Trading off regional and overall energy system design flexibility in the net-zero transition

Abstract

The transition to net-zero emissions in Europe is determined by a patchwork of country-level and EU-wide policy, creating coordination challenges in an interconnected system. We use an optimisation model to map out near-optimal energy system designs for 2050, focussing on the planning flexibility of individual regions while maintaining overall system robustness against different weather years, cost assumptions, and land use limitations. Our results reveal extensive flexibility at a regional level, where only few technologies (solar around the Adriatic and wind on the British Isles and in Germany) cannot be substituted. National policymakers can influence renewable energy export and hydrogen strategies significantly, provided they coordinate this with the remaining European system. However, stronger commitment to solar in Southern Europe and Germany unlocks more design options for Europe overall. These results on regional trade-offs facilitate more meaningful policy discussions which are crucial in the transition to a sustainable energy system.

Figures (6)

• Figure 1: Robust trade-offs between annualised onshore wind and solar investment in Europe, consistent with the net-zero emissions target for 2050. (a) shows the overall robust design space projected onto the onshore wind and solar dimensions, with the cost slack marked, as well as the optima in all the individual scenarios marked. Note that the optima from individual scenarios are not necessarily robust (i.e. may not be feasible or near-optimal in other scenarios). (b) -- (c) show how the robust design spaces changes (in dashed lines) subject to various levels of investment in offshore wind and hydrogen infrastructure respectively. For instance, investing only 20 bn EUR into offshore wind requires more investment in solar and/or onshore wind. Note that as cost-optimal European-wide hydrogen infrastructure investment is on average 67 bn EUR, investment beyond this level takes away cost slack from other technologies (here, solar and onshore wind) and forces their values closer to optimum, reducing the extent of the design space. (d) -- (j) show the trade-offs between investment in solar PV and onshore wind for the different focus regions and display the vast opportunities of robust and cost-effective onshore wind-solar substitution. Solar investment is limited in robust designs by the scenarios allowing only a maximum average installation density of 1.7 MW/km$^2$ for utility solar within available land (corresponding to roughly 1.2% of the total land area of the modelling region); see Supplementary Figures S3--5 for alternative versions of this figure with a looser land-use restriction for solar and including land-use restrictions for wind power. The conversion guide from bn EUR (annualised) to GW for different technologies, shown at the bottom, is valid for all figures in the baseline scenarios. In scenarios where solar or wind power are more expensive, the same investment will result in smaller capacities.
• Figure 2: Comparison of minimal regional and European robust investment (in bn EUR, annualised) and average optimal investment across all scenarios. We present the investment levels in the four key technologies required for all robust designs ("PV": solar, "On": onshore wind, "Off": offshore wind, "H2": hydrogen infrastructure) as well as minimal total investment in wind power ("W": on- and offshore wind) and in renewable generation ("RES": solar, onshore wind and offshore wind). The European investment values are the minimum across all focus regions' robust solutions (for the average optimal investments, they are the average across all scenarios and focus regions). Only indispensable investments above 1 bn EUR are annotated. For example, across robust system designs Germany requires investment of at least 19.7 bn EUR in total renewable generation, 8.3 bn EUR of which in wind power. Neither of solar, onshore or offshore wind power on its own is strictly necessary in Germany. Across the 12 scenarios (Methods), German optimal investment in renewable generation is on average 44.2 bn EUR. See Supplementary Figures S8--10 for versions of this figure excluding the restrictive solar land-use scenarios as well as including restrictive wind land-use scenarios; minimum investment levels for wind power of any kind drop significantly for all regions as well as Europe overall if higher solar investment is allowed. For instance, Figure S8 shows that minimum overall European wind power investment drops from 165.8 to only 26.1 bn EUR when the restrictive solar land use scenarios are dropped, increasing the land available for utility solar by threefold.
