Contribution of bulk energy storage to integrating …storage to integrating variable renewable energies in future European electricity systems Karl Anton Zach* and Hans Auer Remarkably

Advanced Review

Contribution of bulk energystorage to integrating variablerenewable energies in futureEuropean electricity systemsKarl Anton Zach* and Hans Auer

Remarkably higher expected shares of variable renewable energy sources forelectricity generation (RES-Electricity) than those available today will be a greatchallenge for the European power system. Bulk electricity storage technologies—that is, pumped hydro energy storage—are considered a key component whilefacing these future challenges. The conducted analysis is based on the modelingresults of the Blue storyline of the SUSPLAN project that is characterized by ahigh growth of RES-Electricity deployment. The RES-Electricity deploymentresults on a country level for the years 2030 and 2050, the age structure of theexisting thermal power plant portfolio, and hourly values of electricity demandand RES-Electricity generation are taken as input for a linear optimization algo-rithm, which is implemented to analyse the importance of bulk electricity storagetechnologies for integrating large amounts of variable RES-Electricity inEuropean regions. Two regions are analysed in detail: Central Western Europeand the Iberian Peninsula. The results show that due to the age-related phaseout of thermal power plants in the future, additional new power plant capacitieswill be needed in many European regions by the year 2030. Only some Europeanregions also have sufficient flexible generation capacity available to cover theresidual load in the long term. Furthermore, RES-Electricity feed-in will exceedelectricity demand for some time during the year 2050 in several regions. TheIberian Peninsula shows especially high RES-Electricity feed-in exceeding elec-tricity demand about half the time during the year 2050. This surplus RES-Electricity generation can be used for bulk electricity storage and/or for exportsto neighboring European regions. © 2016 John Wiley & Sons, Ltd

How to cite this article:WIREs Energy Environ 2016. doi: 10.1002/wene.195

INTRODUCTION

Remarkably higher expected shares of variablerenewable energy sources for electricity genera-

tion (RES-Electricity) than those available today willbe a great challenge for the European power systemthat is calling for more flexibility of back-up power

plants to cover the residual load and higher transmis-sion capacities to balance supply and demand inwider areas. During times when high RES-Electricityfeed-in and low electricity demand result in a nega-tive residual load (i.e. renewable surplus electricity)in the electricity system, bulk electricity storage tech-nologies (EST)—i.e., mainly pumped hydro energystorage (PHES) and compressed air energy storage(CAES)—are considered a key component to facethese future challenges. The long-term scenario ana-lyses of infrastructure integration of RES-Electricityperformed in this study take into account the interde-pendencies between regional and trans-national

*Correspondence to: [email protected]

Energy Economics Group, Institute of Energy Systems and Electri-cal Drives, Vienna University of Technology, Vienna, Austria

Conflict of interest: The authors have declared no conflicts of inter-est for this article.

© 2016 John Wiley & Sons, Ltd

energy systems and also the importance of bulk ESTsfor integrating large amounts of variable RES-Electricity in future electricity systems.

The conducted analysis is based on the model-ing results of long-term RES-Electricity deploymentin the Blue storyline of the SUSPLAN project. TheBlue storyline is characterized by a high growth ofRES-Electricity deployment and also a high growthof electricity demand. The RES-Electricity deploy-ment results on a country level for the years 2030and 2050, the age structure of the existingthermal power plant portfolio, and hourly values ofelectricity demand and RES-Electricity generation aretaken as input for a linear optimization algorithm,which is implemented to analyse interdependencies ofneighboring regions and the importance of bulk ESTsfor integrating large amounts of variable RES-Electricity (including mitigation of their effects). TwoEuropean regions are analysed in detail: CentralWestern Europe and the Iberian Peninsula, whichshow great differences concerning their future RES-Electricity penetration and thermal power plantportfolio.

ANALYSIS OF THE EUROPEANELECTRICITY MARKET REGIONS

The modeling approach for the analysis of the elec-tricity sector of the SUSPLAN project1 providesresults for RES-Electricity generation and installedcapacities as well as corresponding transmission gridinvestment costs and needs on a country and annuallevel2 for four different storylines until the year2050.3 Future European RES-Electricity deploymentsfor the different storylines were generated basedon modeling results derived from GreenNet,4 a least-cost RES-Electricity deployment simulation tool.GreenNet provides future scenarios on annualRES-Electricity capacity installations and electricitygeneration per country under a variety of differentpossible policy settings and constraints.

In this paper, a complementary analysis is con-ducted for one of these four storylines—the Bluestoryline—that is characterized by a high growth ofRES-Electricity deployment and also a high growthof electricity demand. The analysis provided hereinalso shows interdependencies of neighboring regions/countries and the importance of bulk EST (i.e. mainlyPHES and CAES) for integrating large amounts ofvariable RES-Electricity on the one hand and mitigat-ing their effects on the other hand. More precisely,(residual) electricity load duration curves are derivedand analysed for so-called ‘European electricity

market regions’ (EEMRs) on an hourly basis for theyears 2030 and 2050 for the Blue storyline of theSUSPLAN project.

For the consideration of bulk EST in the analy-sis, the assessment of the geographical allocation offuture potentials for EST in Europe was a precondi-tion. In the second step of the analysis, Europeancountries where clustered into nine different EEMRsaccording to physical constraints in the transmissiongrid. To derive the residual electricity load durationcurves of the different EEMRs in the years 2030 and2050 under consideration of the operation of thebulk EST of the respective region, a linear optimiza-tion algorithm was established in the General Alge-braic Modelling System (GAMS).5 The task andconstraints of this algorithm are given in the respec-tive section below.

Geographical Allocation of FuturePotentials for Bulk Electricity Storage inEuropeThe geographical allocation of future potentials forbulk EST in Europe is estimated based on the mostrelevant existing work and studies on a country levelas well as on European level:

• PLATTS—European electric power plantdatabase6

• ENTSO-E—System Adequacy Forecast7

• EURELECTRIC—Hydro in Europe report8

• National renewable energy action plans(NREAP)9

• Green-X database10

• JRC – PHES—potential for transformationfrom single dams11

Furthermore, internet research for gathering data ofutilities in the different countries was carried out inorder to get more detailed information of local (P)HES systems. The result of this potential estimationfor bulk EST in Europe is shown in Figure 1.

