This action might not be possible to undo. Are you sure you want to continue?
for transformation from single dams
Analysis of the potential for transformation of non-hydropower dams and reservoir hydropower
schemes into pumping hydropower schemes in Europe
Roberto Lacal Arántegui, Institute for Energy and Transport, Joint Research
Centre of the European Commission, Petten, the Netherlands.
Niall Fitzgerald and Paul Leahy, Sustainable Energy Research Group,
University College Cork, Cork, Ireland.
EUR 25239 EN - 2012
The mission of the JRC-IET is to provide support to Community policies related to
both nuclear and non-nuclear energy in order to ensure sustainable, secure and
efficient energy production, distribution and use.
Joint Research Centre
Institute for Energy and Transport
Address: Westerduinweg 3, NL-1755 LE Petten, the Netherlands
Tel.: +31 224 565 390
Fax: +31 224 565 616
Neither the European Commission nor any person acting on behalf of the
is responsible for the use which might be made of this publication.
Europe Direct is a service to help you find answers
to your questions about the European Union
Freephone number (*):
00 800 6 7 8 9 10 11
(*) Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.
A great deal of additional information on the European Union is available on the
It can be accessed through the Europa server http://europa.eu/
EUR 25239 EN
Luxembourg: Publications Office of the European Union
© European Union, 2011
Reproduction is authorised provided the source is acknowledged
Printed in the Netherlands
1able of contents
1 Deíinition oí a methodology íor transíormation 5
1.1 Di·ersity oí topologies íor PlS transíormation 5
1.2 Introduction to countries selected as case studies 5
1.3 Methodology íor transíormation 6
1.4 Country potential 10
1.5 GIS implementation 11
2 Database oí dams and hydropower schemes 11
2.1 Countries included in database 12
2.2 Data sources 13
2.3 Database íields 1¯
2.4 Disclaimer oí warranty and citations oí data utilised in database 18
3 Potential íor transíormation to PlS 19
3.1 Data required and data processing 19
3.2 Design oí ArcGIS model 24
3.3 Scenario design 1A & 1B 2¯
3.4 Result oí the analysis 29
4 Barriers to the realisation oí this potential 3¯
4.1 1opographical barriers 3¯
4.2 Lconomic barriers 38
4.3 Social barriers 41
4.4 Ln·ironmental and planning barriers 41
4.5 \ater supply barriers 42
4.6 llood protection 42
4.¯ Conclusions oí the barriers analysis 42
5 1opics íor íuture research 43
5.1 Post modelling site analysis 43
5.2 luture model de·elopment 44
5.3 lurther related research 45
5.4 Potential co-operation with GRanD 46
6 Methodological remarks and conclusions 4¯
¯ Conclusions 48
lurther reading 51
Llectricity supply and demand has to be balanced íor the system to work smoothly and íor
achie·ing this balance systems use demand-, supply- and system-based elements and measures.
In modern electricity grids electricity storage is a major system resource to keep that balance,
and currently the only widespread, large-scale electricity storage installed are reser·oir-based
hydropower and pumped hydropower storage ,PlS, |1[. One diííerence between both is that
PlS can act as load when electricity is spare in the system thus absorbing the excesses.
1he decarbonisation oí electricity systems requires increased use oí renewable energies, the
íastest-growing oí those, solar and wind, are dependent on natural resources that are not
necessarily a·ailable when electricity is most demanded. Increased penetration oí wind and
solar electricity is thereíore dependent, among other íactors, on electricity systems de·eloping
larger storage capacities. 1he problem is compounded by the eííects oí climate change on the
a·ailability oí rain and thus hydropower generation.
1he potential íor íurther con·entional hydropower in Lurope is limited because oí
en·ironmental considerations, lack oí adequate sites and certain social acceptance issues. New
PlS schemes are subject to similar limitations, but this is likely not to be the case íor PlS
resulting írom the transíormation oí existing hydropower and non-hydropower reser·oirs.
Reasons include that an existing reser·oir, candidate íor transíormation to PlS, already caused
eííects ,e.g. en·ironmental, long time ago and currently íorms part oí a more stable system
where -hopeíully- those problems ha·e been alle·iated. 1he two-dam system PlS is, in this
context, a closed-circuit whose PlS-related impact is unlikely to signiíicantly spread beyond the
system. 1he transíormation oí single reser·oirs to PlS thereíore becomes the simplest way to
add electricity storage capacity, has lower costs than new PlS and lower en·ironmental impact
than new reser·oir hydropower.
