Seasonal optimal mix of wind and solar power in a future, highly renewable Europe

Abstract

The renewable power generation aggregated across Europe exhibits strong seasonal behaviors. Wind power generation is much stronger in winter than in summer. The opposite is true for solar power generation. In a future Europe with a very high share of renewable power generation those two opposite behaviors are able to counterbalance each other to a certain extent to follow the seasonal load curve. The best point of counterbalancing represents the seasonal optimal mix between wind and solar power generation. It leads to a pronounced minimum in required stored energy. For a 100% renewable Europe the seasonal optimal mix becomes 55% wind and 45% solar power generation. For less than 100% renewable scenarios the fraction of wind power generation increases and that of solar power generation decreases.

Full text

Seasonal optimal mix of wind and solar power in a future,
highly renewable Europe
Dominik Heide
a
, Lueder von Bremen
b
, Martin Greiner
c
,
d
,
*
, Clemens Hoffmann
c
,
Markus Speckmann
e
, Stefan Bofinger
e
a
Frankfurt Institute for Advanced Studies (FIAS) and Frankfurt International Graduate School for Science, Johann Wolfgang Goethe Universität,
Ruth-Moufang-Straße 1, D-60438 Frankfurt am Main, Germany
b
ForWind eCenter for Wind Energy Research, University of Oldenburg, Marie-Curie-Str. 1, D-26129 Oldenburg, Germany
c
Corporate Research and Technology, Siemens AG, D-81730 München, Germany
d
Aarhus School of Engineering and Institute of Mathematical Sciences, Aarhus University, Ny Munkegade 118, 8000 Aarhus C, Denmark
e
Fraunhofer Institut für Windenergie und Energiesystemtechnik (IWES), Königstor 59, D-34119 Kassel, Germany
article info
Article history:
Received 13 January 2010
Accepted 14 March 2010
Available online xxx
Keywords:
Wind power generation
Solar power generation
Large-scale integration
Power-supply system design
Storage
abstract
The renewable power generation aggregated across Europe exhibits strong seasonal behaviors. Wind
power generation is much stronger in winter than in summer. The opposite is true for solar power
generation. In a future Europe with a very high share of renewable power generation those two opposite
behaviors are able to counterbalance each other to a certain extent to follow the seasonal load curve. The
best point of counterbalancing represents the seasonal optimal mix between wind and solar power
generation. It leads to a pronounced minimum in required stored energy. For a 100% renewable Europe the
seasonal optimal mix becomes 55% wind and 45% solar power generation. For less than 100% renewable
scenarios the fraction of wind power generation increases and that of solar power generation decreases.
Ó2010 Elsevier Ltd. All rights reserved.
1. Introduction
We investigate the design of a future European power supply
system based on a very high share of Renewables. Some key ques-
tions are: How much wind, solar, hydro and geothermal power is
good for Europe? Is there an optimal mix between them? How much
storageand balancing is needed? Howmuch transmission is needed?
Is it better to design the future power-supply system for one Europe
or for several separate regions?
These days wind power has emerged as the most dominant
contributor torenewablepower generation, withpotentialfor more in
the future. A straightforward and simple answer to the first question
has been pointed out [1,2]: there are enough wind resources around
the globeto supplyall continentswith only windpower. However, due
to its weather-driven fluctuations wind power generation is not
a simple stand-alonesolution. An enormous amountof balancing and
storage will be needed on top of this wind-only scenario.
The balancing and storage needs are already now very impor-
tant issues and will increase even more in the near future, when
wind power is totaling up to twenty percent of the overall European
power generation. Currently, for such a relatively small share, the
weather-driven wind power fluctuations are dominated by time
scales of the order one day and below. The respective spatial scales
range from the level of a single transformer to the level of inter-
connected regions. This paves the way for regional smart grids,
which will integrate short-term wind power fluctuations with
market-driven balancing and load flexibility.
In a renewable future well beyond 2020 the share of wind power
generation may well increase beyond fifty percent [2]. For such
a large amount, the spatial and temporal scales that have to be
looked at increase substantially. Regions with a momentary excess
of wind power will try to export it, whereas deficit regions are
depending on import. This spatial horizon beyond regions and
countries helps to smoothen short-term wind power fluctuations
and to reduce the short-term balancing and storage needs. In [3] it
has been shown that in a 100% wind-only scenario the need for
stored energy due to the mismatch between wind generation and
load is 41% lower in a European-wide smoothing of fluctuations
compared to selfish balancing in several regions in Europe.