PHES is the most mature EST today; it is usedfor different functions in the power system andwidely applied in Europe.12,13 The ability of PHESplants to integrate (variable) RES-Electricity has beenproved in various studies.14–17

At present, only two commercial CAESsystems—the second major bulk EST—have beenbuilt worldwide (one existing and one advanced-adiabatic is under development in Germany).18,19

Due to a lack of data sources on future deployment

Advanced Review wires.wiley.com/energy


and potential of CAES systems, Figure 1 shows esti-mations on future potentials of (P)HES systems inEurope [including already existing (P)HES systems]only. However, CAES systems might also play animportant role in the future European electricity sys-tem, but the utilization potential of CAES systems isdependent on the availability of appropriate under-ground air storage capacities (especially salt caverns)and on further technological developments needed inorder to reach higher overall efficiencies and to elimi-nate the need for using additional fossil fuels (i.-e. advanced adiabatic CAES).20

It is important to note that the majority of theused sources for the estimation of the European (P)HES potential only provides scenario figures includ-ing currently operating, planned (P)HES, and (P)HESsystems under construction (see respective sources formore details). Therefore, the potential figures in ouranalysis represent economical rather than technicalpotential figures for (P)HES. The actual technicalpotential of (P)HES systems in the European coun-tries might be higher.

Figure 1 shows that the highest total potentialsfor (P)HES are located in parts of Scandinavia, Cen-tral, and Western Europe. Besides Norway and Swe-den, European countries in mountainous areas, such

as the Alps and the Pyrenees, have significant poten-tial for deployment of PHES systems. Luxembourg,Switzerland, and Austria have a particularly highPHES potential in relation to their land area. Coun-tries with a rather flat landscape like Denmark, Fin-land, Latvia, and the Netherlands have no existing(P)HES power plants and also no development plansfor new PHES systems at the moment (no data isavailable for Albania).

Norway, often referred to as the ‘green battery’of Europe, has almost half of Europe’s reservoircapacity because of the topographical advantage ofhigh mountainous lakes or reservoirs. Typically60–70% of Norway’s average annual hydropowergeneration (123 TWh) is produced by (conventional)HES power plants offering high flexibility in electric-ity generation. As the Norwegian electricity generat-ing system already comprises a lot of flexibilitythrough its hydropower plants, the construction ofnew PHES (or upgrading of existing HES) powerplants has so far not been economically profitable.This is also a reason why Norway currently has1336 MW of PHES systems only, occasionally used topump seasonal excess water into higher reservoirs.8

Other possible bulk ESTs like hydrogen storageand methane storage (i.e. power-to-gas) are not

(P)HES Potential 2050

0 GW

0–1 GW

1– 2 GW

2–5 GW

5–10 GW

10–15 GW

15–30 GW

(P)HES Potential 2050

0 kW/km2

0–10 kW/km2

10–20 kW/km2

20–40 kW/km2

40–80 kW/km2

80–160 kW/km2

160+ kW/km2

FIGURE 1 | (Pumped) hydro energy storage [(P)HES] potential in Europe until the year 2050: power capacity per country (GW) (a) and powercapacity per land area per country (kW/km2) (b).

WIREs Energy and Environment Contribution of bulk energy storage to integrating variable renewable energies


considered in the analysis. Additionally, these alter-native technologies proved to be economically infe-rior to PHES as a day storage plant.21 However,there might be a field of application for hydrogenand methane storages as seasonal storage in thefuture if their costs can be reduced considerably.

Clustering of European Countries intoNine European Electricity MarketRegionsDue to physical constraints in the European (cross-border) transmission grid, electricity transfer betweencountries is limited, with the consequence of differentwholesale electricity markets and prices. To take thisinto account in the analysis, European countries wereclustered into nine different electricity market regionsaccording to the different wholesale electricity marketplaces (as a consequence of physical constraints inthe transmission grid):

• Iberian Peninsula: Portugal and Spain

• Central Western Europe (CWE): France, theNetherlands, Luxembourg, Belgium, West Den-mark, Germany, Switzerland, and Austria

• Central Eastern Europe (CEE): Poland, CzechRepublic, Slovakia, and Hungary

• South-Eastern Europe (SEE): Romania, Bul-garia, and Greece

• Western Balkan: Slovenia, Serbia, Croatia, Bos-nia and Herzegovina, Republic of Macedonia,and Montenegro

• Italy

• The United Kingdom (UK) and Ireland

• Nordic Region: Norway, Sweden, Finland, andEast Denmark

• Baltic Region: Lithuania, Latvia, and Estonia

This clustering coincides with relevant EC documents(e.g. EC infrastructure package)22,23 and theEuropean Network of Transmission System Opera-tors for Electricity (ENTSO-E)’s ‘Ten Year NetworkDevelopment Plan’ (TYNDP)24 and is shown inFigure 2.

Integration of Surplus RES-ElectricityGeneration with Pumped Hydro SystemsBulk ESTs are expected to be one of the key enablingtechnologies for the integration of large amounts ofvariable RES-Electricity generation. In particular, the

ability to quickly discharge large amounts of storedelectricity and to store surplus RES-Electricity gener-ation especially in off-peak times can mitigatemany challenges that arise with high shares of varia-ble RES-Electricity generation in the electricitysystem.

Therefore, after derivation of the residualload curves in the analysed EEMR (CWE and theIberian Peninsula are shown in more detail) forthe years 2030 and 2050, the capability of PHES sys-tems to integrate surplus RES-Electricity generationand provide peak-load power in the regions wasanalysed in a subsequent step. For this, the individualresidual load curves were taken as input for asmall linear optimization algorithm, which waswritten in GAMS. The task of this algorithm is themaximization of the revenue of PHES systems in theEEMRs:

maxXt2T

RL tð Þ* PTurb tð Þ−PPump tð Þ� �( ); ð1Þ

RL(t) is the residual load in hour t; PTurb(t) is the tur-bine power of PHES system in hour t; and PPump(t) isthe pumping power of PHES system in hour t.

As the actual hourly electricity prices are notavailable for the years 2030 and 2050, the values ofthe hourly residual load are taken as proportionalindicators for the future electricity price; high resid-ual load values are equivalent to high electricityprices. Surplus RES-Electricity generation might espe-cially be available at a low cost or even negativeprices. The revenue is optimized under the followingconstraints:

forRL tð Þ ≥ 0 : RL tð Þ−PTurb tð Þ ≥ 0 and PPump tð Þ ≤RL tð Þ;ð2Þ

RL tð Þ−PTurb tð Þ +PPump tð Þ ≤RLmax−PmaxTurb; ð3ÞforRLmin +PmaxPump ≤ 0 : RL tð Þ−PTurb tð Þ +PPump tð Þ

≥RLmin +PmaxPump;

ð4Þ

RLmax, RLmin is the maximum and minimum resid-ual load; PmaxTurb is the maximum turbine power ofthe PHES system; and PmaxPump is the maximumpumping power of PHES system.

