The power market in CWE towards 2030 - a perspective

Introduction

In the coming decade the Northwestern European (NWE) power system will see unprecedented changes, as its countries rally to decarbonize by phasing out coal and transition to renewable energy sources. This is obviously good news seen from an environmental perspective, but also creates other challenging problems. Most notably, as the share of renewables increases, the amount of available supply becomes increasingly dependent on the intermittent nature of the weather, making the system more prone to periods of supply deficits. To add to the complexity, both Belgium and Germany are slated to phase out their nuclear fleets entirely, and Great Britain is significantly reducing theirs, in effect removing a large amount of reliable base load supply. To cope with these changes, substantial grid infrastructure improvements and additions are planned, creating another dimension of uncertainty regarding the future dynamics of the power system and its effect on the power market.

To quantify the impact of the aforementioned changes, we in Refinitiv Power Research have developed a fundamental power market model, which solves the generation dispatch problem across the NWE region for 25 historical weather years. The model runs in an hourly resolution and calculates the generation of conventional power plants on a unit level, flows and power prices from now until the end of 2030. This article below focuses on Germany and discusses the results from our base case run (Run Date: 5th June 2021), which includes our best guess assumptions and uses the traded fuel prices. The consumption used in our model is the public grid demand, which excludes transmission losses and self-consumption. For full access to all the input and output data from our new model, please use our desktop platform Eikon.

Outlook

German conventional generation to decrease substantially in the coming decade

In our base case scenario, we stick to the coal phaseout timeline set by the German government, which targets 15 GW of hard coal and lignite capacity in 2023 and 8 GW of hard coal and 9 GW of lignite capacity in 2030. In between the target dates we linearly scale down the hard coal capacity, removing the least efficient plants first. The closure dates of individual lignite plants are already known and is incorporated in the forecast.

In our base case scenario fossil fuels provide 200 TWh of generation in 2022, or about 40 percent of the yearly German public grid demand. We expect fossil fueled generation to halve by 2030, driven by sustained high carbon prices, the closures of hard coal and lignite plants and the expansion of wind and solar capacity.

The decline in German coal generation is expected to be especially pronounced in the first part of the decade, going from 131 TWh (94 TWh lignite and 37 TWh hard coal) in 2022 to 66 TWh (49 TWh lignite and 17 TWh hard coal) in 2025. After 2025, coal-fired generation is still declining, but at an increasingly slower pace. Gas-fired generation, on the other hand, gradually increases in the start of the decade, peaking at 82 TWh in 2024, before slowly decreasing to 65 TWh in 2030. It becomes the largest fossil fueled generation source in Germany by 2023 in our base case scenario.?

We also observe that fossil fuel generation in Germany will be increasingly sensitive to the weather conditions, as renewable generation becomes a larger part of the generation mix. This is especially pronounced during the winter months, where the amount of wind output can vary substantially from year to year, and the load is high. Taking hard coal as an example, we see the difference between a windy and non-windy year can mean a three-folds increase in hard coal generation in the near end, and up to a ten-folds increase at the end of the decade.

In addition to the coal phaseout, Germany is also exiting nuclear power. In accordance with the German nuclear exit law, 3 reactors totaling 4 GW are slated to go offline at the end of 2021, leaving 3 operational reactors. At the end of 2022 the 3 remaining reactors are scheduled to go offline, shutting off a base load supply which provided 60.8 TWh of electricity in 2020, and which we expect to generate 67 TWh in 2021 and 34 TWh in 2022.

Germany to become net importer from 2022 and onwards

The supply deficit resulting from the German coal and nuclear phaseout will make Germany a net importer of electricity. Already from 2022 we expect Germany to become a yearly net importer for the first time, with net imports totaling 28 TWh. We expect net imports to continue to rise to just under 100 TWh in 2026, before remaining stable at this level for the rest of the decade. This means that in our base case scenario Germany gets between 18 and 20% of its yearly public grid demand from neighboring countries in the latter part of the decade.

Germany far away from 65% RES target with current trajectory

In our base case run we adhere to the renewable buildout trajectory laid out by the German transmission system operator’s medium-term forecast until 2025. From 2026 and onwards we assume solar and wind power capacity to grow linearly, reaching the 2030 targets currently stipulated in the German Renewable Energy Sources Act (100 GW of solar and 91 of GW wind).