• Figure 3: Internal dynamics between wind power and hydrogen on the British Isles (a) and the effects of British wind power on continental renewable investment (b). The left panel shows that low investment in wind power on the British Isles (of at least 12.3 bn EUR as by \ref{['fig:investment-ranges']}) is connected with low investment in hydrogen infrastructure, as sufficient affordable electricity for green hydrogen is lacking. The annotated 2030 target represents the sum of UK britishgovernment-2022 and Irish irishgovernment-2023 offshore wind goals amounting to 50 GW and 5 GW respectively. Existing onshore and offshore capacities (at the end of 2023) are retrieved from IRENAirena-2024 and sum to 20.3 GW of onshore wind and 14.8 GW of offshore wind. In both cases, capacities in GW are converted to annualised investment in bn EUR using the capital cost assumptions also used elsewhere in the model. The right panel shows the significant impact that British onshore wind has on robust invest levels in renewables in continental Europe. Renewable investment outside the British Isles is decomposed into onshore + offshore wind ($x$-axis) and solar ($y$-axis); the full space of robust solutions is shaded in turquoise. The dashed lines show how this design space changes depending on onshore wind investment on the British Isles. Very low onshore wind investment in the British Isles reduces the overall design flexibility of the rest of the system: the 2 bn EUR level shows a much smaller space. At the highest levels, on the other hand, the maximum viable wind investment in continental Europe is reduced. Supplementary Figures S12--14 are versions of this figure without the restrictive solar land-use scenarios, as well as including scenarios with limited onshore wind power land-use.
• Figure 4: Left, the robust design space for solar and onshore wind inside Germany, with two particular designs / points marked as blue and yellow dots. Current (April 2024) arbeitsgruppeerneuerbareenergien-statistik-2024 and targeted capacities for 2040 germangovernment-2023 are marked with crosses; these are converted from GW to bn EUR annualised investment using the technology-specific capital costs also used elsewhere in the model. Right, robust investment ranges for remaining technologies inside Germany and all technologies under consideration outside Germany, at the investment levels in onshore wind and solar inside Germany given by the two points on the left. Blue bars correspond to the blue point and likewise for yellow. The ranges are expected to be of similar relative proportions but wider or narrower with higher or lower slack levels, respectively. Observe that the investment ranges for the blue point, being located closer to the boundary of the robust design space, are significantly smaller (as in \ref{['fig:regional-impact-europe']}). German as well as European policymakers have an interest in ensuring large near-optimal feasible investment ranges (i.e. a large near-optimal space), as this translates to greater design robustness. Supplementary Figures S15--17 are versions of this figure without the restrictive solar land-use scenarios, as well as including scenarios with limited onshore wind power land-use; both Germany and the rest of Europe have wider ranges of options when more solar can be built.
• Figure 5: Effects of regional investment decisions on the design flexibility of the rest of the European system. The calculation of the flexibility indicator (where 1 is maximally flexible) that is plotted is explained in the Methods. The levels of regional investment leading to maximal outside flexibility are marked with black vertical bars. The outer limits of each bar represent the minimum and maximum robust regional investment levels for each technology. For instance, investment in solar in Poland and the Baltic countries has a positive effect on design flexibility in the rest of the system from 0 and up to about 14 bn EUR, with only a marginal decrease in flexibility upon further investment. The blue and yellow dots in \ref{['fig:DE_triple-plot']} are also examples of designs with low and high flexibility indicators, respectively. Robust onshore wind investments in the Nordics are cut off at 130 bn EUR (\ref{['fig:overview-onwind-solar']}(e)). While robust investment ranges widen with increasing slack levels, the investment levels maximising system-wide flexibility are not expected to change significantly under alternative slack levels. Supplementary Figure S18 shows the "converse" to this figure, with inside and outside dimensions swapped. Supplementary Figures S19 and S20 are more detailed versions of this figure and its converse, showing the derivative of the mean design width (proportional to the derivative of the flexibility indicator). See also Supplementary Note S5 for additional details on methodology.