A PlS scheme, while not necessarily adding more electricity oí renewable origin ,PlS
electricity is as renewable as the electricity that was used to pump the water up in the íirst time,
minus the cycle eííiciency losses oí 15-30 °,, would allow the integration oí more renewable
·ariable electricity |2[. Other system beneíits include the replacement oí expensi·e peak-ser·ing
power plants íuelled by oil or natural gas and, beyond electricity production, the contribution to
ílood control and water supply that are typical oí any reser·oir-based hydropower plant.
1he analysis oí hydropower potential has been widely explored. lor example, the Luropean
Ln·ironmental Agency commissioned ALA to deíine a methodology to estimate the Luropean
en·ironmentally compatible potential íor small hydropower ,SlP, |3[. lowe·er, a literature
search exercise showed that the PlS potential has hardly been analysed. 1he analysis oí the
potential íor PlS -or íor transíormation to PlS íor this matter- was carried out in Lurope at
project le·el by pri·ate companies ,e.g. R\L,, at regional ,e.g. Canary Islands, and national
,e.g. Ireland, le·els, but a similar analysis was ne·er carried out íor the whole oí Lurope.
1he objecti·e oí this report is to deíine a methodology íor íinding the potential íor
transíormation to pumped hydro schemes ,PlS, under two gi·en topologies, and to test this
methodology in two cases, Croatia and 1urkey. 1he methodology uses a geographical
iníormation system ,GIS, tool and a purposely-de·eloped database oí Luropean reser·oirs.
Section 1 oí this report de·elops the methodology under two diííerent topologies. Section 2
deíines the database oí hydropower and non-hydropower reser·oirs on which the methodology
is based, and how it was created. Section 3 applies the methodology to the data in the database
and describes and quantiíies the potential íor transíormation to PlS íor the two case studies.
Section 4 then identiíies barriers to the realisation oí this potential whereas Section 5 identiíies
topics íor íuture research and Section 6 summarises the íindings.
J Definition of a methodology for transformation
J.J Di·ersity oí topologies íor PlS transíormation
A PlS scheme requires the existence oí an upper and a lower reser·oir between which water is
pumped up -mostly in oíí-peak periods- to store hydraulic potential energy and then released
down through a turbine -mostly in peak periods- to produce electricity. One or two penstocks
join the two reser·oirs and a power house, oíten built inside the mountain, contains the
pumping and generation as well as any ancillary equipment. A dynamic representation oí how a
PlS works is a·ailable at Verbund`s web site |5[.
1he íollowing topologies íor transíormation would be possible.
• •• • 1opology A (1A): when a reser·oir exists already 1A consists oí adding a second
reser·oir, normally at a higher ele·ation, plus penstock and equipment. 1he Dinorwig PlS
plant project in Gwynedd, North \ales, UK, is an example oí 1A.
• •• • 1opology B (1B): when two reser·oirs already exist and are within suitable distance and
diííerence in ele·ation, 1B consists oí adding generation and pumping equipment between
them. Lxisting natural lakes can also be considered one oí the two reser·oirs in this
topology, and these dams might be in the same ri·er or in parallel ·alleys. Limberg II PlS
plant in Kaprun, Salzburg, Austria, a 480-M\ plant due íor completion in 2011, is an
example oí a 1B PlS transíormation |5[.
1his research de·elops those two topologies, but it would be possible to analyse the potential
íor transíormation under the íollowing three more topologies:
• •• • 1opology C: when an old, abandoned pit or quarry is a·ailable this one could take the role
oí the existing reser·oir in 1A abo·e, or used as new ,e.g. upper, reser·oir ií geography so
• •• • 1opology D: pump-back` in an existing 2-dam system a penstock and a pump are added
to send water back írom the lower reser·oir to the upper one.
• •• • 1opology L: the lower reser·oir is the sea and the upper reser·oir is build abo·e cliíís
close to the sea. 1his topology was implemented íor the íirst time in Okinawa, Japan |52[.
1his paper will not extent on the latter three topologies nor in other ideas proposed by
1hroughout this document the terms dam` and reser·oir` are used as equi·alents.
J.2 Introduction to countries selected as case studies
1urkey and Croatia were selected as the countries that will be analysed íor the potential
transíormation oí dams to PlS using the methodology deíined in Section 0.
1he total a·erage electricity generation in Croatia is 12 500 G\h per year, oí which
hydropower plants contribute 5 ¯00 G\h ,Lurostat a·erage írom 2005 to 2009
,, and account
íor 2 0¯6 M\ oí installed power. 1otal annual consumption reaches around 18 000 G\h and
thereíore local hydropower plants supply 31 ° oí Croatian consumption. Croatia currently has
3 PlS plants in operation: RlL Velebit |¯[ ,generation capacity 2¯6 M\ ,2x138,, pumping
Calculation írom Lurostat tables nrg_105a and nrg_10¯2a in Lnergy Statistics ,nrg_10, - supply, transíormation,
capacity 240 M\ ,2x120,,, luzine ,generation capacity 4.6 M\, pumping capacity 4.8 M\,
and Lepenica |8[
,generation capacity 1.14 M\, pumping capacity 1.25 M\,.
