The Moving Landscape of Stationary Storage: A short introduction to electricity markets
Discover more from Intercalation Station
The Moving Landscape of Stationary Storage: A short introduction to electricity markets
Building on our series on stationary storage, brought to you by Gaël! Click here for part 1 (talking about behind- and front-of-the-meter) and part 2 (defining the Levelized Cost of Storage).
In the second episode of our series on stationary storage, I concluded that LCOS (Levelized Cost of Storage) is a very useful techno-economic indicator for storage systems, but somewhat oversimplifies the revenue side of operating a Battery Energy Storage System (BESS) on electricity markets.
I already showed a few examples of In-Front-Of-The-Meter markets in part 1 of our series (such as FCR and intraday), now I would like to go into more detail about how these prices are calculated, how storage systems can operate on these markets, and where profitability may be optimal for battery on the mid to long-term.
Section 1 will provide some context on why there is a grid frequency. Section 2 will provide some context on how and why it can be regulated through electricity markets. Section 3 will focus on batteries, and the role they currently play on the grid. Finally, section 4 and 5 will provide some future pathways for battery revenue on electricity markets.
The article is intended to be concise, but don’t hesitate to jump to the variety of links provided throughout for more details!
1. History of the grid in (very) brief
Here is a grid history crash course in thirty seconds:
1880s: The first International Expositions of Electricity take place in Paris. Among the exhibitors, Edison demonstrates the efficiency of the electric lightbulb. His intuition is that the real money lies not in selling the bulbs themselves, but rather selling electricity to customers equipped with lightbulbs.
1880s-1890s: The war of the currents. Tesla and Edison (and their respective companies, Westinghouse Electric and General Electric) disagree on two strategies for carrying electrical current in wires: Edison advocates for Direct Current (DC), while Tesla advocates for Alternating Current (AC). The AC system presents the advantages of lower losses over long distances, but requires high-voltage transformers which operate at a constant frequency.
1896: AC wins. The Niagara Falls powerplant is linked to the city of Buffalo via AC lines, marking the beginning of a worldwide rise of AC connections thanks to their lower losses and better economy of scale.
1896-1926: The Wild West of electric utilities. Many electric utility companies emerge with a relatively simple business model: building generation power plants (coal, oil or hydro), connecting them to customers with transmission equipment, and selling electricity as a service. One challenge though: deviations between production and consumption cause the frequency to change, leading to equipment failure and blackouts. Some utilities start forming joint operations to share peak load coverage and backup power.
1926: The National Grid. In the UK, local electrical networks merge to form the National Grid, a nationwide synchronous AC grid running on 132 kV, 50 Hz lines. This further demonstrates the scale effect of AC electric grids: more customers lead to smoother demand and higher load factors for the production power plants.
1926-1974: The centralization of electric grids. Following the example set by the UK, many industrializing countries build their own centralized AC networks to power their growing industries. The standardization of grid nominal frequency becomes a question, with no definite answers. The US settle on 60 Hz, Europe on 50 Hz and Japan… on both. Transmission System Operators (TSOs) emerge, tasked with coordinating power producers to meet demand and keep the frequency constant over time. In many countries, fossil-based generation dominates, and electricity monopolies form.
1973-1974: The oil shock in the 1973 oil crisis. The sudden increase in oil prices sends electricity prices up, causing a global slowdown in economic output. Fossil fuels alternatives such as solar panels, wind turbines, large-scale batteries and hydrogen start seeing some investment. The fossil fuel-powered national grid model is questioned.
1980s-2000s: The liberalization of electricity markets. The idea that a shift towards private ownership and competition, as opposed to state-control, would improve the efficiency of the grid starts gaining traction. First in Chile, the US and Thatcher’s UK, then to the rest of the world. National system operators start devising financial instruments to optimise cross-border energy exchanges and enable power trading.
2000-present: The rise of renewables. Solar panels and wind turbines are installed in growing numbers on the world’s grids, and electricity supply is now being impacted by weather forecasts. Electricity markets become an increasingly important tool to match supply and demand at all times.
2017-present: Batteries as frequency regulation instruments. The Hornsdale Power Reserve in Australia marks the emergence of batteries as frequency regulation instruments for the world’s AC grids, a position that was until then occupied by gas-peaking plants.
In the following section, I will go into more detail about how modern electricity markets are designed.
2. Electricity markets in (very) brief
I will paint a broad (and short) picture, for more details feel free to head to this write-up on understanding electricity markets by Art Lapinsch.