In case of a positive residual load, the con-straints in Eq. (2) assure that, on the one hand, thePHES system does not turbinate beyond zero (i.e. nonegative residual load through turbining) and, on theother hand, limits the pumping power to the residual



load in order to set the right price signals for theoptimization. Without this constraint, the algorithmwould also use full pumping power in case of a verylow residual load, leading to a considerable increasein the residual load and therefore electricity prices.The constraint limits this load and price increase inorder to reflect the market power of the PHES sys-tem. Equations (3) and (4) assure that the positiveand also negative peak-loads are maximally reducedby the PHES system. With these constraints, the opti-mization algorithm also guarantees a maximal inte-gration of excess RES-Electricity generation in theEEMRs.

Further constraints applied in the algorithm aretypical limits to the storage capacity of the PHESsystem:

• 80% cycle efficiency20

• Maximum energy storage capacity was definedper EEMR

• Start and end value of stored energy must bethe same

• Start-value for the energy stored is set to half ofthe maximum energy storage capacity

Furthermore, all PHES systems of an EEMRare modeled as one big PHES system with symmetri-cal upper and lower storage reservoirs and nonatural inflow or drainage. The solver ‘OSICPLEX’

(free version of CPLEX that comes free of chargewith the GAMS base system) was used for executingthe optimization.

NordicRegion

CentralEasternEurope

WesternBalkan

Italy

South-EasternEurope

BalticRegion

Central Western Europe

Iberian Peninsula

UK and Ireland

FIGURE 2 | Clustering of countries to nine different European electricity market regions.



SETUP AND INPUT DATA

Establishment of Hourly RES-ElectricityGeneration Datasets for the ElectricityMarket Regions for the Years 2030and 2050For the derivation of the load duration curves andresidual load curves in the different EEMRs for theyears 2030 and 2050, the following input datawere used:

• Electricity demand: Electricity demand data onan hourly basis and on a country level wastaken from ENTSO-E for the year 2011.25 Forthe analyses of the EEMRs, different growthrates of electricity demand were applied for dif-ferent time periods in the Blue storyline: 1.2%for the period 2010–2030 and 1.1% for theperiod 2030–2050. The ENTSO-E electricitydemand data is including the network lossesbut excludes the consumption of PHES (pump-ing mode) and generation auxiliaries. With thisdata, the regional load duration curves wereestablished for the years 2030 and 2050.

• Hourly CSP, PV, and wind electricity genera-tion: To establish an hourly concentrated solarthermal power (CSP), photovoltaic (PV), andwind electricity generation dataset for eachEEMR, real hourly CSP, PV, and wind genera-tion data was used from selected countriesfor the year 2011 and linearly up-scaled accord-ing to the RES-Electricity deployment in theBlue storyline in the years 2030 and 2050(cf. Table 1): the Spanish dataset from the Span-ish transmission system operator (TSO) RedElectrica26 for the Iberian Peninsula and theGerman27–30 and Austrian31 datasets for theCWE region.

• Hourly electricity generation of other renew-ables: Other RES-Electricity technologies(e.g. biomass, biogas, geothermal, etc.) wereapproximated as constant generation bandsthroughout the year.

• Hourly run-of-river hydroelectricity generation:The hourly run-of-river (RoR) hydroelectricitygeneration for the Iberian Peninsula was estab-lished again by the linear up-scaling of the realRoR hydro generation dataset of the SpanishTSO Red Electrica for the year 2011. As no realdatasets were available for Germany or Austria,average daily water-level values of the Rhineand Danube river32 were taken to establishhourly RoR electricity generation data (linearlyscaled between the measurement points andmultiplied by the respective installed hydro-power plants (HPP) capacities in the region) forthe CWE and Western Balkan region.

An example for the established sets of regionalresidual load curves is shown in Figure 3, where theload duration and different residual load curves forthe CWE region in the Blue storyline in the year2030 are given. In general, the residual load is calcu-lated as follows:

RL tð Þ=ED tð Þ−REPV tð Þ−RECSP tð Þ−REWind tð Þ−REOther−REROR tð Þ; ð5Þ

RL(t) is the residual load in hour t; ED(t) is the elec-tricity demand in hour t; REPV(t) is the RES-Electricity from PV installations in hour t; RECSP(t) isthe RES-Electricity from CSP installations in hour t;REWind(t) is the RES-Electricity from wind installa-tions in hour t; REOther is the electricity from otherRES-Electricity technologies; and RERoR(t) is theRES-Electricity from RoR HPP in hour t.

TABLE 1 | RES-Electricity Capacities of the CWE Region and Iberian Peninsula in the Blue Storyline (input from the SUSPLAN project1)

RES-Electricity Capacities in the Blue storyline (GW)

CWE Iberian Peninsula

2030 2050 2030 2050

Other RES-Electricity 23.1 28.4 7.5 9.3

RoR HPP 41.8 41.8 19.7 19.7

Photovoltaics 42.8 59.1 13.8 34.4

Solar thermal electricity 0.0 0.0 16.8 33.6

Wind onshore 86.0 120.7 64.3 81.4

Wind offshore 57.2 160.3 2.2 12.7

Total RES-E capacities 250.9 410.3 124.4 191.0



It is important to note that in Figure 3, severaldifferent residual load (RL) curves are presented. Asindicated in the figure, ED−REPV depicts the classifiedRL curve after only RES-Electricity from PV isdeducted, and ED−REPV−REWind depicts the classi-fied RL curve after RES-Electricity from PV and windare deducted from the electricity demand.

As power (capacity) is indicated on the verticalaxis and time (number of hours) is indicated on thehorizontal axis, the area between two RL curvesrepresents the total electricity produced by a (RES-Electricity) technology over a year, for eample, theyellow area between the load duration curve and ED-REPV represents the total electricity generated by PVsystems over a year. Similarly, the other colored sur-faces represent annual wind (purple area), otherRES-Electricity (green area), and run-of-river hydro(blue area) electricity generation.

Note that every RL curve is a new classified RLcurve after adding—or better deducting—the contri-bution of another technology. The values are in nodirect relationship per hour to the former RL curve(above) as after every subtraction of respective RES-Electricity feed-in from the load, the RL curve issorted again from the highest to lowest RL values.Therefore, a vertical cross-section from the loadcurve to the RL curve does not determine RES-Electricity feed-in at a certain point in time simplybecause it represents RES-Electricity feed-in datafrom different points in time.

