The Dutch electricity sector - part 4: Changing electricity markets present opportunities and risks for businesses and households

The table below presents a summary of this article. If you are not familiar with the different players in the electricity sector and the workings of the various Dutch electricity markets, we recommend that you read at least the first two parts of this series.

Forward and futures market

The Dutch futures electricity market is not very liquid.[1] Though there is some trading in contracts for baseload power[2] for the following calendar year and the year after, there is little to no trading in contracts for shorter periods or for peak load. The German futures market is the most liquid market in Europe, and German prices are fairly similar to Dutch prices.

Table 2 shows a declining trend in prices for baseload contract prices, which implies that the market expects average power prices to drop in the coming years. This is not surprising, as future natural gas prices are currently also relatively stable and even declining slightly.[3] Of course, this outlook could change overnight due to unforeseen circumstances such as geopolitical developments.

[1] This means there is little trading.

[2] Baseload refers to a fixed amount of electricity offtake/production over a certain period of time.

[3] Source: ICE Endex Dutch TTF Natural Gas Futures.

In recent years, fixed-price PPAs have been concluded with major creditworthy electricity customers for many large-scale solar and wind farms in the Netherlands. But these customers are rare, which makes it unlikely that all newly built parks can also conclude long-term fixed-price PPAs. We expect developers of new renewable energy projects (mainly offshore wind) to increasingly lock in the price of future electricity production using futures. This will increase trading on the futures market. However, certainty about prices in the long term (ideally for at least ten years) is crucial for the financeability of new parks, especially if there is no revenue guarantee in the form of subsidies.

Day-ahead market

The day-ahead market is a volatile market in which the price can vary significantly per hour. Until 2020, this volatility was relatively limited, but the energy transition and the war in Ukraine have changed that. Figure 1 shows the impact.

To date, the energy transition has at certain times[4] resulted in lower prices. Hours with low and negative prices have increased, and in 2023, the price dropped to EUR -500 per MWh for the first time (see Table 3).[5] Especially the war in Ukraine and the ensuing skyrocketing natural gas prices impacted the maximum price levels in 2021 and 2022, resulting in extreme volatility in the day-ahead market in these years. Because the price of natural gas has since stabilized considerably, the maximum prices of electricity have also fallen, and volatility has decreased. However, the market has yet to return to 2020 levels.

[4] There is no evidence to date that the energy transition has resulted in lower energy bills for consumers.

[5] At that time (and also at the time of publication of this article), this is the prevailing minimum price. The maximum price is currently EUR 4,000/MWh.

Due to the increase in weather-dependent production capacity, the day-ahead market is expected to remain relatively volatile in the coming years, compared to 2020. This volatility results in an increasing number of hours with negative prices and a lower average price. Still, we do not expect volatility to return to the levels of 2021 and 2022: price spikes will be lower,[6] and there has been an increase in flexible capacity, such as battery storage and demand-side management (DSM) by businesses and households.

From 2030 onward, however, the number of hours with (very) high electricity prices may increase. This is due in part to the planned closure of existing coal-fired power plants, which will negatively impact the capacity to match the expected supply of electricity to the expected demand. This problem will arise especially during longer periods (days to weeks) with little sun and wind. Whereas natural gas or coal prices currently often determine the level of price spikes in the day-ahead market, a scarcity of non-weather adjustable generation capacity may influence price spikes starting in 2030. At the same time, we will see fewer hours with negative prices. This is due to the fact that more assets, such as wind turbines or solar panels, will be switched off in the event of negative prices,[7] further increasing flexible capacity.

In recent years, the yearly traded volume in the Dutch day-ahead market has increased to about 50 TWh, while electricity consumption has remained largely the same. This indicates that more trades are taking place, despite only minimal change in demand.

Intraday market

As we explained in part 2 of this series, prices in the intraday market are set in a “pay-as-bid” process. Because there is no uniform (marginal) pricing, insight into price trends is limited. However, TenneT's Annual Market Update 2021 does show that the number of transactions and volumes traded in this market increased between 2019 and 2021. According to TenneT, this suggests a growing divergence between day-ahead market forecasts and reality, which BRPs try to correct for on the intraday market. This unpredictability can be attributed to the growing influence of weather on electricity production. Most intraday market transactions take place in the afternoon, followed by the evening; the fewest transactions take place at night and in the morning. This is because the time between the closure of the day-ahead market and the actual production and delivery of electricity is shortest during these parts of the day. Most of the transactions in the intraday market take place one to three hours before delivery.