We observe that with the above capacity buildout trajectory, and considering the average of 25 weather scenarios, the amount of yearly wind/solar generation is expected to total 179+80 TWh in 2030, combined covering a little over 50% of the total yearly public grid demand, compared with 37.5% in 2020.?

With hydro power buildout expected to be limited, we foresee hydro power generation stable at a yearly output of 15 TWh. If we assume biomass generation to stay at the same level as in 2020 (45.5 TWh), renewables stand for 319.5 TWh, or 62% of the public grid demand in 2030 in our base case scenario. This is considering a 2030 public grid demand of 516 TWh, which means a conservative 7% growth compared with 2019. Assuming self-consumption, pumped hydro demand and grid losses remain constant at 2019 levels (68 TWh), we get a gross consumption of 584 TWh in 2030. This means that when using the German government’s current wind and solar expansion targets and a conservative consumption growth, we still only see renewables covering roughly 55% of the gross demand in 2030, far below the official target of 65%.

Flexible assets to become increasingly important

With the German power system set to have a significantly larger share of renewables, and substantially lower dispatchable capacity, the amount of available generation in the system will be highly contingent on the weather conditions. As a result, we observe in our simulations that in periods with low RES generation and cold temperatures, the amount of generation in the German system plus the maximum imports is not enough to cover the consumption. To not jeopardize the future security of supply there is thus an increasing need for flexible assets in the form of storage, demand-side management, and new gas-fired power plants. In our most extreme weather scenarios, where cold weather coincides with very low RES output, the single hour supply deficit reaches 20 GW in the winter of 2030. Figure 5 shows the generation stack on the 24th of January 2030 simulated with weather year 1997. The circled areas indicate the amount of electricity that would have to be covered by flexible assets, in this case amounting to a daily total of 65 GWh.

By the same token, our model also foresees increasingly frequent periods where the amount of RES generation is larger than the consumption plus the maximum export capacity, resulting in the curtailment of renewable energy. This occurs throughout the year, but is the most significant in the spring months, where solar power output approaches its peak, and wind power output can still be high. The figure below shows the generation stack on the 28th of April 2030, simulated with weather year 2003. In this weather scenario we see that the combined solar and wind output totals 85 GW in the middle of the day, 40 GW above the demand. We observe curtailment for an 8-hour period, reaching a maximum of nearly 20 GW, and totaling 77 GWh over the whole period. This further emphasizes the need and value of storage solutions, where the surplus electricity could be stored and utilized at a later stage or converted into another energy carrier such as hydrogen and used as feedstock in industrial processes.

Prices to decline, sensitivity to weather to rise

We expect prices to decline throughout the decade, with prices averaging 62.1 EUR/MWh from 2022 to 2025 and 57.8 EUR/MWh from 2026 to 2030. The prices reflex the prices of the fuel complex from the 5th June 2021. The downward trend can mainly be attributed to renewables covering an increasing amount of the demand, displacing more expensive conventional power plants. The impact of the weather on the price is increasingly visible, especially when looking at the average monthly price. In the winter months, the price differentials between scenarios with high and low renewables output are already substantial from the winter of 2021/2022. The differentials continue to increase further out on the curve, as the thermal stack decreases, and the system becomes increasingly dependent on intermittent renewable generation.

The price differentials between the weather scenarios are comparatively low in the spring and summer months, at least in the near end. The reason for this is two-fold; firstly, solar radiation does not vary in magnitude from year to year to such a large extent as wind speed does in the winter. Secondly because the temperature is higher, the load to cover is lower. This means that there is a lower chance of ending up in tight supply-demand situations, where high priced thermal assets need to supply power. This being said, we see that as the share of renewables increase, the price becomes increasingly sensitive to the weather also in the summer and spring months.

Methodological note

Our long-term price model covers 6 countries (DEU, FRA, AUT, BEL, NLD, GBR). The model runs with traded fuel prices, and with traded forward prices for the border countries. The consumption used in our model is the public grid demand, which excludes transmission losses, pumped hydro demand and self-consumption. The results of the model are based on 25 weather years (1991-2015), weighted equally. The weather and all the model inputs and outputs are in hourly resolution. The run date of the model presented above is the 5th of June 2021.

For questions, please contact me or the Refinitiv European Power Research Team at?[email protected]