Consideration of the Age Structure of theExisting Thermal Power Plant Portfoliofor the Years 2030 and 2050The currently existing thermal power plant portfolio(e.g. nuclear, lignite, coal, etc.) within each of the dif-ferent EEMRs was also considered in the analysis. Bydoing this, the age structure and the phase-out of theexisting thermal power plant portfolio (see Figure 6for an example) were generated from the PLATTSdatabase6 for all different electricity market regions.Installations of new thermal power plant capacitiesup to the year 2015 are already considered withinthe database.

Figure 4 shows an example of the status quo ofthe age structure of the thermal power plant portfolioin the CWE region. It can be seen that—ceteris pari-bus (assuming current decisions on lifespans ofnuclear power plants left unchanged)—the majorityof nuclear capacities phase out by 2040, and onlyfew new gas power plants are constructed until 2015in the CWE region.

Coverage of the Future Residual Loadswith Existing Thermal and Pumped-Hydro Power Plants in the Years 2030and 2050After deriving the age structure of the thermal powerplant portfolio in the different regions, established

275

250

225

200

175

150

125

Resid

ual lo

ad (

GW

)

100

75

50

25

00 500 1000

Residual load curves of CWE in the Blue Storyline 2030

PV

ED – REPV

ED – REPV – REWind

ED – REPV – REWind – REOther

ED – REPV – REWind – REOther – RERoR = RL

Other RES-E

Wind

RoR HPP

Load duration curveResidual load

1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500

Hours over a year (h)

FIGURE 3 | Classified load duration and residual load curves for the Central Western Europe (CWE) region in the Bluestoryline in 2030.



RL curves for the years 2030 and 2050 were‘filled up’ with the still-existing thermal power plantcapacities (cf. Figure 5). The thermal power plantcapacities are drawn as constant bands, startingwith the base-load and least-costly power plants(i.e. nuclear, lignite, and coal) followed by gas andoil power plant capacities. In order to also incorpo-rate power plant availabilities (i.e. offline periods dueto maintenance etc.), installed thermal power plantcapacities were multiplied by a factor of 0.8(nuclear)33 and 0.85 (all other thermal powerplants),34 respectively.

Additionally, currently existing, licensed, andplanned installed capacities of PHES systems in therespective region are depicted as constant bands indi-cated downward from the top of the RL curves inorder to show their potential for providing peak-loadpower (cf. Figure 5). The electricity consumption ofexisting and future PHES systems in pumping modeis not incorporated in the RL curves in this stage ofthe analysis. However, in general, this additionaldemand would only alter the RL values in times ofhigh RES-Electricity feed-in/low RL (i.e. right side ofthe RL curve).

300Age structure of thermal power plant-portfolio in CWE

Other thermal

Oil

Gas

Coal

Lignite

Nuclear

Total

250

200

150

100

Insta

lled c

apacity (

GW

)

50

02005 2010 2015 2020 2025

Year

2030 2035 2040 2045 2050

FIGURE 4 | Age structure of the thermal power plant portfolio in the Central Western Europe (CWE) region.

250

225

200

175

150

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

125

100

75

50

25

0

–250 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

GapOil

PHESOther TPPGasLigniteResidual load

CoalNuclear

7500 8000 8500


Residual load curve of CWE in the Blue Storyline 2030

FIGURE 5 | Coverage of the 2030 residual load of the Central Western Europe (CWE) region with existing thermal power plants and pumpedhydro energy storage (PHES) in the Blue storyline.



In many analysed regions and time horizons, agap (indicated in white colour) remains between thePHES band and the upper band of the thermal powerplants, meaning that there is not enough installedpower plant capacity available within the region tomeet regional electricity demand.

Figure 5 shows the coverage for the 2030 RL inthe CWE region with existing thermal power plantsand PHES systems in the Blue storyline. It can beseen that in the CWE region, a capacity gap of about35 GW remains for the year 2030, meaning that newthermal and/or PHES power plants will be needed tocover electricity demand in this scenario (higher RES-Electricity deployment would lower this capacitygap). The major still-existing and available thermalpower plants (TPP) in 2030 in the CWE region aregas-fired TPP (about 68 GW) followed by nuclear(about 52 GW) and coal-fired TPP (about 35 GW).

Integration of Surplus RES-ElectricityGeneration with Pumped Hydro SystemsIn the final step of the analysis, the established PHESoptimization algorithm is executed. Based on theconducted literature review, the maximum availableturbine/pumping power of the PHES system in theCWE region for the year 2030 was set to 35/30 GW,and the maximum storage capacity was defined to4000 GWh.

The result of the PHES optimization for theCWE region in the year 2030 is shown in Figure 6.

The light blue area indicates the difference betweenthe original RL curve (black dashed line in Figure 6)and the updated RL curve that also incorporates thePHES operation (blue line in Figure 6). It can be seenthat the PHES system provided peak-load power (i.-e. reduction of the original RL curve due to turbineoperation on the left side of Figure 6 until about4000 full-load hours) and created additional electric-ity demand due to pumping in off-peak times (i.-e. increase of the original RL curve from about 4000full-load hours on in Figure 6). By doing so, the full-load hours of base-load TPP (i.e. nuclear, coal, andlignite) are also increased, and therefore, these tech-nologies get more economical. Furthermore, thesmall amount of excess RES-Electricity in the CWEregion for the year 2030 is integrated in the electric-ity system due to storage operation of the PHES sys-tem. The storage capacity use of CWE’s PHESsystem for the year 2030 is shown in Figure 7. It canbe seen that the energy storage (i.e. upper reservoirof PHES system) is fully charged and fully emptiedseveral times per year by the optimization algorithmwithout seasonal differences. Due to the optimizationconstraint (i.e. same end and start value), the storedenergy reaches 2000 GWh at the end of the year.

RESULTS FOR THE CENTRALWESTERN EUROPE REGION

In the Blue storyline, RES-Electricity feed-in exceedselectricity demand for some time in the year 2050 in

250

225

200

175

150

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

125

100

75

50

25

0

–250 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

Gap

Oil

PHES

Other TPP

Gas

Lignite

Residual loadResidual load w/PHES

Coal

Nuclear

7500 8000 8500



FIGURE 6 | Coverage of the 2030 residual load after pumped hydro energy storage (PHES) optimization in the Central Western Europe (CWE)region with existing thermal power plants in the Blue storyline.



the CWE region (cf. Figure 8 and dashed line inFigure 9). This RES-Electricity excess generation canbe used for (large-scale) electricity storage (e.g. thiselectricity might be available at low cost for pumpingpurposes in a PHES system) and/or for exports toneighboring regions.