Croatia has set a target to increase the share oí electricity írom renewable energy sources, and
this includes 1200 M\ oí wind by 2020. lor the purpose oí pro·iding incenti·es, Croatia does
not take into account RLS electricity írom large hydropower ,capacity oí 10 M\ or more, |9[.
1he high penetration oí hydropower in Croatia and their commitment to a 20 ° share oí
renewable energy in its total consumption by 2020 |10[, along with its numerous electricity
interconnections, make Croatia a suitable candidate íor this study. Croatia is directly
interconnected to Slo·enia, Bosnia and lerzego·ina, Serbia and lungary, which creates the
potential to store surplus wind generation írom these neighbouring countries.
1he total a·erage electricity generation in 1urkey is 195 000 G\h per year, oí which
hydropower plants contribute 38 000 G\h ,Lurostat a·erage írom 2005 to 2009
, and account
íor 14 550 M\ |11[ oí installed power at the end oí 2009, thus hydropower plants supply 20 °
oí 1urkish demand. DSI |12[ suggests that only 35 ° oí estimated economic potential íor
hydropower is utilised in 1urkey, and the 1urkish go·ernment hopes that hydropower capacity
will expand to 35 000 M\ by the year 2020 |13[. 1able 1 presents a broad o·er·iew oí dam
projects in operation and under construction.
Projects In operation Under construction Planned
Large dams 260 63
Small dams 413 83
lydroelectric plants ,no., 1¯2 148 1 418
lydroelectric capacity ,M\, 13 ¯00 8 600 22 ¯00
Annual a·erage generation ,G\h,yr,
48 000 20 000 ¯2 000
1able 1: list oí existing dams and dams under construction in 1urkey in 2008 |61[
1urkey is a suitable country íor this study due to the large number oí dam sites, and thus the
large number oí potential transíormation sites, and oí its target to increase the country`s
installed wind power capacity to 20 000 M\ by the year 2023 |14[. Both íactors make oí the
1urkish case representati·e oí the potential transíormation oí dams to PlS in some other
J.3 Methodology íor
1he methodology íor the
transíormation oí existing reser·oirs
into PlS, under both 1A and 1B, is
set out below. A high le·el
methodology ílow chart is described
in ligure 1 which shows the ílow oí
decisions which need to be
implemented. In the subsequent
sections each stage oí the
methodology is described and
implementation details are pro·ided.
Note the discrepancies with Lurostat data íor generation
It has to be noted that the íinal parameters used were e·en more restricti·e, see e.g. 1able 10
Initial physical characteristics for transformation
Minimum size oí existing reser·oir ,m
, 1 million
,or, minimum hydropower capacity ,M\, 1
Max distance between reser·oirs ,dams, ,km, 5
Minimum head ,m, 150
1opology A, assumed new reser·oir suríace , m°, ¯0 000
Minimum distance írom inhabited sites to new dam
Minimum distance írom existing transportation
inírastructure to new dam inírastructure ,m,
Minimum distance to a UNLSCO site ,km, 5
Potential site should not be in a Natura 2000 area
Maximum distance to suitable grid connection ,km, 20
1able 2: summary oí parameters used íor analysing the potential
J.3.J 1ransíormation topography, physical characteristics and assumptions
lirst, the topography and physical characteristics íor transíormation must be deíined, and
assumptions must be made on distances to key íeatures ,e.g. inhabited sites,, sources oí data,
etc. Lach site
can then be
assessed in a
is presented in
on the major
ha·e a water
than 1 million
m´ and ha·e a
1 M\. 1he
either be the
upper or the lower reser·oir oí a potential transíormation site under 1A. In the case oí
topology B both reser·oirs ha·e to already exist and the assumption is that the penstock,
generation and pumping equipment must be added.
Dam types. It was assumed that all types oí dams are suitable íor transíormation
regardless oí the dam construction type ,rock-íill, concrete etc.,.
ligure 1: methodology ílow chart. 1he dashed line at grid transmission capacity exists
indicates elements not applied because oí lack oí appropriate data
Constra|nts 1heoretico/ potentio/
between 1A & 18
Crld Lransmlsslon llnes presenL
Crld Lransmlsslon capaclLy exlsLs
uotobose of reservoirs
Mln. reservolr/MW slze
ulsLance beLween dams
Mln. new reservolr slze (1A)
1he distance between the existing and the prospective reservoirs under 1A, or
between the two reser·oirs under 1B, must not be greater than 5 km. Ií greater, the
transíormation to PlS will be deemed not ·iable. 1his distance is normally measured
between their dams.
