2023 unwrapped: A BESS focussed review of another year in the GB market
Last year, the cost-of-living crisis dominated energy headlines. This year, rising inflation rates and the increasing cost of capital have been at the fore. The growing penetration of intermittent, renewable generation, especially wind, has increased the need for system flexibility. In recognition of this, National Grid Electricity System Operator (NGESO) has made significant headway in addressing the challenge. They have updated connection rules so that assets can connect faster and consulted on new balancing services to support this. It has also been a good year for technology. Improved technology supports market providers to adopt more sophisticated optimisation strategies and NGESO to dispatch assets more efficiently in the balancing mechanism. Nevertheless, across all of these, we must do more to achieve our net zero targets.
1. Record levels of renewable generation and low grid carbon intensity, but will we reach our targets?
Following a decade of increases, wind generation had another record-breaking year (Fig. 1). In January, wind generation peaked above 21.6 GW and met more than 50% of the UK’s electricity demand. Increasing levels of renewable generation led to an all-time low GB grid carbon intensity of 27gCO2/kWh and a 17%[1] decrease in annual average grid carbon intensity from 185gCO2/kWh to 153gCO2/kWh. With the government targeting an average carbon intensity of 50gCO2/kWh by 2030, 50 GW of offshore wind and easing restrictions on onshore wind, wind generation should regularly break records providing flexibility can keep pace. 50 GW of offshore wind generating close to full capacity would be enough to power every home in Britain today. However, starting from just over 14 GW offshore wind capacity today, it’s a large expansion.
Despite increasing levels of wind generation, curtailment volumes have stayed relatively stable for the last three years (Fig. 2). This suggests that flexibility is helping to integrate wind onto the system.
Reaching our offshore wind targets is needed for decarbonising the power sector. We still must procure a further 25 GW wind capacity (Fig. 3) and this year’s Contract for Difference (CfD) auction did not result in any contracts. The strike price of £44/MWh did not make sense given project costs. Hopefully the next CfD auction will make more progress: the Government will increase the strike price by 66% to £73/MWh. Despite recent increases in cost, offshore wind is still cheaper for the end consumer than gas that currently set the average wholesale price of £98/MWh. Even once we build the turbines, integrating large intermittent generation volumes will be a challenge for system operators, especially given the currently limited options for long duration storage.
2. Increasing levels of intermittent generation means frequency products are working harder than ever
Last month, NGESO phased out the procurement of Dynamic Firm Frequency Response (DFFR) and launched the Enduring Auction Capability (EAC) which allows providers to stack the three frequency response products, dynamic containment (DC), dynamic regulation (DR) and dynamic moderation (DM). Stacking them allows the ESO more optionality to select the lowest cost bids for the consumer. These products respond up to ten times faster than DFFR to keep system frequency at 50Hz which is essential in an increasingly renewable world. Retiring fossil fuel generation reduces system inertia which increases the grid’s vulnerability to frequency swings. Looking back at the legacy DFFR requirements, BESS delivering frequency services are now working twice as hard to maintain grid frequency than eight years ago (Fig. 4).
3. However, the value pool for frequency services is limited
To date, GB BESS returns have been dominated by ancillary services. But, for years it has been speculated that the revenue stack would move towards a more trading centric business model.
This year, as installed BESS capacity increased, we saw ancillary service market saturation drive a structural transition in the revenue stack. NGESO’s frequency services demand is currently between 1.0–2.5GW, compared to an installed BESS of 3.1GW. Average prices have accordingly also started to converge with the average DA price spread (opportunity cost of participating in other markets) (Fig. 5).
4. A trading centric model led by renewable buildout will move value closer to real time
As intermittent renewable generation on the system increases, closer to real time markets should become more volatile and consequently more valuable (Fig. 6). This is reflected by a more volatile system imbalance price over the last 6 years.
Today, CCGTs and coal still play a huge role in balancing when renewable generation is low — making up over 80% of accepted offers (to turn up generation). However, gas assets waiting to be used in the BM are often running at lower efficiencies than fully utilised gas assets and therefore their emissions per MWh are higher.
BESS currently makes up less than 1% of all offers. Therefore, BM revenues have made a relatively minor contribution to the BESS revenue stack. This has historically been due to a combination of strong ancillary service returns, meaning assets are not always available to be called upon in the BM, alongside it being challenging for the NGESO to dispatch smaller assets. NGESO doesn’t have visibility on real-time state of charge and the technology required to bulk dispatch smaller assets wasn’t sufficient. This made it difficult to understand a battery capability and to use multiple batteries to supply a larger volume, therefore making it safest and fastest to use thermal units.
This year, NGESO committed to updating parameters to ensure that batteries can be dispatched for longer actions. Earlier this week, the launch of the Open Balancing Program illustrated the benefits of the bulk dispatch tool which saw BOA volumes for BESS double (Fig. 7).
