the pricing mechanism, the industry is eagerly awaiting more bankable revenue structures so that it can invite cheaper financing to the party. We urge the Government to be bold and to do what is needed to make this happen.