The intraday market is much smaller than the day-ahead market. Currently in the Netherlands, about five to nine TWh are traded in the intraday market per year, but this volume may grow significantly in the coming years. This increase would be due to Balancing Responsible Parties (BRPs) increasingly wanting to update their portfolios at the last minute, based on changeable factors such as weather information.

[6] Price spikes will be lower on the condition that natural gas prices remain relatively stable.

[7] As an example, the Amer power plant currently often continues to run even when electricity prices are negative, because it also provides heat for a district heating grid. However, they will stop doing so, and this change will cause an increase in flexible capacity. We also expect that residential rooftop solar panels will increasingly be smartly controlled and will therefore also respond to price signals. Finally, large-scale wind and solar farms that by that time will no longer receive subsidies, will switch off more quickly in the event of negative prices.

Demand for balancing capacity and energy will rise in the coming years, as more solar panels and wind turbines are installed (see part 3 of this series). The availability of wind and solar is difficult to predict with exact precision, so when large volumes of solar and wind capacity are installed, even a small deviation can significantly impact the balance of the power system. At the same time, the supply of flexible power is rising, and we expect stable to slightly declining natural gas prices. Given all of this, we expect average fees for providing balancing capacity and energy to remain stable and possibly even decrease in the coming years. During (prolonged) periods of low wind and sun availability, however, the fees for balancing capacity may actually be high.

FCR market

The size of the frequency containment reserve (FCR) market is related to the outage of the largest generator or end consumer in the synchronized European grid. Currently, a total of 3,000 MW of FCR capacity must be contracted by the joint TSOs. Of this total capacity, TenneT must contract a minimum of 110 MW (and a maximum of 210 MW) by 2024. They are obligated to contract at least 30% (that is, 33 MW) of that 110 MW in the Netherlands and permitted to contract the remaining capacity abroad. Conversely, foreign TSOs are also allowed to contract part of their FCR capacity in the Netherlands. So, the capacity that does not have to be contracted in the Netherlands has to compete with international prices.

Data from Entso-e indicates that, in the Netherlands, an average of about 55 MW of FCR capacity has been contracted daily over the past five years. This indicates that Dutch FCR capacity is relatively expensive compared to foreign FCR capacity. The increasing deployment of battery storage in the Netherlands is expected to lead TenneT to contract more FCR capacity within the Netherlands in the coming years, rising to 120 MW in 2033. Nevertheless, the FCR market remains relatively small.

As described earlier in this series, the FCR market operates with marginal pricing.[8] In 2021 and 2022, extremely high natural gas prices caused the average price to increase (see Table 4). Gas-fired thermal power plants played a bigger role in the supply of FCR power at the time than they do today. Today, batteries supply the vast majority of FCR power. But natural gas prices can still affect FCR prices: high natural gas prices lead to higher prices in the spot markets (at times when natural gas power plants are price-setting), which increases the opportunity cost[9] of deploying batteries in the FCR market and, as a result, normally increases FCR prices.

[8] The price of the most expensive selected provider sets the prices for all selected bids.

[9] Opportunity costs reflect the cost of the best missed opportunity. If a battery can make a lot of money in the day-ahead market, why offer it at lower fees in the FCR market?

In 2023, the average fee for FCR power would have dropped to its pre-war level. However, the average was influenced by an extremely high price of nearly EUR 20,000 paid by TenneT on November 2, 2023, between 4:00pm and 8:00pm, about a thousand times more than normal. Due to a combination of circumstances, there were not enough other bids available, leading to the selection of this exorbitant bid. Since then, the number of bids (with lower prices) for FCR has increased, and so TenneT considers it relatively unlikely that such a high bid will ever set the price again. Without this outlier, the average price in 2023 was around EUR 13 per MW per hour. Between July 2020 (when the FCR market switched from contracts per day to contracts per time block) and April 2024, the 8:00am to 12:00pm time block yielded on average the lowest fees and the 4:00am to 8:00am time block yielded the highest. These averages don’t include the one-time outlier of November 2023 (see Figure 2).

We expect stabilized natural gas prices and increasing battery storage to ensure that future fees do not increase and potentially continue to decline.

aFRR market

For now, automatic Frequency Restoration Reserve (aFRR) capacity for the Netherlands is exclusively contracted by TenneT. Since 2021, this has averaged about 340 MW of upward capacity and 390 MW of downward capacity per day.[10] Although more downward capacity is contracted on average than upward capacity, average prices for upward capacity are about 25% higher (see Table 5). aFRR fees also clearly show the effect of the high natural gas prices in recent years, with prices in 2024 so far averaging more than 3.5 times lower than in the peak year of 2022, and price volatility dropping sharply.