If you should remember only one thing about electricity markets: they are financial instruments aimed at matching the supply and demand of electricity and therefore keeping the grid frequency constant at all times. Markets are differentiated based on how long in advance electricity is traded, from seconds to months before physical delivery.
Since all countries use different market structures as well as different nomenclature, the following graph is not completely accurate but should constitute a good high-level view:
Futures markets allow to hedge against long-term risks. For example, a pumped hydro powerplant might fill its reservoirs during spring (when glaciers are melting) and commit its available energy to a buyer eager to secure an electricity supply for the upcoming summer.
As forecasts in future markets can not account for up-to-date events, The day-ahead market is where electricity is traded for the next day. The Transmission System Operator (TSO) takes sellers’ bids and calculates a clearing price based on the demand curve:
Gas, coal, nuclear and renewable plant operators will submit bids based on their operating and startup costs, and the TSO will select them from the cheapest to most expensive source.
For an illustration, I like this YouTube video on a hypothetical 2030 Algerian market from Bilal Dahlab, and this one on intraday and day-ahead trading in Great-Britain from Modo Energy.
Of course, the forecasted curves may not represent accurately the actual energy that is produced (if weather forecasts are inaccurate and wind farms produce less than expected, for example) or consumed (if the temperature is lower than expected and more electric heaters are turned on than expected, for example). The intraday market is useful to adjust the quantities of electricity traded closer to delivery time.
Finally, the frequency reserve markets are instruments that are directly tied to the grid frequency: if it deviates too much from its nominal value, the TSO activates the reserves (primary first, then secondary and tertiary if needed) to regulate the frequency up or down. On the primary reserve market it’s usually not energy (in MWh) that is traded, but rather capacity (in MW), and often must be delivered within a few seconds.
One example I like to illustrate is the importance of frequency reserves: TV pickups. Football is pretty popular in the UK, and so is tea. Consequently, what do millions of Britons do at half-time when the Three Lions are playing?… Turn on their electric kettle to boil water. This causes significant spikes in electricity demand and therefore deviations in the grid frequency. Some of the largest TV pickups events include the penalty shoot-out between England and Germany in 1990 and the 2019 rugby world cup final between England and South Africa.
3. And batteries, in all that?
Batteries as frequency regulation devices
Batteries tend to be very responsive systems but hold relatively small volumes of energy (especially in a grid context, where you compare them to pumped hydro systems). They are therefore very suited to bid on the primary frequency reserves.
Below is an illustration of how bidding actually works on the European FCR market, from the European ENTSO-E transparency platform:
However, electricity markets obey the same rules as regular markets: more competitiveness usually means lower margins. With a growing number of batteries being installed on the grid, the bids on the FCR markets become more and more competitive (see the “non-awarded bids” on the top right of the figure), eventually leading to market saturation and revenue losses for the battery systems.
Towards a market saturation soon?
According to Rystad Energy, we should see a significant growth of grid-operating batteries in the coming decade:
For reference, in France, the average need for the FCR service is about 500 MW, while grid-scale batteries will soon be nearing 1 GW. Given the pace of the installation of batteries worldwide, it is obvious that there will soon be too many non-awarded bids for the FCR market to remain profitable in the long term.
Thus, battery system operators are increasingly turning their attention to other markets in order to diversify their revenue streams. From this 2020 article published in PV magazine:
This type of revenue diversification is already well underway for battery projects in Britain. When the market started to grow in 2016, the revenue stack for batteries was concentrated on primary reserve (FFR) [Fast Frequency Response, an ancillary servive market used in GB] and capacity market [a market where batteries are rewarded for committing capacity, even if they don’t end up being used]. But as more battery projects were built, FFR prices dropped quickly from £20/MW/hour to below £5/MW/hour [The prices here are indicated in £/MW/h and not £/MWh because if the service is used, even for 5 minutes, the battery is rewarded on the MW rate for the whole hour]. There has been a recent upturn in prices, but the market is now saturated and, with more than 1 GW of batteries commissioned, assets are unable to win contracts at a viable level on a long-term basis. Projects are now assuming that they will trade in energy markets [spot markets] and the balancing mechanism [a real-time market used in GB] for a large proportion of their revenue rather than depending solely on ancillary services [the frequency regulation markets] and the capacity market [the market where batteries are rewarded for comitting capacity, even if it isn’t used].
Same opinion in Germany, in this market review of stationary storage systems from Figgener et al.:
Prices for FCR [Frequency Containment Reserve, a primary reserve market used in mainland Europe] have halved since 2015 and almost the entire FCR market will soon be served by [batteries] that will result in stronger competition and ongoing price decreases. It is assumed, that the profitability of this market will become ever more constrained and this will disincentive new projects.