Head. 1he head íor a transíormation to PlS should be 150 m or greater, ií not then the
transíormation to PlS will be deemed to be not ·iable.
Volume and surface of a new reservoir. 1he analysis must assume a standard area íor
the size oí a prospecti·e new reser·oir in order íor constraints to be applied. Based on the
requirement oí a minimum ·olume oí 1 million m´, and on an indicati·e reser·oir depth oí
20 m, the resulting minimum indicati·e reser·oir suríace is 50 000 m°. In order to take into
account embankments and other inírastructure, a minimum indicati·e size oí any
prospecti·e reser·oir site íor 1A should be ¯0 000 m°.
Human presence. 1he restriction
on inhabited sites is that ií there is
an inhabited area within 200 m oí a
new construction, either a new
reser·oir or the corresponding
penstock, in a transíormation site
then the transíormation to PlS will
be deemed to be not ·iable. ligure 2
illustrates this restriction íor 1A.
Both the dams already exist íor 1B
so the inhabited constraint is that
there should be no generation,
pumping and penstock placed on or
within 200 m oí an inhabited site.
1ransport infrastructure. Ií there is transport inírastructure within 100 m oí a
transíormation site then the transíormation to PlS will be deemed to be not ·iable.
1ransportation inírastructure reíers to public roads, train lines and bridges ,road or rail
Grid infrastructure. Ií the transíormation site is a no hydro-dam then there must be
suitable grid inírastructure within 20 km, ií not the transíormation to PlS will be deemed
to be not ·iable. Application oí this constraint will be subject to public a·ailability oí data,
which may not be the case due to security or other considerations.
J.3.2 Lnergy storage
1he potential energy storage will be analysed íor each potential site, but it will also be used
when merging the results oí 1A and 1B into a global country potential to eliminate o·erlaps.
1he theoretical energy storage a·ailable írom a reser·oir can be expressed as:
L ~ energy storage capacity in \h
_ ~ eííiciency ,in general ranging 0.¯5 to 0.80,
p ~ density ,kg,m
, ,~ 1000 kg,m
g ~ acceleration oí gra·ity ,9.81 m,s
h ~ íalling height, head ,m,
V ~ Volume oí water in the upper reser·oir ,m
Lx|st|ng dam Þotent|a| dam s|te
Area 70 000 m²
ligure 2: 1opology A - minimum distance to inhabited sites
η ρ ∗ ∗ ∗ ∗
V h g
J.3.3 Methodology deíinition
lor each oí topology A and B the íollowing steps and deíinitions apply to a database oí
reser·oirs abo·e the minimum size as in 1able 2:
1heoretical potential: the theoretical potential is the result oí applying in the GIS programme
the restrictions oí minimum reser·oir size, maximum distances between reser·oirs, and
minimum head to e·ery reser·oir in the database.
Physical potential: the theoretical potential will be íiltered íor distances to UNLSCO sites, to
natural reser·es, to transport inírastructure and to inhabited sites to eliminate any sites which
do not meet the deíined speciíication íor a site to be suitable íor transíormation to PlS. 1he
result will be the physical potential íor both 1A and 1B.
Infrastructural potential: when grid maps are a·ailable the íilter oí distance to the grid
connection will be applied to obtain the Inírastructural` potential. lor the time being it was
not possible to obtain data on capacity oí the power grid and thus this aspect was not analysed.
Country potential: both inírastructural potentials ,1A and 1B, will be analysed to eliminate
o·erlaps to realise each country potential. 1he rules oí how the global country potential will be
reached will be deíined in section 1.4.
Lach dam,reser·oir in the country being analysed will be gi·en a unique ID, henceíorth termed
Dam ID`, used to identiíy each dam during the analysis.
1opography. 1he topographical analysis is the íirst stage oí íiltering down the potential
transíormation sites, and uses the distance between dams and minimum head íilters. Lach oí
the reser·oirs will be analysed under 1B to íind out whether another dam is within 5 km oí the
Dam ID and whether the head is greater than 150 m. lor 1A the GIS analysis will check
whether there is a suitable plateau within 5 km oí Dam ID and at least 150 m abo·e Dam ID`s
Dam ID sites passing this phase will ha·e one or more potential .ecova re.erroir associated, which
will constitute the theoretical potential íor transíormation. 1hen all Dam IDs ha·ing passed
this topography íilter will ha·e the íollowing listed constraints applied to them.