However, due to the phase-out of the majorityof nuclear and lignite power plants in the region andhigh electricity demand increase, large amounts ofintegrated RES-Electricity cannot hamper the growth

of the gap of missing generation capacity in compari-son to the 2030 scenario (cf. Figure 5 and Figure 9).The generation of newly installed wind power plants(on- as well as off-shore) especially contributes tolowering the RL substantially. Only some generationcapacities of gas- and coal-fired TPP are still availa-ble in the CWE region for 2050. Therefore, the ‘gen-eration gap’ amounts to about 190 GW in 2050.

Coverage of the 2050 RL after PHES optimiza-tion in the CWE region with existing thermal power

4000

3500

3000

2500

Sto

rage c

apacity (

GW

/h)

2000

1500

1000

500

00 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500


Storage capacity of PHES systems in CWE in 2030

FIGURE 7 | Storage capacity of the pumped hydro energy storage (PHES) system in the Central Western Europe (CWE) region in theyear 2030.

350

300

250

200

Resid

ual lo

ad (

GW

)

150

100

50

00 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

WindRoR HPP

PV

Other RES-E

Residual load Load duration curve

7500 8000 8500


Residual load curves of CWE in the Blue Storyline 2050

FIGURE 8 | Classified load duration and residual load curves for the Central Western Europe (CWE) region in the Blue storyline in 2050. RES-E, renewable energy sources for electricity generation; RoR, run-of-river.



plants can be seen in Figure 10. For the optimizationin the year 2050, the same parameters of the PHESsystem were used as in 2030. Again the typical PHESsystem operation scheme—provision of peak-loadpower and consumption of off-peak power—can beobserved in the CWE region for 2050. Additionally,about 80% of the annual surplus RES-Electricitygeneration can be integrated (i.e. stored) in the inher-ent electricity system. The high negative peak of theRL precludes that all surplus RES-Electricity

generation can be integrated in the system; the PHESsystem is not sufficient, and more pumping power isneeded for full RES-Electricity integration.

RESULTS FOR THE IBERIANPENINSULA

Figures 11 and 12 show the construction of the RLcurve and its coverage with existing thermal power

350

300

250

200

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

150

100

50

0

–50

–100

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

GapPHES

Other TPPGas

Lignite

Residual load

Coal

Nuclear

7500 8000 8500



FIGURE 9 | Coverage of the 2050 residual load of the Central Western Europe (CWE) region with existing thermal power plants and pumpedhydro energy storage (PHES) in the Blue storyline. TPP, thermal power plants.

350

300

250

200

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

150

100

50

0

–50

–100

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

GapPHES

Other TPPGas

Lignite

Residual loadResidual load w/ PHES

Coal

Nuclear

7500 8000 8500



FIGURE 10 | Coverage of the 2050 residual load after pumped hydro energy storage (PHES) optimization in the Central Western Europe(CWE) region with existing thermal power plants in the Blue storyline. TPP, thermal power plants.



plants and PHES in the Blue storyline for the IberianPeninsula in the year 2030. The RES-Electricity feed-in already exceeds electricity demand for quite sometime (about 2000 h) of the year 2030 in the IberianPeninsula. Wind and solar (PV and CSP) generationespecially lead to low and even negative RL values.

Furthermore, it can be seen that sufficient (ther-mal) power plant capacity is available in the region(i.e. no capacity gap between supply and demand)for 2030 (cf. Figure 12); there is an overcapacity ofmore than 34 GW of gas-fired TPP in the IberianPeninsula. This fact has two reasons: on one hand,

high RES-Electricity deployment in the region and,on the other hand, large amounts of still-existingTPP in the year 2030, especially due to high invest-ments in gas-fired thermal power plants in the last10 years in the Iberian Peninsula.

This set of flexible gas-fired electricity genera-tion technologies [i.e. CCGTs and conventionalopen-cycle gas turbines (OCGT)] is perfectly quali-fied to balance electricity systems and to providereserve capacities in electricity systems with highshares of variable RES-Electricity generation. Thesegas-fired generation technology types are needed and,

50

60

40

30

Resid

ual lo

ad (

GW

)

20

10

00 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

Wind

PV

RoR HPP

Load duration curve

CSP

Other RES-E

Residual load

7500 8000 8500


Residual load curves of the lberian Peninsula in the Blue Storyline 2030

FIGURE 11 | Classified load duration and residual load curves for the Iberian Peninsula in the Blue storyline in the year 2030. CSP,concentrated solar power; HPP, PV, photovoltaics; RES-E, renewable energy sources for electricity generation.

40

30

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

20

10

0

–10

–20

–30

–40

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

PHESGas

Lignite

Residual load

Coal

Nuclear

7500 8000 8500


Residual load curve of the Iberian Peninsula in the Blue Storyline 2030

FIGURE 12 | Coverage of the 2030 residual load of the Iberian Peninsula with existing thermal power plants and pumped hydro energystorage (PHES) in the Blue storyline.



alongside bulk ESTs (e.g. PHES, CAES, etc.), theyare key candidates for maintaining smooth electricitysystem operation, especially peripheral areas of elec-tricity systems, that is, in a European context, in theIberian Peninsula, Italy, UK, and Ireland (necessitydepending also on the future interconnection withthe Nordic region) and also other areas, such as theBalkan region (in the future, most probably passedthrough by a gas pipeline like ‘South Stream’). Thesecountries also have access to natural gas (either ownresources and/or transit countries of natural gas cor-ridors/hubs) as a primary energy carrier.

For the optimization of the PHES system opera-tion, in the Iberian Peninsula, for the year 2030, themaximum storage capacity was set to 2000 GWh,7

and the maximum turbine/pumping power was setto 14/13 GW. The results of the optimization algo-rithm are shown in Figure 13. It can be seen that theresidual peak-load in the Iberian Peninsula canbe lowered to about 18 GW for the year 2030.Furthermore, more than 90% of the domestic excessRES-Electricity generation can be integrated in theelectricity system by the PHES system. For a fullRES-Electricity integration, a higher pumping powerwould be needed. Due to the PHES system operation,the maximum full-load hours of TPP are increased toabout 7250 h. For about 1100 h, the RL is reducedto zero by the PHES system; the electricity system isbalanced with 100% RES-Electricity generation.

RES-Electricity feed-in exceeds electricitydemand about half the time for the year 2050 in theIberian Peninsula; additional solar generation unitsespecially contribute to this. Until 2050, a further

increase in RES-Electricity deployment leads to evenlower maximum full-load hours for TPP in the Ibe-rian Peninsula (cf. Figures 14 and 15). However, CSPplants equipped with thermal energy storage systemscould, to some extent, also operate in a more flexibleway,35 that is, as dispatchable power plants compa-rable to gas-fired TPP. When operating CSP plantslike TPP, that is, taking them out of the calculationof the RL in Eq. (5), the maximum full-load hours ofTPP can be increased again to about 6500 h.