Frequency regulation still profitable, despite forecasts of saturation
In the UK, Modo Energy publishes regular updates, podcasts and articles on the state of the UK market. Here is one of their updates for Q2 2023:
Points 2 and 3 are particularly interesting, as it seems that frequency response is still alive and well. So, despite a move towards the Balancing Mechanism [a market used in GB, feel free to head to this video from Modo Energy for more details] and wholesale markets from some battery operators, there are still viable strategies in pure ancillary services.
In Mainland Europe, it seems that battery operators are turning their attention to the secondary reserve markets, or automatic Frequency Restoration Reserve (aFRR).
4. Is the future wholesale?
In a podcast, again from Modo Energy, entitled “Trading flexibility in wholesale market” with Enspired Trading:
What we believe is that wholesale markets is the place to be. The old aggregator models were based on ancillary services (offering flexibility to the TSO) which are starting to saturate. We also see that there is no clear correlation between increasing share of renewables and demand from TSOs [on ancillary services], but we see that increase very clearly on spot markets. So, spot markets are actually the markets which help actively integrating renewables before ancillary services have to jump in. We believe wholesale markets [spreads] will grow with renewables.
The last point, in particular, was recently echoed by Power Magazine on the “duck curve” starting to look like a “canyon curve”, a phenomenon coined by the California Independent System Operator (CAISO) to describe the impact of solar on the grid:
The “canyon curve” in particular leads to large spreads with prices that sometimes go negative, which means good business for batteries operating on wholesale markets. This is not exclusive to California, as Modo Energy reported in a recent Linkedin posts significant revenue from negative wholesale price for the Hawkers Hill battery system
Another point that I find interesting from Enspired Trading is the following
Quentin (Modo) : What about operating in all these different countries? Different TSOs, different credit cover requirements, … The overhead cost must be huge. Jürgen (Enspired) : Wholesale is 100x easier to scale than ancillary services. Ancillary services are totally different in UK and in Germany. Wholesale-wise? It’s not that bad. The route to market in Nord Pool or EPEX is basically the same. The data you need [for forecasting and trading algorithms] is basically the same. So, from a business model perspective there is an advantage there.
Yes, ancillary services can be a bit of a headache. From DCL, DCH, DRL, and DRH in Great Britain, to FFR, FCR, mFRR, and aFRR in mainland Europe, to RRS, ECRS, and NSR in Texas. All of these perform roughly the same functions but the rules for registration, participation and/or stacking vary greatly. The day-ahead market, on the other hand, is pretty similar everywhere aside from a few details on tradable products and gate closure time.
The only problem, as Jürgen mentions in the interview, relates to collateral requirements, the “pay-to-play” fee to trade on the markets, which scales linearly with the average power price.
This is less of a problem if you are trading short-term positions on a continuous intra-day market, but becomes much more of a problem in futures markets where you can accumulate huge positions and run into liquidity problems.
So, while continuous intraday markets seem very appropriate for batteries, day-ahead requires higher collaterals and more accurate price forecasts (a day in advance rather than a couple of hours), and futures may not be very attractive as the collaterals are high and the energy available from a battery is quite low.
5. Revenue stacking
The ultimate strategy for batteries operating on the network is to stack revenue from different sources.
This LinkedIn post from Pr. Oliver Schmidt sums up different ways to stack revenues.
An example of stacking would be:
Bid some of the battery energy on wholesale markets, especially when prices turn negative
Bid the remaining capacity on frequency regulation services when opportunities arise
Commit to capacity markets whenever the profitability of other options is low.
6. Conclusion
Electricity markets are constantly evolving, even more so with the increasing penetration of renewables and batteries. While exciting, this also makes it difficult to perform accurate, future-proof modelling of revenue streams for batteries operating in electricity markets.
On the flip side, this also means that there is a very fast developing job market in the stationary battery sector: not only for building and supervising the day-to-day operation of the battery but also in optimizing its revenue streams by supervising the bidding and trading on different markets.
And as per the conclusion of my Intercalation article BattGPT or AI bubble?, I believe electricity trading and battery management are highly suited applications for machine learning, so pay attention to the startups emerging in this field!
🌞 Thanks for reading!
📧 For tips, feedback, or inquiries - reach out
🌐 Follow us on Twitter, LinkedIn, and our website
😇 Learn more about angel investing through Intercalation Collective