Inhabited sites. lor 1A ií there is any inhabited site within 200 m oí the potential site íor a
second reser·oir or oí the direct penstock link between it and Dam ID, this potential reser·oir
is considered unsuitable íor transíormation and the potential site is dropped. 1he question is
then repeated íor any subsequent potential second-dam site íor the same Dam ID, then íor all
the other Dam IDs. lor 1A the íilter is applied to the direct penstock link between two Dam
IDs making up a potential site íor transíormation.
1he íinal application oí this constraint will depend on the public a·ailability oí detailed
iníormation on settlements, but the presence oí settlement cluster,s,, rather than an indi·idual
dwelling, will be deemed to constitute a suííicient le·el oí habitation to apply the constraint in
any particular case.
1ransport infrastructure. 1his part analyses whether there is any transport inírastructure site
within 100 m oí the potential site íor a second reser·oir or oí the direct penstock link between
it and Dam ID, as in the inhabited sites case íor 1A and 1B.
1ransport inírastructure here reíers to public roads, bridges and railways. As in the pre·ious
case ií the answer is yes the potential site is dropped. 1he question is then repeated íor any
subsequent potential site íor the same Dam ID, then íor all the other Dam IDs.
UNLSCO and natural spaces. 1his aspect oí the analysis íocuses on nature conser·ation
sites and archaeological and historic location sites
and more concretely Natura 2000 areas,
those associated to the LUROPARC íederation |15[, included in UNLSCO Biosphere Reser·e
|16[ or \orld leritage lists |1¯[, special areas oí conser·ation ,SAC,, or national parks -most oí
which are already included in the Biosphere Reser·e or LUROPARC lists.
Grid infrastructure. Suitable grid inírastructure must be within 20 km oí Dam ID. \hen Dam
ID is an existing hydropower reser·oir the answer is already yes - howe·er, this approach
ob·iates whether the existing hydropower site could be enlarged as a result oí this analysis
because then a higher-capacity export line could be needed. lor non-hydropower reser·oirs this
constrain is most rele·ant. \hen suitable GIS-shaped iníormation on the capacity oí the grid
transmission lines becomes a·ailable, this íilter could be applied at this stage thus impro·ing the
quality oí the inírastructural potential.
J.4 Country potential
It is assumed that when both generation and pumping equipment is required this will be in the
íorm oí pump-turbines. 1hen only one penstock will be required.
1ransíormation íollowing 1A will always require the construction oí a new reser·oir, pumping,
equipment and the associated penstock,s, and normally, unless it is a pump-back PlS ,not
analysed here,, the installation oí generation equipment. 1ransíormation íollowing 1B will not
require the construction oí a second reser·oir but will require the installation oí generation and
pumping equipment and the associated penstocks. 1able 3 below illustrates the required
modiíication needed to the existing reser·oirs, whether a no-hydro or hydro reser·oir, under
1A and 1B.
1opology A 1opology B
Add new reser·oir \es \es No No
Add generation \es \es
Add pumping \es \es \es \es
Add penstock \es \es \es \es
1able 3: matrix analysis oí the modiíications needed íor the diííerent con·ersion options
Note: - 1he existing generation equipment installed at a hydropower reser·oir would not be
suitable, in most cases, to be used íor transíormation to PlS. 1he reason is that existing
generation equipment was designed íor the head oí the existing hydro scheme. In 1A the new
reser·oir will normally be the upper reser·oir, thus new generation equipment is required. In
1B existing generation equipment will ne·er be suitable gi·en the diííerence in heads, ílows and
capacities. One case under 1A where existing generation equipment may be used is íor pump-
back PlS ,topology D,. In this case, the new reser·oir is constructed directly below the
existing reser·oir, thus the head will be the same as the existing hydropower scheme. lor
pump-back PlS the generation equipment may be replaced with pump turbines
scenario is to install a separate pump unit and related penstock, which would reduce the impact
on the current generation unit while the transíormation is occurring.
1hls llsL of consLralnLs ls from Lhe uk LnvlronmenLal Agency, good pracLlce guldellnes Lo Lhe envlronmenLal agency
handbook, 1he envlronmenLal assessmenL of proposed low head hydropower developmenLs,
Þump-Lurblnes refer Lo a unlL whlch ls reverslble so lL can boLh generaLe and pump and share Lhe same pensLock.
lurLher analysls of each lndlvldual hydropower dam LransformaLlon slLe would be requlred ln order Lo access Lhe
poLenLlal Lo replace currenL generaLlon equlpmenL wlLh pump Lurblne equlpmenL for a pump-back ÞPS.