Besides the short-term time scales ranging from minute to a few
days, theweather follows a distinct seasonal time scale. Winds across
Europeare stronger duringwinter than in summer. As a consequence,
*Corresponding author.
E-mail address: greiner@imf.au.dk (M. Greiner).
Contents lists available at ScienceDirect
Renewable Energy
journal homepage: www.elsevier.com/locate/renene
ARTICLE IN PRESS
0960-1481/$ esee front matter Ó2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2010.03.012
Renewable Energy xxx (2010) 1e7
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind and solarpower in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

the wind power aggregated over all of Europe is larger in winter than
in summer. This is shown in Fig. 1. In fact, the winter maximum is
about double the summer minimum.
If wind were the only power source in a fictitious future Europe,
then the seasonal wind power curve has to be directly compared to
the European load curve. This is also illustrated in Fig.1. In this 100%
wind-only scenario the yearly average of wind power generation
and load is the same. However, the seasonal behavior is different.
The seasonal load curve also comes with a maximum in winter and
a minimum in summer, but the seasonal variation strength is much
smaller than for the wind power generation. As a consequence an
enormous amount of stored and balancing energy is required. Over
summer the storage and balancing plants have to feed the deficit.
During winter the large wind power excess is put into the storage.
Like wind, also the solar community has its own solar-only
answer to the first question of the first paragraph [2]. If solar were
the only power source in another fictitious European future, then
the seasonal generation curve would look like the orange one in
Fig. 1. The solar power generation is much larger during summer
than in winter. Since it anticorrelates with the seasonal load curve,
a 100% solar-only scenario will lead to even larger seasonal storage
and balancing needs than for the wind-only case.
Let us summarize Fig. 1 in another way. For Europe the seasonal
wind power generation nicely correlates with the seasonal load
behavior. The seasonal solar power generation anticorrelates with
the seasonal load behavior. The seasonal wind and solar power
variation strengths are roughly the same. Both are significantly
larger than for the seasonal load.
When listening to these facts set by weather-driven mother
nature, an idea is created immediately. Future Europe is able to
counterbalance seasonal wind with solar power generation! Their
share should be almost the same, with a small extra contribution
from wind power due to its seasonal correlation with the load. Fig. 2
takes 60% from the wind curve and 40% from thesolarcurve of Fig. 1.
The resulting curve is able to nicely follow the seasonal loadcurve. It
is expected that this optimal mix brings seasonal storage and
balancing needs to a minimum.
In this paper we will further quantify the seasonal optimal mix
between wind and solar power generation in Europe, and the
resulting seasonal storage needs. Due to the expected dominance of
wind and solar power, all other renewable sources are neglected for
the moment. Section 2focuses on the European 100% wind-plus-
solar-only scenario. Section 3generalizes to transitional scenarios,
where wind-plus-solar power generation contribute less than the
load demand and where the rest is coming from fossil and nuclear
power. The conclusion is given in Section 4. The Appendix describes
the weather-driven time series modeling of the wind and solar
power generation and the estimation of the load curve across all of
Europe.
This Paper is the first within a series of three. The two followups
address the remaining questions of the first paragraph. They focus
on a detailed analysis of the balancing and transmission needs in
a future Europe with a very high share of wind and solar power
generation [4,5].
2. The 100% wind-plus-solar-only scenario
Based on seasonal time series such as shown in Figs. 1 and 2 it is
straightforward to quantify a seasonal optimal mix between wind
and solar power generation in a 100% wind-plus-solar-only
scenario for a future Europe. Key to such quantifications is the
mismatch energy
D
ðtÞ¼aWðtÞ
hWiþbSðtÞ
hSiLðtÞ
hLi:(1)
W(t) represents the total European wind power generation during
month t,andhWiits average over all 96 months contained in the
eight-years-long time series. S(t)andL(t)aretherespectivesolar
powerand load timeseries. The coefficients a¼hWi/hLiand b¼hSi/hLi
tell how much of the load is on average covered by wind and solar
power generation. For the 100% wind-plus-solar-only scenario these
coefficients are constrained to aþb¼1.
