Summary

Figure 3.28 summarizes the results obtained in Figure 3.27 and Table 3.2 in a different way. For fixeda = 0.6 it compares theγ dependence of the smaller daily-based with the larger hourly-based storage energies, balancing energies and balancing quantiles.

With this separation of time scales we anticipate a separation into a short-term and long-term storage.

The long-term storage takes care of the daily mismatch. Over the day its hourly contribution would be more or less constant and sum up to the daily mismatch. Its storage/balancing energy and power quantiles would correspond to those based on the daily mismatches. With another look into Table 3.2 at for example γ = 1.5 and a= 0.6 the required numbers would be E_H(η = 1) = 0.004 and Q_B(q = 0.99) = 0.246 for ideal round-trip storage, which, given the annual European load of 2007, translate into E_H(η = 1) = 15TWh andQ_B(q= 0.99) = 90GW. For hydrogen storage the respective numbers are E_H(η = 0.6) = 0.010 = 35TWh and Q_B(q= 0.99) = 0.246 = 90GW. For one-way storage reservoirs like storage lakes the respective numbers are EB= 0.009 = 30TWh and Q_B(q = 0.99) = 0.246 = 90GW.

The short-term storage takes care of the hourly mismatch around the daily mismatch.

Roughly, its required balancing power corresponds to the difference between the quantile based on the hourly mismatch and the quantile based on the daily mismatch. Again for γ = 1.5 anda = 0.6, this results in Q^hourly_B (q= 0.99)−Q^daily_B (q= 0.99) = 0.40 = 140GW.

A safe upper bound for the required storage energy for a roundtrip storage would then be E_H^short−term = 140GW×12h = 1.68TWh <2TWh. The multiplication with 12h is due to the night hours, where balancing is needed; see again Figure 3.25. Candidates for such a short-term storage would be pumped hydro, electric cars, compressed air, and any combination thereof.

3.8. Summary

20% solar power generation, for balancing power it is 90% wind and 10% solar power generation. Furthermore, all optimal mixes turn out to be more or less independent from the amount of excess power generation. For transitional scenarios with a high fraction of fossil-nuclear power generation left in the system, the share of wind power generation increases when minimizing the required storage size and decreases when minmizing balancing energy compared to the only renewable scenario. Note, that the different optimal mixes are mainly caused by the intra-day mismatch dynamics between wind plus solar power generation and load. Once the intra-day time scales are neglected, the optimal mixes for storage energy, balancing energy and balancing power would become identical at around 60% wind and 40% solar power generation. The storage energy for all of Europe amounts to around 10% of the average annual load. Compared to other scenarios like wind-only or solar-only, the optimal mix reduces the need for storage energy by a factor of two.

With no excess wind plus solar power generation, the required storage needs turn out to be very large. However, they decrease very fast with the introduction of excess power generation. In the following we list the concrete numbers for the required storage needs, given the European consumption load of 3240 TWh for 2007; consult again with Table 3.1. With the objective to minimize the storage energy, the required needs for roundtrip storage with ideal efficienciesη_in =η_out = 1 amount toE_H = 320 TWh storage energy and QB(q = 0.99) = 300 GW balancing power for γ = 1 anda = 0.6. For excess generation withγ = 1.5 anda = 0.6 the numbers reduce toE_H = 16 TWh storage energy and Q_B(q= 0.99) = 240 GW. However, the installed wind and solar-photovoltaic power capacities across Europe would each increase from 750 to 1100 GW. For comparison, hydrogen storage with non-ideal efficiencies η_in =η_out = 0.6 would require around 40 TWh and 220 GW for minimum storage energy and non-minimum balancing power at γ = 1.5 and a= 0.7 with installed 1300 GW wind and 830 GW solar power capacities.

If we were to choose the other objective to minimize the balancing power, then the optimal a = 0.9 at γ = 1.5 leads to required 50 TWh (ideal) and 120 TWh (hydrogen) storage energies, and 195 GW balancing power, with installed 1650 GW wind and 275 GW solar power capacities.

For one-way balancing storage the two objectives to minimize either the balancing energy or to minimize the balancing power lead to quite similar results. The optimal share of generated wind power amounts to a≈0.8-0.9. At zero excess generation γ = 1 the required European balancing energy and balancing power result to be 500 TWh and 265 GW. They are reduced down to 160 TWh and 200 GW once the excess generation is increased to γ = 1.5.

The presented results demonstrate that excess wind and solar power generation will be one key to reduce the required storage needs for a fully renewable European power system. In fact, a fully renewable power system is only fully defined when including of a reasonable amount of excess generation. However, with a reasonable amount of excess wind and solar power generation alone the resulting storage energies, balancing energies and balancing powers are still very large. With 50% excess generation, roundtrip storage

still comes with a required storage energy of the order 20-50 TWh and balancing power of the order 200 GW. For one-way balancing storage it will be 160 TWh storage energy and 200 GW balancing power.

With the considered storage technologies, central storage facilities seem to be un-avoidable due to the high energy that needs to be storage in a storage only scenario.

These facilities are required to provide high capacities for balancing power and can be expected to cause large power flows. A paradigm change so that the load follows the intra-day generation could help to overcome these problems. A shift of consumption to day time and less load during night time would reduce both the required balancing power, see Figure 3.25, and the extreme transport flows, as will be shown in Section 6.1.

However, this approach highlights the need for accurate prediction of wind and solar power for at least 24 hours. Local, distributed storage facilities with limited storage energy also have a great potential to match the generation and consumption over a day.

A good understanding of possible future scenarios has to be the basis for deciding on efficient transitions from the existing power system towards a fully renewable scenario.

It has to be noted, that the two extreme scenarios, only storage or only balancing power, are discussed here. Both scenarios, were shown to be feasible. A mix of both approaches potentially further reduces the storage and the balancing needs.