J.4.J Merging topologies A and B
1he process oí merging topologies A and topology B results into an o·erall country potential
and must eliminate o·erlaps. 1he le·el oí modiíication detailed abo·e to transíorm to PlS will
be taken into account in selecting the preíerred option íor each site. In this study and at this
stage this is done by gi·ing 1B a higher priority than 1A when a gi·en Dam ID results in both
options. 1he reason íor this
choice is that 1B being
based in two existing
dams,reser·oirs, it will only
be necessary to add
generation and pumping
equipment, see note in
1able 3 and the diííerence
in cost that is discussed in
GIS shapeíiles ,layers, will
be required to build up a
íull country map íor each oí
the proposed countries.
Digital terrain maps will be
used to pro·ide topographic
Additional data will include:
country maps ,ri·ers, water
bodies,, topography layer
,ele·ation data,, inhabited
sensiti·ity ,Natura 2000,
SAC, and national parks,,
and electricity grid both at
distribution and transmission
2 Database of dams and hydropower schemes
A suitable database oí dams with or without a linked hydropower scheme was necessary íor
íinding the potential íor transíormation. Initial screening oí public and pri·ate databases
showed that none oí them included the íull range oí details needed íor this project, the most
problematic oí which were the geographical coordinates and ele·ation, and the reser·oir
capacity. In this assessment is included a íocus on reser·oirs that either ha·e a water storage
capacity larger than 1 million m´ or ha·e a turbine capacity oí 1 M\ and a geographical
co·erage including the members oí the Luropean Union ,LU, |18[, the Luropean lree 1rade
Area |19[, the \estern Balkans, LU candidate countries and LU potential candidate countries
country Mop li/e
£/evotion uoto li/e
Cvera|| country map made up of shapef||e |ayers
ligure 3: GIS shapeíile layers to íorm an o·erall country map
1his section describes the sources and methodology used to build the database and discusses
the reasons why some data were una·ailable. 1he íields in the database are also explained.
2.J Countries included in database
Country Status Country Status
Albania Potential Candidate Country
Lat·ia Member State oí LU
Austria Member State oí LU
Liechtenstein Luropean lree 1rade Area
Belgium Member State oí LU
Lithuania Member State oí LU
Bosnia-lerzego·ina Potential Candidate Country
Luxemburg Member State oí LU
Bulgaria Member State oí LU
Malta Member State oí LU
Croatia Candidate Countries
Montenegro Potential Candidate Country
Cyprus Member State oí LU
Netherlands Member State oí LU
Czech Republic Member State oí LU
Norway Luropean lree 1rade Area
Denmark Member State oí LU
Poland Member State oí LU
Lstonia Member State oí LU
Portugal Member State oí LU
linland Member State oí LU
Romania Member State oí LU
lormer \ug. Rep. Oí Macedonia Candidate Countries
Serbia Potential Candidate Country
lrance Member State oí LU
Slo·akia Member State oí LU
Germany Member State oí LU
Slo·enia Member State oí LU
Greece Member State oí LU
Spain Member State oí LU
lungary Member State oí LU
Sweden Member State oí LU
Iceland Candidate Countries
Switzerland Luropean lree 1rade Area
Ireland Member State oí LU
1urkey Candidate Countries
Italy Member State oí LU
Ukraine Other Luropean Country
Koso·o ,Under Unscr 1244, Potential Candidate Country
United Kingdom Member State oí LU
1able 4: list oí countries included in the database and their LU status
1otal gross electricity
Albania 4 200 4 250 LIA
Austria 36 34¯ 65 500 55 ° Lurostat
Belgium 362 86 5¯¯ 0 ° Lurostat
Bosnia-lerzego·ina 5 050 12 260 41 ° LIA
Bulgaria 3 568 44 636 8 ° Lurostat
Croatia 5 446 12 365 44 ° Lurostat
Cyprus 0 4 ¯45 0 ° Lurostat
Czech Republic 2 261 84 664 3 ° Lurostat
Denmark 25 39 349 0 ° Lurostat
Lstonia 21 10 6¯¯ 0 ° Lurostat
linland 14 142 ¯¯ 884 18 ° Lurostat
lormer \ug. Rep. Oí Macedonia 1 240 6 250 20 ° LIA
lrance 5¯ 9¯3 5¯4 150 10 ° Lurostat
Germany 20 339 632 803 3 ° Lurostat
Greece 4 242 62 014 ¯ ° Lurostat