As already seen in the results for the year 2030,sufficient generation capacities (gas and PHES powerplants) are available in the system to cover the RLalso in 2050. Again, there is an overcapacity in theregion—about 9 GW of gas-fired TPP.

For the PHES system optimization in the year2050, the turbine as well as the pumping power wereincreased by 1 GW to 15 and 16 GW, respectively.The storage capacity was held constant, indicating anupgrade of existing PHES facilities in the Iberian Pen-insula between 2030 and 2050. Coverage of the2050 RL after PHES optimization in the Iberian Pen-insula with existing thermal power plants is depictedin Figure 16. Again, the residual peak-load is loweredto about 25 GW. Due to the low RL in the IberianPeninsula in the year 2050 in general, the remainingRL is also curtailed to a large extent; maximum full-load hours of TPP are therefore further reduced toabout 4000 h. About 60% of the surplus RES-Electricity generation can be integrated in the IberianPeninsula’s electricity system by the PHES system.However, large amounts of excess RES-Electricitygeneration remain, about 26 TWh in the year 2050.

40R

esid

ual lo

ad/insta

lled c

apacity (

GW

) 30

20

10

0

–10

–20

–30

–40

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

PHESGas

Lignite

Residual loadResidual load w/ PHES

Coal

Nuclear

7500 8000 8500



FIGURE 13 | Coverage of the 2030 residual load after pumped hydro energy storage (PHES) optimization in the Iberian Peninsula withexisting thermal power plants in the Blue storyline.



This indicates that there is an additional need forpumping power in the region.

ROLE OF CROSS-BORDERTRANSMISSION GRID EXPANSIONAND EXTREME WEATHER EVENTSIN THE EUROPEAN ELECTRICITYMARKET REGIONS

The contribution of possible future transmission gridexpansion between neighboring EEMRs for better

matching variable RES-Electricity generation,bulk EST/other flexible electricity generation technol-ogies, and load centers (e.g. balancing ‘stressed’ con-tinental European electricity systems with bulkstorage energy from PHES from the Alps) isqualitatively discussed in Figure 17. Furthermore,the contribution of existing and new PHES andflexible thermal power plant units within aregion and the management of extreme weatherevents (e.g. high correlation of wind acrossEurope) within a region and between regions areassessed.

60

70

50

40

30

Resid

ual lo

ad (

GW

)

20

10

00 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

CSPPV

Wind

RoR HPP

Load duration curve

Other RES-E

Residual load

7500 8000 8500


Residual load curves of the Iberian Peninsula in the Blue Storyline 2050

FIGURE 14 | Classified load duration and residual load curves for the Iberian Peninsula in the Blue storyline in the year 2050. CSP,concentrated solar power; HPP, PV, photovoltaics; RES-E, renewable energy sources for electricity generation; RoR, run-of-river.

20

40

30

10

–10

0

–20

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

–30

–40

–50

–60

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

PHESGasCoalResidual load

7500 8000 8500



FIGURE 15 | Coverage of the 2050 residual load of the Iberian Peninsula with existing thermal power plants and pumped hydro energystorage (PHES) in the Blue storyline.



Due to a lack of available (flexible) power plantcapacities in the future, the CWE region can hardlycontribute to balancing services in neighboringregions, even in case of significant transmission gridexpansion (cf. red export arrows of the CWE regionin Figure 17). However, imports from the Nordicregion, the Iberian Peninsula, and Italy could helpmitigate the effects of variable RES-Electricity genera-tion in the CWE region in case of significant trans-mission grid expansion and negative correlation ofwind. In case of moderate transmission grid expan-sion or positive correlation of wind between theregions, the contribution from these regions is limited(cf. yellow-green import arrows of the CWE regionin Figure 17). Imports of flexibility from the CentralEastern Europe (CEE), Western Balkan, and the UKand Ireland regions are only limited (even in case ofsignificant transmission grid expansion and negativecorrelation of wind, cf. yellow import arrows of theCWE region in Figure 17). Table 2 explains the situa-tion for the CWE region in detail.

In addition to the information given inFigure 17 (i.e. contribution from outside and contri-bution to other regions), Table 2 also indicates thatnew and existing PHES and thermal power systemscontribute highly to mitigating the effects of variableRES-Electricity generation in the CWE region.

The Western Balkan region could only providelimited flexibility to the CWE and the SEE regions.Due to current low transmission capacity betweenItaly and the Western Balkan region, a significanttransmission expansion is necessary for reciprocalflexibility provision between these two regions.

As the Iberian Peninsula is located on the outeredge of the European electricity system, its import-ing/exporting capabilities from/to neighboringregions are limited. However, in case of significanttransmission grid expansion to the CWE region, theIberian Peninsula could export flexibility from gas-fired thermal power plant units.

The CEE region is very similar to the CWEregion in terms of possible contribution to neighbor-ing electricity regions as it also has only limited capa-bility. The region could profit from Nordic, SEE, andWestern Balkan PHES capacities in case of significanttransmission grid expansion and negative correlationof wind.

Because of its vast amounts of flexible (P)HESsystems, the Nordic region could contribute to bal-ance neighboring European regions in case of signifi-cant transmission grid expansion and negativecorrelation of wind. If additional generation capaci-ties are needed, UK and Ireland could contribute withtheir flexible thermal power plant portfolio.

For the remaining EEMRs, the following canbe observed: as Italy’s electricity system is very simi-lar to the Iberian Peninsula’s (i.e. high shares of gas-fired thermal power plants and PHES systems), Italycould export flexibility to CWE and the Western Bal-kan region. Because of high shares of PHES capaci-ties in its electricity system, the SEE region couldexport flexibility to its neighboring regions. On thecontrary, due to low overall installed power plantcapacities, the Baltic region’s flexible power plantportfolio is only of minor importance in a Europeancontext. In the UK and Ireland region, PHES systems

20

40

30

10

–10

0

–20

Resid

ual lo

ad/insta

lled c

apacity (

GW

)

–30

–40

–50

–60

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000

PHESGasCoalResidual loadResidual load w/ PHES

7500 8000 8500



FIGURE 16 | Coverage of the 2050 residual load after pumped hydro energy storage (PHES) optimization in the Iberian Peninsula withexisting thermal power plants in the Blue storyline.



will have little contribution to the electricity systembecause of low future PHES potentials. Nevertheless,the UK and Ireland region could provide flexibilityfrom gas-fired thermal power plant units to neighbor-ing regions.