[10] Until 2021, capacity was contracted collectively. Since 2021, this has been split into separate contracts for upward and downward capacity.

The total amount of contracted FRR capacity (aFRR+mFRR) will increase significantly around 2030, when the large landfall cables of new wind farms in the North Sea are installed. The failure of such large cables could potentially cause a lot of imbalance at once, and so TenneT expects the total amount of contracted upward capacity for FRR to increase from 1,260 MW in 2023 (340 aFRR + 920 MW mFRR), to 2,000 MW in 2030. The total amount of contracted downward capacity for FRR is expected to increase from 1,280 MW in 2023 (400 aFRR + 880 mFRR) to 1,800 MW in 2030. So the demand for FRR capacity will increase. Supply should also increase, especially if FRR capacity markets are converted from 24 to four hours. TenneT expects that batteries in particular and even wind farms will then offer both upward and downward capacity. In the future, TenneT may also join with other TSOs to contract aFRR capacity collectively, just as it already does with FCR capacity. The idea is that this will reduce the costs for TSOs and, by extension, the revenues for providers. All in all, average aFRR capacity fees are expected to fall until 2030, although prices for upward capacity may be temporarily high during (prolonged) periods of low solar and wind. We expect prices to stabilize once the FRR capacity markets start to expand.

Unlike the FCR market, FRR markets provide remuneration for activated capacity. Please note: When prices for activated downward capacity are set and published by TenneT, positive amounts mean that the Balancing Service Providers (BSPs) have to pay TenneT. When the amount is negative, TenneT has to pay the BSPs. To clarify this, the positive amounts for activated downward capacity are shown in red in Table 6. Compared to the peak year of 2022, less aFRR capacity was activated last year. This indicates that the market was better able to match supply to demand in advance.

When it comes to prices, it’s worth noting that BSPs are generally willing to pay to activate downward capacity in the aFRR market. If downward capacity is supplied by curtailing generation, BSPs are often willing to pay because the electricity they now no longer need to supply has likely already been sold, through the day-ahead market, for example. The BSPs get to keep these revenues.[11] They’re also saving on the use of fuel that would be required to generate electricity with a thermal power plant. As a result, providers of downward capacity can make a profit even when they are paying to curtail electricity production.[12]

BSPs can also provide downward capacity by consuming more electricity, such as by charging a battery. If battery operators expect to be able to sell this electricity for a higher amount at a later moment, they may be willing to pay to charge the battery. Interestingly, the average price for activated downward capacity in 2024 has so far been negative. As explained earlier, this means that TenneT has to pay BSPs, for example, to curtail a wind farm or charge a battery.

In the years that natural gas prices were exceptionally high, so was the average price of both activated upward and downward capacity. Now that natural gas prices are less extreme, average prices for activated aFRR capacity have also dropped. However, price volatility has not. In fact, the price volatility of activated upward capacity is higher than ever. This means that the time of activation is a major factor in determining how much money an aFRR capacity provider stands to make. Because the deployment of battery storage does not incur fuel costs and the supply of batteries in this market is increasing, we expect that the average price for activated aFRR capacity will not increase in the coming years and may continue to decrease.

Imbalance market

As explained in detail in part 2 of this series, the imbalance market is not a real market but a financial settlement for BRPs that deviate from their own E-program (the forecast of electricity offtake and supply of a BRP's portfolio per quarter-hour). Deviations can happen by chance, e.g., due to more or less wind than expected, but may also be deliberate, when a BRP expects to make money by deviating. How much money a BRP can earn or lose depends on the imbalance price at that moment and on the so-called regulation state.

TenneT outlines four regulation states. In regulation state -1, the TSO must activate capacity via the aFRR market due to a surplus of electricity in the system. The reverse is true for regulation state +1. In regulation state 2, both upward and downward capacity is activated.[13] In the event that no capacity is activated at all in a quarter of an hour through the aFRR market, regulation state 0 applies.

The imbalance price per quarter-hour is typically determined by the highest activated price for supplying upward capacity and the lowest (considering negative amounts) activated price for supplying downward capacity in the aFRR market. However, in regulation state 2 the imbalance price is usually determined a different way.

[11] The supply obligation for this electricity will be taken over by other parties.

[12] Guarantees of Origin (GOs) and subsidies can also still generate revenue.