lungary 203 3¯ 900 1 ° Lurostat
Iceland ¯ 156 9 308 ¯¯ ° Lurostat
Ireland ¯48 2¯ 840 3 ° Lurostat
Italy 36 8¯5 312 ¯09 12 ° Lurostat
Koso·o ,Under Unscr 1244,
Lat·ia 2 966 4 960 60 ° Lurostat
Lithuania 418 13 ¯96 3 ° Lurostat
Luxemburg 113 4 00¯ 3 ° Lurostat
Malta 0 2 140 0 ° LIA
Montenegro 1 640 2 ¯40 60 ° LIA
Netherlands 101 102 855 0 ° Lurostat
1otal gross electricity
Norway 132 05¯ 134 862 98 ° Lurostat
Poland 2 18¯ 158 551 1 ° Lurostat
Portugal 8 156 4¯ 210 1¯ ° Lurostat
Romania 1¯ 931 62 185 29 ° Lurostat
Serbia 10 090 34 610 29 ° LIA
Slo·akia 4 382 29 960 15 ° Lurostat
Slo·enia 3 584 15 419 23 ° Lurostat
Spain 24 044 303 0¯5 8 ° Lurostat
Sweden 6¯ 441 150 205 45 ° Lurostat
Switzerland 33 368 65 146 51 ° Lurostat
1urkey 38 232 182 058 21 ° Lurostat
Ukraine 11 590 180 940 6 ° LIA
United Kingdom 4 943 395 501 1 ° Lurostat
1able 5: o·er·iew oí the percentage oí hydropower generation. Sources: Lurostat |6[, LIA |21[
1able 5 presents an o·er·iew oí the percentage oí electricity generated by hydropower, on
a·erage 2005 to 2008 when the source is Lurostat |6[, in the countries included in the database.
Data írom LIA |21[ is 4- or 5-yr a·erage and total generation is net instead oí gross, and the
percentages are thereíore calculated on a diííerent basis yet they all are representati·e. It is clear
írom this table that hydropower plays an important role in the electricity generation portíolios
oí most countries in Lurope.
2.2 Data sources
As pointed out abo·e, error-íree global data sets describing reser·oir characteristics and
geographical distribution are largely incomplete. 1he best and most comprehensi·e global dam
database, the \orld Register oí Dams, is compiled by the International Commission on Large
Dams ,ICOLD, |22[ and currently lists more than 33 000 records oí large reser·oirs and their
attributes. lowe·er, this database is not geo-reíerenced thus its use is limited íor this project.
Despite this and because it is the most comprehensi·e database a·ailable, it íorms the primary
data source íor this study.
Lach data source that was used to compile the database will be described below. lor each oí
the countries a justiíication oí eííort taken to gather the data will be made in the cases where
the dataset is not íully complete.
2.2.J International Commission oí Large Dams ,ICOLD,, \orld Register oí Dams
1be ívtervatiovat Covvi..iov ov íarge Dav. ;íCOíD) i. a vov·gorervvevtat ivtervatiovat orgavi.atiov
rbicb proriae. a forvv for tbe e·cbavge of /vorteage ava e·perievce iv aav evgiveerivg. 1be Orgavi.atiov
teaa. tbe profe..iov iv ev.vrivg tbat aav. are bvitt .afet,, efficievtt,, ecovovicatt,, ava ritbovt aetrivevtat effect.
ov tbe evrirovvevt. ít. origivat aiv ra. to evcovrage aaravce. iv tbe ptavvivg, ae.igv, cov.trvctiov, operatiov,
ava vaivtevavce of targe aav. ava tbeir a..ociatea cirit ror/., b, cottectivg ava ai..evivativg reteravt
ivforvatiov ava b, .tva,ivg retatea tecbvicat qve.tiov.. ´ivce tbe tate .i·tie., focv. ra. pvt ov .vb;ect. of cvrrevt
covcerv .vcb a. aav .afet,, vovitorivg of perforvavce, reavat,.i. of otaer aav. ava .pittra,., effect. of ageivg
ava evrirovvevtat ivpact. More recevtt,, ver .vb;ect. ivctvae co.t .tvaie. at tbe ptavvivg ava cov.trvctiov
.tage., barve..ivg ivtervatiovat rirer., ivforvatiov for tbe pvbtic at targe, ava fivavcivg. ,Background
description írom ICOLD |22[,
ICOLD produce the world register oí dams. 1his database is compiled by ICOLD by accessing
data through the ICOLD representati·es oí the member countries, is a comprehensi·e database
oí hydropower and no-hydropower reser·oirs and pro·ides detailed iníormation on each
1he ICOLD database has some drawbacks íor this project mainly because oí the co·erage and
accuracy oí the data pro·ided. 1he primary oí these drawbacks is that it does not pro·ide geo-
reíerencing or ele·ation iníormation. 1here are also issues with the accuracy oí the storage
capacity and area oí some oí the reser·oirs. 1he GRanD project team ha·e also highlighted this
as an issue.