Afirst approach to quantify the seasonal optimal mix is to find
the minimum of the standard deviation
s
D
¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
D
D
2
Eh
D
i
2
r(2)
of the mismatch energy as a function of a¼1b. Since any
mismatch in the system requires balancing and the use of stored
energy,
s
D
can be regarded as a simple measure for balancing costs
0.6 0.8 1.0 1.2 1.4 1.6
time
normalized power
2000 2002 2004 2006 2008
Fig. 1. Normalized wind power generation (blue), solar power generation (orange) and
load (red) time series aggregated over Europe. Each series is shown in one-month
resolution and is normalized to its 8 years average. More details on the calculation of all
three seasonal curves are given in the Appendix. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article).
0.8 0.9 1.0 1.1 1.2 1.3
time
normalized power
2000 2002 2004 2006 2008
Fig. 2. Same as Fig. 1, only that the wind and solar power generation time series are
combined with a 60%/40% weighting (green). (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article).
D. Heide et al. / Renewable Energy xxx (2010) 1e72
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind andsolar power in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

[6]. The standard deviation is shown in Fig. 3a. Its minimum comes
at a¼0.62.
Another approach to the seasonal optimal mix constructs
a simple storage model out of the mismatch energy [1]:
HðtÞ¼Hðt1Þþ
h
in
D
ðtÞif
D
ðtÞ0;
h
1
out
D
ðtÞif
D
ðtÞ<0:(3)
Whenever the mismatch is positive, the surplus generation is
stored with efficiency
h
in
. In case of a negative mismatch the
generation deficit is taken out of the storage with efficiency
h
out
.
The time series H(t) describes the filling level of the storage. Its
maximum and minimum determines the maximum required
stored energy
H
max
¼max
t
HðtÞmin
t
HðtÞ:(4)
It is equivalent to the size of the storage capacity. This quantity is
shown in Fig. 3b as a function of the wind fraction a¼1b.Itcomes
with a rather flat minimum at a¼0.47. The storage efficiencies have
been set to
h
in
¼
h
out
¼1. Also shown in Fig. 3b is the 90% quantile Q
(95%) Q(5%) of the stored energy, which is determined from the
distribution p(H)andR
Q
0
pðHÞdH¼0:95 (0.05). This second variant
of the required stored energy reveals a pronounced minimum at
a¼0.57.
The used idealized storage efficiencies
h
in
¼
h
out
¼1 are not
realistic. Pumped hydro has
h
in
¼
h
out
¼0.9 and hydrogen storage has
h
in
¼
h
out
¼0.6. Since efficiencies smaller than one lead to storage
losses, the wind and solar power generation has to be increased in
order to compensate for the losses. The surplus generation factor
g
¼aþb>1 is determined from the requirement that the storage
level H(t¼8y) ¼H(t¼0) reached after 8 years is equal to the initial
storagelevel. It is important to notethatdue to the storage losses
g
is
depending on the time resolution. This is why for Fig. 4 we switch
from monthly to hourly wind, solar and load time series. Fig. 4a
shows that the smaller the storage efficiencies turn out to be the
larger the maximum stored energy becomes. However, the location
of the seasonal optimal mix does not change. The surplus generation
factor is illustrated in Fig.4b. For a¼bit amounts to
g
¼1.05 (pumped
hydro) and 1.28 (hydrogen).
0.0 0.2 0.4 0.6 0.8 1.0
a = 1−b
σΔ
0246810
0.0 0.2 0.4 0.6 0.8 1.0
a = 1−b
stored energy
1 1.5 2 2.5 3
max
90% quantile
Fig. 3. (a) Standard deviation (2) of the mismatch energy (1) and (b) stored energy
(3) as a function of the wind fraction a¼1bin a 100% wind-plus-solar-only scenario
for a future Europe based on monthly time resolution. For the storage energy its
maximum (black) and 90% quantile (red) is shown. The unit of the stored energy is in
average monthly load. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article).
a/(a+b)
stored energy
no loss
Hydro
H2
0.0 0.2 0.4 0.6 0.8 1.0
0.0 0.2 0.4 0.6 0.8 1.0
a/(a+b)
γ
1.5 22.5 3 3.5
1 1.2 1.4 1.6 1.8
Hydro
H2
Fig. 4. (a) Comparison of the maximum stored energies for an idealized storage
(black), pumped hydro (green) and hydrogen storage (red), derived from hourly wind,
solar and load time series. The unit of the stored energy is in average monthly load.