In general, it can be observed that existing andnew PHES and flexible thermal power plant units arestrongly needed in the majority of the European elec-tricity regions to cope with the effects of variableRES-Electricity feed-in in the future. Sufficient (flexi-ble) power plant capacity is already in the electricitysystem only in the Iberian Peninsula and Italybecause of high investments in new gas-fired thermalpower plants in the last 10 years and also due to highamounts of already implemented PHES schemes.

SUMMARY AND CONCLUSION

The long-term scenario analyses performed in thispaper showed interdependencies between regionaland trans-national energy systems and also theimportance of bulk ESTs for integrating largeamounts of variable RES-Electricity in electricity sys-tems in the future. The analysed Blue storyline withhigh RES-Electricity deployment and high electricitydemand growth showed the following results.

Due to age-related phase-out of thermal powerplants in the future, additional new power plantcapacities are will be needed in many European elec-tricity regions by the year 2030. These gaps of elec-tricity generation capacity can be either filled up with

FIGURE 17 | Contribution of transmission grid expansion for mitigation of variable renewable energy sources (RES)-Electricity generation(e.g. wind) within the regions.



new PHES systems (as far as additional potential isavailable in the region) or new flexible thermalpower plants; higher RES-Electricity generation alsoreduces these gaps. However, in the electricity marketregions of the Iberian Peninsula, sufficient (thermal)power is available (i.e. no capacity gap between gen-eration and demand). This has two reasons: on onehand, high RES-Electricity deployment in the region(especially wind, but also PV) and, on the otherhand, large amounts of still-existing thermal powerplants in the year 2030/2050 due to high investmentsin gas-fired thermal power plants in the last 10 years.

Furthermore, in the analysed Blue storyline,RES-Electricity feed-in exceeds electricity demand for

some time of the year 2050 in several of theanalysed regions. The Iberian Peninsula especiallyshows high RES-Electricity feed-in exceeding electric-ity demand about half the time of the year 2050.This surplus RES-Electricity generation can beused for (large-scale) electricity storage (e.g. this elec-tricity might be available at low cost for pumpingpurposes in a PHES system) and/or for exports toneighboring European regions if there are enoughtransmission grid capacities in place. As shown in theanalysis, PHES systems in the CWE region and Ibe-rian Peninsula can help to integrate large amounts ofthe surplus RES-Electricity, 80% and 60%,respectively.

TABLE 2 | Contribution of Transmission Grid Expansion for the Mitigation of Variable RES-Electricity Generation (e.g. Wind) within the CWERegion and Management of Extreme Weather Events

Contribution of Transmission Expansion for Mitigation of Wind within the Regions and Management of Extreme Weather Events

Central Western Europe

Within the Region

Contribution in CWE

PHES Thermal

Contributions from Outside (Imports’)

Transmission Expansion to the Nordic Region

Mode Rate

Limited

Limited

Significant

Significant

Moderate

Limited

Moderate

High (Anticorrelation Wind, PHES)

Limited (Anticorrelation Wind)

High (Anticorrelation Wind, Thermal)

Limited (Correlation Wind)



Transmission Expansion to UK and Ireland

Transmission Expansion to the Iberian Peninsula

Transmission Expansion to Italy

Transmission Expansion to the Western Balkan

Transmission Expansion to CEE

Moderate

Moderate

Limited

Limited

Limited

Significant

Significant

Moderate Significant

High (Anticorrelation Wind, Thermal)






Significant

Contributions to Other Regions (’Exports’)

Transmission Expansion to the Nordic Region

Moderate

Low

Transmission

Moderate

Low

Low (Anticorrelation Wind)

Low (Correlation Wind)

Expansion to UK and Ireland

Significant



Transmission Expansion to theIberian Peninsula

Moderate

Low

Transmission Expansion to Italy

Moderate

Low

Significant



Transmission Expansion to the Western Balkan

Transmission Expansion to CEE

Moderate

Moderate

Low

Low

Significant

Significant





Significant



Significant

High (Existing) High (Existing)

High (New) High (New)



Because of its vast amounts of flexible (P)HESsystems, the Nordic region could significantly con-tribute to balance the electricity systems of

neighboring regions in case of significant transmis-sion grid expansion and negative correlation of windbetween the regions.

FURTHER READINGSBessa R, Moreira C, Silva B, Matos M. Handling renewable energy variability and uncertainty in power systems operation.WIREs Energy Environ 2014, 3:156–178. doi:10.1002/wene.76.

Tuohy A, Kaun B, Entriken R. Storage and demand-side options for integrating wind power. WIREs Energy Environ 2014,3:93–109. doi:10.1002/wene.92.

Inage SI. The role of large-scale energy storage under high shares of renewable energy. WIREs Energy Environ 2015,4:115–132. doi:10.1002/wene.114.

REFERENCES1. SUSPLAN Consortium. Available at: www.susplan.eu.

(Accessed May 12, 2015).

2. Auer H, Zach KA, Lettner G. Synthesis of Results fromthe Regional Scenario Studies. Deliverable D2.10 fromthe project “SUSPLAN”, 2010.

3. Auer H, Zach KA, Totschnig G, Wiessensteiner L. Sce-nario Guidebook: Definition of a Consistent ScenarioFramework. Deliverable D1.2 from the project “SUS-PLAN”, 2009.

4. Huber C, Faber T, Resch G, Auer H. The IntegratedDynamic Formal Framework of GreenNet. ProjectReport (WP8), 2004.

5. GAMS Software GmbH. Available at: http://www.gams.com/. (Accessed May 12, 2015).

6. PLATTS. Electric power plant database – Europe.2010. Available at: www.platts.com. (Accessed May12, 2015).

7. ENTSO-E. System Adequacy Forecast 2010 – 2025.Report and scenarios, 2010.

8. EURELECTRIC. Hydro in Europe: powering renew-ables. 2011. Available at: http://www.eurelectric.org.(Accessed May 12, 2015).

9. Beurskens LWM, Hekkenberg M. Renewable EnergyProjections as Published in the National RenewableEnergy Action Plans of the European Member States.ECN-E--10-069. 2011. Available at: http://www.ecn.nl/docs/library/report/2010/e10069.pdf. (Accessed May12, 2015).

10. Huber C, Faber T, Haas R, Resch G, Green J, Ölz S,White S, Cleijne H, Ruijgrok W, Morthorst PE,et al. Action plan for deriving dynamic RES-E policies.Report of the European research project Green-X –

funded by the EC-DG Research, Vienna University ofTechnology, 2004.