[13] Regulation state -1 also applies to situations where there is also upward regulation, but where the difference between the activated upward regulation and the activated downward regulation power is continuously decreasing or stable. The reverse applies to regulation state +1. In regulation state 2, there is no constant or continuous difference between activated upward and downward regulation. Source: TenneT, 2022

Some examples

Suppose a BRP has informed TenneT in advance that its customers will take off and supply a total net of 0 MWh of electricity the following day between 10:00am and 10:15am. However, it turns out that this BRP's customers actually supplied a net of 30 MWh of electricity, while the system was short of electricity at the time (regulation state +1). In that case, for the extra 30 MWh supplied (the surplus in the BRP's portfolio), the BRP receives the imbalance price set for that quarter-hour for shortages in the market. In 2023, this averaged EUR 198 per MWh (see Table 7), higher than that year's average day-ahead price of EUR 96 per MWh (see Table 3). However, if the BRP had contributed to the deficit in the market, they would have had to pay EUR 20 per MWh (see Table 8).

Another example: a BSP sees that TenneT has activated upward capacity through the aFRR market between 10:00am and 10:08am because there is a shortage of electricity in the system (regulation state +1). The BRP decides to provide passive upward capacity, expecting to receive EUR 198 per MWh for it. Many other parties assume the same, and also start supplying passive upward capacity. As a result, the system deficit turns into a surplus and TenneT has to activate downward capacity between 10:13am and 10:15am. This brings regulation state 2 into effect, which means that the imbalance price for that quarter-hour is determined in a different way. Now, the BSPs that provided upward capacity do not receive EUR 198 per MWh, but only EUR 9 per MWh (see Table 7). This rate may not be enough to cover the marginal costs, depending on how the upward capacity was supplied. Moreover, the activated upward capacity would have made more money if it had been sold through the day-ahead market. So, in hindsight, providing (this amount of) passive upward capacity was a bad decision.

Of course, average annual prices don't say much, and it's better to look at the prices prevailing on a quarter-hourly basis. This reveals that the past few years, 93% of the time, the day-ahead price was higher than the imbalance price for upward capacity at the times when regulation state 2 was in effect. So, when regulation state 2 applies, it is usually unattractive to provide passive upward capacity. Indeed, where BSPs expect to receive money for passively delivered upward capacity, so far this calendar year, in more than half of the cases, they actually had to pay money during regulation state 2. This percentage has increased significantly over the past two years. In addition, imbalance prices are very volatile, more volatile this year than in 2023, and often even more volatile than in the very difficult years of 2021 and 2022.

This means that there is potentially more money to be made in the imbalance market than in the day-ahead market, but the risk of trading in the imbalance market is also greater. This is because the prices are not set in advance. TenneT deliberately makes prices in regulation state 2 unattractive in order to prevent too much passive balancing, which can turn a system deficit into a surplus or vice versa.

Table 9 shows that so far this calendar year there were still proportionally more quarter-hours with regulation state 2 than in previous years. This may be because more battery systems - both large-scale and small-scale - are being used for passive balancing. The question is whether or not the increase in quarter-hours with regulation state 2 will become a trend.

mFRR

Although TenneT contracts more mFRR than aFRR power, the prices of contracted mFRR power are significantly lower. While the average contracted price of upward capacity in the aFRR market is higher than that of downward capacity, the circumstances are not as clear in the mFRR market. For example, the average price of contracted downward capacity this year so far is higher than the price for contracted upward capacity (see the second and fifth columns of Table 10). Moreover, the price for contracted upward capacity is relatively volatile.

Currently, combined heat and power plants (CHPs) of horticulturists and other thermal power plants mainly provide emergency power. It has been agreed that greenhouse horticulture will be climate neutral by 2040. This means that horticulturists must switch from natural gas to other, more sustainable ways of heating their greenhouses. Therefore, it is uncertain how the installed capacity of CHPs at horticulturists will develop in the coming years. Large industrial companies, which also provide some of the emergency power, also need to become more sustainable and will mainly electrify. This could drive up the price of contracted mFRR capacity, especially that of upward capacity. On the other hand, TenneT expects that heat pumps combined with aggregates will also supply mFRR power in time. An increase in the supply of other technologies in this market may actually cause prices to remain relatively stable.

Emergency power was used much more frequently in 2021 and 2022 than in previous years, which is indicative of the chaos in the electricity markets in those years. It has since stabilized considerably (see Table 11). So far, in 2024, significantly more downward than upward capacity has been activated, which is typically the other way around. There seems to be an upward trend in the deployment of downward capacity. This may have to do with the increase of quarter-hours with surplus renewable electricity.