Within numerical uncertainties the black curve is identical to the black curve of Fig. 3b
based on the monthly time series. (b) Surplus generation of wind and solar power
needed to compensate the storage losses. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article).
D. Heide et al. / Renewable Energy xxx (2010) 1e73
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind and solarpower in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

The different approaches based on the standard deviation of the
mismatch energy and the stored energy lead to a seasonal optimal
mix of 50e60% wind and 50e40% solar power generation. These
results agree nicely with the intuition obtained from the intro-
ductory Figs. 1 and 2.
From these findings on the seasonal optimal mix a few more
conclusions can be given for a 100% wind-plus-solar-only scenario in
a future Europe. Just for demonstration we freeze the average yearly
load for all of Europe to its 2007 value, which is 3240TWh. 55% of this
makes 1780TWh and requires 670 GW of installed wind power
capacity; here the wind load factor 0.30 has been used, which has
been directly determinedfrom the used weather data.The remaining
45% make 1460TWh and require 810 GW of installed solar photo-
voltaic power; here the PV load factor 0.21 has been used, which
again has been directly determined from the used weather data.
670 GW of installed wind power capacity across Europe translate
into 335 000 2 MW turbines, or 135 000 5 MW turbines, or 4000
wind farms of the size of the first offshore wind farms Horns Rev I
and Nysted. As a rule of thumb [7], one MW of installed capacity
requires 0.07 km
2
onshore and 0.11 km
2
offshore, respectively. This
translates 670 GW into 50,000 km
2
onshore or 75,000 km
2
offshore.
For comparison, Denmark has an area of 43,000 km
2
.
The spatial and temporal mean global radiation 169 W/m
2
is
computed from the weather data and translates 1460TWh-
producing 810 GW-installed solar photovoltaic power into a PV-
panel area of 5000 km
2
. For comparison, Germany has the potential
to cover 1330 km
2
of roofs with ideal slopes and direction [8].
As can be seen in Figs. 3b and 4a, the required maximum stored
energy has to be 1.5 (without storage losses) and 1.8 (for hydrogen
storage) times the monthly load and amounts to400 and 480 TWh,
respectively. These are very large numbers. They will even double
once future Europe decides to switch to a wind-only or a solar-only
scenario.
Currently, Germany has about 190 GWh of pumped hydro
facilities in operation, with only little room for more. The exact
amount of pumped hydro across all of Europe is not known to us.
Even if it is a factor of ten more, still two orders of magnitude are
missing to reach the required 400e480 TWh.
New forms of bulk storage like hydrogen will be needed. Its
storage density is 187 kWh/m
3
, assuming a pressure difference of
120 bar and an efficiency of 0.4. This translates 400e480 TWh
stored energy into a volume of 2.2e2.6 km
3
, which does not appear
to be completely out of reach.
3. Transitional scenarios with wind, solar and fossil-nuclear
power
The investigations on the 100% wind-plus-solar-only scenario
can be extended to transitional scenarios by modifying the
mismatch energy (1) to
D
ðtÞ¼aWðtÞ
hWiþbSðtÞ
hSiþcFðtÞ
hFiLðtÞ
hLi:(5)
F(t)¼hFirepresents fossil-nuclear power generation and is
assumed to be time-independent. It may even include a contribu-
tion from geothermal power.
The three coefficients aþbþc¼1 add up to one and match the
average load. The choice a¼0.27, b¼0.00, c¼0.73 leads to Fig. 5.
The seasonal power generation curve follows the seasonal load
curve more closely than for the a¼0.60, b¼0.40, c¼0.00 example
shown in Fig. 2. This indicates already, that as long as a fraction of
fossil-nuclear power remains in the generation mix the need for
stored energy will be smaller than for the 100% wind-plus-solar-
only scenario.