11. Joint Research Centre (JRC). Pumped-hydro energystorage: potential for transformation from single dams.

JRC Scientific and Technical Report. ISBN 978-92-79-23182-7. 2012. Available at: http://setis.ec.europa.eu/newsroom-items-folder/report-pumped-hydro-energy-storage-potential-for-transformation-from-single-dams.(Accessed May 12, 2015).

12. Deane JP, Ó Gallachóir BP, McKeogh EJ. Techno-economic review of existing and new pumped hydroenergy storage plant. Renew Sustain Energy Rev 2010,14:1293–1302.

13. Steffen B. Prospects for pumped-hydro storage in Ger-many. Energy Policy 2012, 45:420–429.

14. Kapsali M, Kaldellis JK. Combining hydro and varia-ble wind power generation by means of pumped-storage under economically viable terms. Appl Energy2010, 87:3475–3485.

15. Kaldellis JK, Kapsali M, Kavadias KA. Energy balanceanalysis of wind-based pumped hydro storage systemsin remote island electrical networks. Appl Energy2010, 87:2427–2437.

16. Duque ÁJ, Castronuovo ED, Sánchez I, Usaola J. Opti-mal operation of a pumped-storage hydro plant thatcompensates the imbalances of a wind power pro-ducer. Electr Pow Syst Res 2011, 81:1767–1777.

17. Dena. Analyse der Notwendigkeit des Ausbaus vonPumpspeicherwerken und anderen Stromspeichern zurIntegration der erneuerbaren Energien. DeutscheEnergie-Agentur GmbH (dena), 2010.

18. Radgen P. Energy storage: from compressed air energystorage to E-mobility. In: Presentation at User ForumPower Plant Technology, April 21, 2009, Hannover,E.ON Energie AG, 2009.

19. RWE Power. ADELE – adiabatic compressed-airenergy storage for electricity supply, brochure; 2011.Available at: www.rwe.com. (Accessed December28, 2015).



http://www.susplan.eu

http://www.gams.com/

http://www.gams.com/

http://www.platts.com

http://www.eurelectric.org

http://www.ecn.nl/docs/library/report/2010/e10069.pdf

http://www.ecn.nl/docs/library/report/2010/e10069.pdf

http://setis.ec.europa.eu/newsroom-items-folder/report-pumped-hydro-energy-storage-potential-for-transformation-from-single-dams



http://www.rwe.com

20. Zach KA, Auer H, Lettner G. Report summarizing thecurrent status, role and costs of energy storage technol-ogies. Deliverable 2.1 of the project “stoRE”. 2012.Available at: www.store-project.eu. (Accessed May12, 2015).

21. Kloess M, Zach KA. Bulk electricity storage technolo-gies for load-leveling operation – An economic assess-ment for the Austrian and German power market. IntJ Elec Power Energy Syst 2014, 59:111–122.

22. European Commission (EC). Energy infrastructurepriorities for 2020 and beyond - A Blueprint for anintegrated European energy network. Communicationfrom the Commission - COM(2010) 677/4, 2010.

23. European Commission (EC). Priority InterconnectionPlan. Communication from the Commission - COM(2006) 846 final/2, 2007.

24. ENTSO-E. Pilot Ten-Year Network Development Plan(TYNDP). Executive Summary 2010. 2010. Availableat: https://www.entsoe.eu/fileadmin/user_upload/_library/SDC/TYNDP/100630_TYNDP_Executive_-Summary.pdf. (Accessed May 12, 2015).

25. ENTSO-E. Data Portal – Consumption Data. 2012.Available at: https://www.entsoe.eu/data/data-portal/consumption/. (Accessed May 12, 2015).

26. Red Eléctrica de España (REE). Sistema de Informa-ción del Operador del Sistema (e � sios). 2012. Availa-ble at: http://www.esios.ree.es/web-publica/. (AccessedMay 12, 2015).

27. 50Hertz Transmission. Kennzahlen. 2012. Availableat: http://www.50hertz.com/de/Kennzahlen. (AccessedMay 12, 2015).

28. Amprion. Netzkennzahlen. 2012. Available at: http://www.amprion.net/netzkennzahlen. (Accessed on May12, 2015).

29. TenneT TSO. Netzkennzahlen. 2012. Available at:http://www.tennettso.de/site/Transparenz/veroeffentli-chungen/netzkennzahlen. (Accessed May 12, 2015).

30. TransnetBW. Kennzahlen – Erneuerbare Energien.2012. Available at: https://www.transnetbw.com/en/key-figures. (Accessed May 12, 2015).

31. APG. Marktinformationen. 2012. Available at: http://www.apg.at/de/markt. (Accessed May 12, 2015).

32. Lebensministerium.at. eHYD. 2012. Available at:http://ehyd.gv.at/. (Accessed May 12, 2015).

33. European Environment Agency (EEA). Availabilityimprovements in nuclear power plants in Europe.2012. Available at: http://www.eea.europa.eu/data-and-maps/figures/availability-improvements-in-nuclear-power. (Accessed May 12, 2015).

34. EURELECTRIC. Power Statistics & Trends 2011 –

Synopsis. 2011. Available at: http://www.eurelectric.org/media/26800/power_statistics_and_trends_synosp-sis-2011-180-0005-01-e.pdf. (Accessed May 12, 2015).

35. International Energy Agency (IEA). Harnessing Varia-ble Renewables – A Guide to the Balancing Challenge.2011. ISBN: 978 92 64 11138 7. IEA Publications, 9,rue de la Fédération, 75739 Paris CEDEX 15.



http://www.store-project.eu

https://www.entsoe.eu/fileadmin/user_upload/_library/SDC/TYNDP/100630_TYNDP_Executive_Summary.pdf



https://www.entsoe.eu/data/data-portal/consumption/

https://www.entsoe.eu/data/data-portal/consumption/

http://www.esios.ree.es/web-publica/

http://www.50hertz.com/de/Kennzahlen

http://www.amprion.net/netzkennzahlen

http://www.amprion.net/netzkennzahlen

http://www.tennettso.de/site/Transparenz/veroeffentlichungen/netzkennzahlen

http://www.tennettso.de/site/Transparenz/veroeffentlichungen/netzkennzahlen

https://www.transnetbw.com/en/key-figures

https://www.transnetbw.com/en/key-figures

http://www.apg.at/de/markt

http://www.apg.at/de/markt

http://ehyd.gv.at/

http://www.eea.europa.eu/data-and-maps/figures/availability-improvements-in-nuclear-power



http://www.eurelectric.org/media/26800/power_statistics_and_trends_synospsis-2011-180-0005-01-e.pdf



Documents

Contribution of bulk energy storage to integrating …storage to integrating variable renewable energies in future European electricity systems Karl Anton Zach* and Hans Auer Remarkably