5. Although trading is the future, revenues are challenging to get comfortable with
The GB BESS fleet predominantly comprises one hour duration batteries (Fig. 8). This is unsurprising given an ancillary service dominated revenue stack[2] remunerate assets on power rather than energy. Moving towards a more trading centric model signifies the potential longer-term shift to multi-hour Li-ion BESS. However, a large proportion of the BESS pipeline are still one hour in duration.
The value achieved on the wholesale market varies based on market volatility (Fig. 9). But, even in periods of high energy arbitrage, many assets will still have opted to participate in ancillary service markets as they are an easier to access without sophisticated optimisation technology[3].
Among the backdrop of varying wholesale volatility, revenues from two-hour BESS would have been approximately 60–70% greater those from one hour BESS. This year the gap rose to over 80% due to the changing relationship between day ahead, intra day and cash out prices enabling two hour BESS to see greater gains when re-optimising closer to real-time. However, there are reasons to why some developers still choose to build one-hour systems, including:
- availability of land / sites
- absolute capex (a 2hr system is approx 60% more expensive than 1hr)
- markets closer to real-time are often harder to capture as they require sophisticated forecasting and optimisation strategies.
In the future, we would expect this trend to continue as the buildout of intermittent generation moves market volatility closer to real-time.
6. The rising role of technology
Moving away from ancillary services towards energy arbitrage highlights the importance of optimiser technology. The best performing optimisers in the market are currently achieving over 50% more than the average, and a large proportion of this revenue can be attributed to energy trading (Fig. 10). However, a lack of market track record makes it difficult to benchmark how players will perform longer term. Historically, an absence of wholesale and balancing mechanism access has not been detrimental to the BESS business case. But the changing revenue stack will open stronger value differentiation across battery optimisers next year. Their trading, optimisation, data analytics and systems capabilities will all matter. Approval of code modification P415 will be welcomed by many optimisers requiring their own access to wholesale markets and will underpin the viability of battery investments going forward.
Asset owners will have the challenge of benchmarking optimiser performance on previous markets that are fundamentally very different to the future. Optimisers will need to not only consider the financial optimisation but also the physical operation. For example, assets that have delivered DC for the first three years of their life could have an energy throughput that’s nine times lower than an asset adopting an energy arbitrage strategy and therefore a very different degradation profile.
7. Reaching net zero depends on low carbon technologies’ ability to compete
Several support mechanisms are still in place for fossil fuelled assets: last year’s Energy Security Strategy references continued support for domestic oil and gas production and the Capacity Market (CM) has already procured approximately 7 GW of carbon intensive assets beyond 2035, the date set by the government for a net zero electricity system. Whilst this ensures security of supply, it highlights that the CM relies on fossil fuel firm capacity because there aren’t sufficient clean dispatchable options.
Allowing carbon intensive generation continued access to support risks hindering low carbon technologies’ ability to compete successfully. Furthermore, it’s ultimately the consumer who funds this support and this is likely to remain high (Fig. 11) because new build capacity has a high probability of setting the price given the existing volume is usually lower than procurement targets.
Additionally, UK ETS prices are now trading over 50% lower than the carbon price in the EU. This is due to an increase in allowances and a decrease in demand[4]. Earlier this year, the Government signalled a boost to allowance supply by 53.5mn tonnes between 2024–2027, the equivalent to half a year’s worth of UK emissions covered by the scheme. Tightening the UK ETS auction parameters could send investors the right investment signal.
Conclusion
It is well recognised that BESS will have a significant role in a future decarbonised GB electricity system. The increasing demand from electrification of heat and transport, and growth in intermittent renewable capacity, will mean that the fundamental requirement for low carbon flexibility will continue to grow.
In the short term, the growing penetration of intermittent generation has already increased the need for flexibility. For BESS, this has highlighted the importance of technology and cross market optimisation. Asset owners will need to ensure they select the right optimiser to manage their assets, and investors will need to select the right locations as the BM becomes a more crucial part of the revenue stack.
By 2030 the UK energy system will be very different from today. Coal will have retired, 6 GW of nuclear will have been decommissioned and 7 GW of ageing CCGTs will retire — a total loss of just under 20 GW or 50% of average electricity demand. Whilst renewable energy will continue to grow, it is unlikely we will reach the government’s targets for offshore wind. Therefore, we need to ensure that revenue streams are in place to accelerate clean dispatchable baseload generation deployment and support short duration demand side flexibility.
[1] From Jan 1st — Nov 30th 2023.
[2] In 2016, Enhanced Frequency Response led to the build out of 400MW of BESS, with an average duration of 30 minutes. 2018, NGESO launched DFFR, contracted over multi year timeframes and leading to assets with 0.5–1hr to be developed.
[3] Stringent technical operational and performance monitoring requirements for DC, DR, DM mean that failing to deliver the service results in loss of revenue.
[4] July brought lower than usual levels of electricity demand, due to mild weather conditions and higher than usual renewable generation, leading to less use of coal and gas.