Fig. 6 shows the required maximum stored energy (4), which has
been deduced from (3) with
h
in
¼
h
out
¼1 and (5). It is a function of
the two independent coefficients cand a/(aþb). The smallest stored
energy is obtained for a¼0.27, b¼0.00, c¼0.73, which are the
values used for Fig. 5. For large fossil-nuclear fractions 1 c0.73
the stored energy reaches a minimum when only wind power
generation covers the remaining fraction a¼1c. This is because the
seasonal wind power generation curve correlates with the seasonal
load curve; confer again Fig. 1. Due to its anticorrelation it is not
favorable to include solar power into the fossil-nuclear-dominated
generation mix.
Solar power is needed once the fossil-nuclear power generation
is reduced below c<0.73. The seasonal solar power generation
then has to counterbalance the seasonal wind power generation.
Otherwise the absolute seasonal wind power variation would
become larger than the absolute seasonal load variation. The
0.85 0.95 1.05 1.15
time
normalized power
2000 2002 2004 2006 2008
Fig. 5. A mix of 27% wind, 0% solar and 73% fossil-nuclear power generation (green) is
also able to follow the (red) seasonal load curve. See Figs. 1 and 2 for comparison.
(For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article).
Fig. 6. Maximum stored energy (4) required for all of Europe as a function of the two
independent coefficients cand a/(aþb). The dashed curve represents the seasonal
optimal mix between wind and solar power generation as a function of the remaining
fossil-nuclear power generation. The unit of the contours is given in average monthly
load over Europe.
D. Heide et al. / Renewable Energy xxx (2010) 1e74
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind andsolar power in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

dashed curve in Fig. 6 follows the bottom of the storage valley and
represents the seasonal optimal mix between wind and solar
power generation as a function of the remaining fossil-nuclear
power generation. In the limit c/0 the wind and solar coefficients
azbz0.5 become about the same. Obviously, this result agrees
with the earlier result obtained in Fig. 3b.
Let us follow the dashed optimal mix-curve once more, from
right to left. At c¼1 the required maximum stored energy amounts
to 1.36 times the average monthly load. From c¼1downtoc¼0.73
the required maximum stored energy decreases down to 1.08 times
the average monthly load. From c¼0.73 to c¼0 the required
maximum stored energy increases again and reaches 1.44 times the
average monthly load at c¼0.
We close this Section with an additional remark. If the seasonal
load curve had come with a maximum in summer and a minimum
in winter, then the optimal mix curve would have been different.
For a large fossil-nuclear fraction cclose to one it would have been
solar only with b¼1cand a¼0. Wind power generation will be
necessary to counterbalance the reduced solar generation in winter
once cbecomes smaller. This might be of relevance for some large
countries outside Europe.
4. Conclusions
Besides short-term fluctuations, wind and solar power genera-
tion across Europe follow the seasonal cycle of the weather. Wind
power generation in winter is much stronger than during summer.
For solar power generation the summer season produces much
larger yields than during winter. In this way mother nature deter-
mines how to design a future European power supply system based
on a very high share of renewables. When mixed together in
aspecific ratio, the opposite strong seasonal behaviors of wind and
solar power generation almost cancel each other and follow the
weaker seasonal load behavior. For a European 100% wind-plus-
solar-only scenario, this seasonal optimal mix is found to be 55%
wind and 45% solar power generation. Compared to other scenarios
like wind-only or solar-only, the optimal mix reduces the need for
stored energy by a factor of two. The reduced stored energy for all of
Europe amounts to 1.5e1.8 times its monthly load. For transitional
scenarios with a fraction of fossil-nuclear power generation left in
the system, the optimal mix between renewables is shifted in favor
of wind power generation. This is because of the seasonal correlation
between wind power generation and load across Europe.
We have addressed only a few of the key questions raised in the
first paragraph of the Introduction. Answers to the other questions
will be given in two subsequent publications [4,5].Withhourlydata
at hand (see Appendix), they will focus on the balancing-power and
power-transmission needs across a future, highly renewable Europe.