A fixed formula determines the remuneration for activated capacity on the mFRR market. The prices set are usually derived from the prices for activated capacity in the aFRR market in that same quarter-hour. This shows that price developments in the aFRR market impact the remuneration developments of the mFRR market, to some extent.

Network operators can issue redispatch requests through the GOPACS platform. The second part of this series explains congestion management and the workings of GOPACS in greater detail. To date, TenneT is the main user of GOPACS, and Liander, Enexis, and Stedin use it to a lesser extent. The other, smaller DSOs are not using it yet.

Interestingly, traded volumes through GOPACS are lagging significantly this year (see Table 12). One explanation is that a relatively large amount of day-ahead congestion management has been done through capacity limiting contracts (CBCs). As a result, grid operators have had to deploy less redispatch. The deployment of CBCs is not yet regulated through GOPACS, but this will change in the future.

Because the demand for flexible capacity to prevent or alleviate congestion problems will increase in the coming years, grid operators are hoping that more parties will start to offer flexible capacity, including through GOPACS.

In 2021 and 2022, the situation on the Dutch electricity markets was exceptional, with high average prices, extreme price spikes, and major price volatility. At the time, hardly anyone had seen the situation coming. This shows the major impact that factors such as geopolitical developments can have, and that expectations sometimes have to be adjusted quickly. Since then, the markets have stabilized considerably, though both prices and volatility remain higher than before 2021.

Apart from geopolitical developments, the energy transition is also effecting change in the Dutch electricity markets. The timing of electricity production increasingly determines the value of the electricity produced. When there is a lot of sun and/or wind, the value may even be negative. Conversely, customers with hourly fluctuating electricity prices can incur high costs if they use a lot of electricity at times without a lot of sun and/or wind.

Because both the supply of electricity and its consumption are increasingly difficult to predict with exact precision, there is an increased risk of imbalance in BRP portfolios. As a result, energy suppliers are charging higher premiums to customers with a relatively high risk of causing imbalance, such as customers with solar panels. Households and businesses that want to stick with fixed-price contracts and non-flexible production and consumption of electricity are therefore increasingly exposed to surcharges. The advantage of such fixed-price contracts is that all risks are invested in other parties, and that customers have more certainty about costs and/or revenues during the contract period. In the longer term, however, it may actually lead to more uncertainty and higher costs, as BRPs and energy suppliers could decide to pass on their costs in a different way.

One example of such a construction is the implementation of feed-in charges for solar panel owners. Feed-in charges encourage households to flexibly arrange and better coordinate their production and consumption of electricity. This way, they feed less electricity into the grid and the electricity supplier incurs less risk. Small consumers can also deliberately assume risk by entering into a dynamic contract. In that case, the customer assumes the risk of hourly fluctuating prices. Large consumers usually have the option of entering into electricity contracts in which they themselves are responsible for the imbalance risk. In return for taking on these risks, they are offered a more favorable price per kWh (on average). Companies or households should only take on such risks if they fully understand what they are getting into, and if they also have the ability to (partially) manage the risks they incur. For many large companies, it is increasingly difficult to leave risk of imbalance to the energy supplier. For those companies, it is especially crucial to strategically consider (flexible) electricity consumption and production.

As concluded in the previous parts of this series, households and businesses stand to make money by changing the way they consume and/or produce electricity. By investing in flexible capacity and deploying it appropriately, small and large consumers can reduce their (imbalance) risk and benefit from favorable electricity prices. They can make money by providing services to grid operators and by trading in various electricity markets. However, this final installment in our series shows that the earning potential in the trading and balancing markets has significantly declined since 2021 and 2022 and is not expected to rise again in the coming years. The imbalance market is an exception. "Trading" in this market could be very profitable, but this also comes with significant risks. In conclusion, households and companies would do well to invest in the flexibility of their electricity consumption and/or production. At the same time, they should realize that the earning potential of flexible power has its limits.

Thanks to Pim Doodkorte for the raw data from Entso-e.

The electricity sector is evolving rapidly. By 2023, half of the electricity produced was renewable. An increasing number of consumers are generating their own electricity, and electrification of transport and heating is also on the rise. All this has an impact on the load on power grids, the supply and demand balance, security of supply, and electricity prices. These developments present both risks and opportunities for electricity users. To better understand these risks and opportunities, RaboResearch is publishing a series of articles on the Dutch electricity sector.

Previously published:

The Dutch electricity sector - part 1: Who are the players and what is their role?

The Dutch electricity sector - part 2: How do the different electricity markets work?

The Dutch electricity sector - part 3: Developments affecting electricity markets