5. Appendix: Modeling of wind-, solar power generation and
loads
Key to the modeling of wind and solar power generation is
a large weather data set with good spatial and temporal resolution
all over Europe. Its convolution with future-projected wind and
solar power capacities reveals how much wind and solar power is
Fig. 7. Expected (a) wind power and (b) solar photovoltaics power capacities [GW] per
grid cell across Europe in 2020. The spatial grid-cell resolution of 47 48 km
2
has been
adapted to theweather data. For a better visualization capacities larger than 0.73 GW for
wind and 0.50 GW for PV are indicated in dark red. (For interpretation of the references
to colour in this figure legend, the reader is referred to the web version of this article).
Fig. 8. Average annual load [TWh] per grid cell in the 50 coarse-grained onshore
regions.
D. Heide et al. / Renewable Energy xxx (2010) 1e75
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind and solarpower in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

generated when and where across Europe. The following subsec-
tions will explain the details. The load modeling is described in the
last subsection.
5.1 Weather data
Weather data for all of Europe is available from various sources
with different spatial and temporal resolutions. For our purposes
three selection criteria have been important: (i) In order to resolve
the passing of synoptic systems related to high winds and opaque
clouds a spatial resolution of atleast 50 50 km
2
is required. (ii) The
correct modeling of intra-day solar and wind power ramps require
a good time resolution of at least 1 h. (iii) In order to gain repre-
sentative and significant statistics covering all possible seasonal and
extreme weather situations a rather long time window is required,
ranging over a couple of years.
These criteria have been met by the private weather service
provider WEPROG (Weather & Wind Energy Prognosis) [9]. With
regional models it downscales medium-resolved analysis data from
the US Weather Service NCEP (National Center for Environmental
Prediction) [10] down to 47 48 km
2
spatial and 1 h time resolu-
tion over an eight-years period (2000e2007).
This high-resolution data provides direct information on the
wind speed and direction 100 m above ground. The solar global
radiation isnot a standard output, but can be computed directly from
the data on the net short wave radiation at the surface, the total
cloud cover, and a standard cloud and surface albedo.
5.2 Wind and solar power capacities
The national 2020 targets serve as guidance for a rough distri-
bution of wind and photovoltaic capacities in Europe. Fig. 7 illus-
trates the installed wind power and solar photovoltaics power
capacities across Europe expected in 2020. They total to 227 and
68 GW, respectively. 66 GW of wind power is assumed to be
installed offshore. The subsequent finer distribution within each
country onto the grid cells of the weather data is done empirically,
giving more capacity to those grid cells with large average wind
speed and large average global radiation, respectively.
5.3 Wind and solar power generation
The conversion of hourly WEPROG wind speeds into wind power
at each grid cell was done using typical wind power curves at 100 m
hub height. Different power curves have been assigned for onshore
and offshore grid points. Losses due to wake effects have been
modeled explicitly foroffshore grid points by assuming a park layout
of 7 7 turbines in offshore wind farms. Additional 7% losses have
been introduced due to electrical losses and turbine non-availability.
The same 7% of losses have also been applied to onshore grid points.
The turbine cut-off due to extreme winds is empirically parameter-
ized by an additional modification of the power curve, which mimics
the gradual power-lowering-behavior of wind turbines with storm-
control.
The solar photovoltaic power generation within the grid cells
has been calculated based on the available meteorological data
(global radiation, air temperature), assumptions on the character-
istics of the photovoltaic plants (tilt angle, orientation, fixed or with
solar tracker) and the geographical coordinate of the grid cell
considered. A mix of different photovoltaic plant technologies was
considered for each grid cell.
This convolution of the weather data with the wind and solar
power capacities produces spatio-temporal power generation
patterns across Europe. These patterns are important for the calcu-
lation of power flows [5]. For the current paper we discard the spatial
part and assume Europe to be one big copper plate. For each hour the
total wind as well as solar power generation is integrated over all grid
cells.
ThebluetimeseriesinFig. 1 shows the monthly wind energy
integral over all of Europe. It is normalized to the monthly average
over the 8 years period. Due to this normalization, also the same time
series would have been obtained, if the wind power capacities of
Fig. 7a had been upscaled. An upscaling factor of 5.2 turns 227 GW
installed wind power capacity into 1180 GW. With 2650 full-load
hours per year directly calculated from the used weather data, this
then produces 3130TWh per year, exactly corresponding to the
average annual European load between 2000 and 2007.
The orange time series in Fig. 1 shows the normalized monthly
solar photovoltaics energy integral over all of Europe. Analogous to
the arguments given for wind, also this time series does not change
upon an upscaling of the solar photovoltaics power capacities of
Fig. 7b. For a 100% solar photovoltaics power scenario the upscaling
factor would be 25.5, turning 68 GW installed solar power capacity
into 1730 GW. With 1800 full-load hours this then produces the
load-equivalent 3130TWh per year. The orange solar time series
may change in a certain direction, once solar thermal power
complements photovoltaics. Since those plants would be installed
mostly in southern Europe, the relative difference between the
seasonal summer maximum and winter minimum is expected to
become smaller to some extent.
5.4 Load modeling
It is impossible to retrieve the load profiles on the spatial
47 48 km
2
grid-cell scale set by the weather data. However,
a coarser resolution is fine for our purposes. For almost all European
countries the load profiles have been downloaded either from the
UCTE-homepage [11] or from the national transmission providers. At
least for the two recent years those have an hourly resolution. For the
remaining six years they have been replicated with the known rela-
tive annual electric power consumption; special care was given to
a proper handling of the weekend effect.
Some countries, especially the larger ones, come with a very
large average load. Those have been further subdivided into regions,
with some spatial correlation to the territories of the respective
network transmission providers. The regional load profiles have
been obtained from the country profiles with a multiplicative factor
obtained from a linear regression between the annual electric
power consumption on the one hand and population and gross
domestic product on the other hand.
Fig. 8 shows the average annual load of the 50 onshore regions
during the years 2000e2007. Offshore regions come with no load
and are not shown. The sum over all regions totals to 3130TWh
annual consumption. Its seasonal dependence is shown in Fig. 1.
Note, that for the seasonal storage calculations in Sections 2 and 3
the European load curve has been detrended.
References
[1] Lu X, McElroy M, Kiviluoma J. Global potential for wind-generated electricity.
Proc Natl Acad Sci USA 2009;106:10933e8.
[2] Jacobson M, Delucchi M. A path to sustainable energy by 2030. Sci Am; Nov
2009:58e65.
[3] von Bremen L, Greiner M, Knorr K, Hoffmann C, Bofinger S, Lange B. A full
renewable power supply scenario for Europe - the weather determines
storage and transport, In: Proc. of the 8th Int. workshop on large-scale inte-
gration of wind power into power systems as well as on transmission
networks for offshore wind farms, Bremen, Germany; October 2009.
[4] Heide D, von Bremen L, Greiner M, Hoffmann C. Storage and balancing needs in
a future Europe with only wind and solar power generation, in preparation.
[5] Heide D, Knorr K, von Bremen L, Greiner M, Hoffmann C. Transmission needs
for a cooperative future Europe with a very high share of wind and solar
power generation, in preparation.
D. Heide et al. / Renewable Energy xxx (2010) 1e76
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind andsolar power in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012

[6] von Bremen L. Large-scale variability of weather-dependent renewable
energy sources. NATO science for peace and security series -C: environmental
security. In: Troccoli A, editor. Management of weather and climate risk in the
energy industry. Springer, ISBN 978-90-481-3691-9; 2009.
[7] DENA, Energiewirtschaftliche Planung für die Netzintegration von Windenergie
in Deutschland an Landund Offshore bis zum Jahr2020, Köln [accessed 24.02.05].
[8] Quaschning V. Systemtechnik einer klimaverträglichen Elektrizitätsversor-
gung in Deutschland für das 21.Jahrhundert, Fortschritt Berichte VDI, Reihe 6,
Nr. 437. Düsseldorf: VDI Verlag; 2000.
[9] http://www.weprog.de/.
[10] http://www.ncep.noaa.gov/.
[11] http://www.ucte.org/resources/dataportal/consumption/.
D. Heide et al. / Renewable Energy xxx (2010) 1e77
ARTICLE IN PRESS
Please cite this article in press as: Heide D, et al., Seasonal optimal mix of wind and solarpower in a future,highly renewable Europe, Renewable
Energy (2010), doi:10.1016/j.renene.2010.03.012