Being the most complete source oí data on dams globally the world register oí dams íorms the
primary source oí data íor this database. \here data are incomplete secondary sources in each
country are also utilised to íill these data gaps.
2.2.J 1he Global Reser·oir and Dam ,GRanD, database |23[
1o aaare.. gap. ava .bortcovivg. iv gtobat aav aataba.e., tbe Ctobat !ater ´,.tev Pro;ect ;C!´P) 21, a
;oivt pro;ect of tbe íartb ´,.tev ´cievce Partver.bip ;í´´P), ivitiatea av ivtervatiovat effort to cottate tbe
e·i.tivg aav ava re.erroir aata .et. ritb tbe aiv of proriaivg a .ivgte, geograpbicatt, e·pticit ava retiabte
aataba.e for tbe .cievtific covvvvit,: tbe Ctobat Re.erroir ava Dav ;CRavD) aataba.e. 1be aeretopvevt of
CRavD privarit, aivea at covpitivg tbe araitabte re.erroir ava aav ivforvatiov; correctivg it tbrovgb
e·tev.ire cro..·ratiaatiov, error cbec/ivg, ava iaevtificatiov of avpticate recora., attribvte covftict., or
vi.vatcbe.; ava covptetivg vi..ivg ivforvatiov frov ver .ovrce. or .tati.ticat approacbe.. 1be aav. rere
geo.patiatt, referevcea ava a..igvea to pot,gov. aepictivg re.erroir ovttive. at bigb .patiat re.otvtiov. !bite tbe
vaiv focv. ra. to ivctvae att re.erroir. ritb a .torage capacit, of vore tbav 0.1 /v, vav, .vatter re.erroir.
rere aaaea if aata rere araitabte. 1be cvrrevt rer.iov 1.1 of CRavD covtaiv. ó,ºó2 recora. of re.erroir. ava
tbeir a..ociatea aav., ritb a cvvvtatire .torage capacit, of ó,1·¨ /v. ;´ovrce: CRavD tecbvicat
1able 6 presents the number oí ICOLD dams that GRanD ha·e geo-reíerenced, and which
contributed a total oí 21° oí the geo-reíerences in the dams in our database. 1he ele·ation
abo·e mean sea le·el ,AMSL, oí these geo-reíerenced dams is also a·ailable and was added to
1able 6: percentage oí ICOLD dams that ha·e been geo-reíerenced by GRanD
2.2.2 Google Larth manual geo-reíerencing oí Croatia and 1urkey
A complete set oí geo-reíerenced dams was required íor the potential transíormation countries,
Croatia and 1urkey, this was a total oí 40 dams in Croatia and 6¯1 dams in 1urkey. As
Albania 308 5 2 ° Lat·ia 5 3 60 °
Austria 168 22 13 ° Liechtenstein 2 0 0 °
Belgium 1¯ 5 29 ° Lithuania 20 2 10 °
Bosnia-lerzego·ina 31 9 29 ° Luxemburg ¯ 1 14 °
Bulgaria 181 46 25 ° Malta 0 0 0 °
Croatia 40 8 20 ° Montenegro 10 3 30 °
Cyprus 52 4 8 ° Netherlands 12 8 6¯ °
Czech Republic 126 35 28 ° Norway 335 125 3¯ °
Denmark 10 0 0 ° Poland 69 29 42 °
Lstonia 2 0 0 ° Portugal 151 53 35 °
linland 56 19 34 ° Romania 246 80 33 °
l\RO Macedonia 18 0 0 ° Serbia 68 19 28 °
lrance 59¯ 114 19 ° Slo·akia 50 16 32 °
Germany 30¯ 60 20 ° Slo·enia 3¯ 2 5 °
Greece 61 19 31 ° Spain 126¯ 252 20 °
lungary 16 4 25 ° Sweden 194 49 25 °
Iceland 25 6 24 ° Switzerland 159 38 24 °
Ireland 18 4 22 ° 1urkey 6¯1 101 15 °
Italy 549 8¯ 16 ° Ukraine 22 9 41 °
Koso·o 2 0 0 ° United Kingdom 515 89 1¯ °
1otal 6424 1326 21 °
presented in 1able 6 GRanD only pro·ides geo-reíerencing íor 20 ° and 15 ° oí dams in
Croatia and 1urkey respecti·ely and thus the remaining dams were geo-reíerenced manually
using Google Larth by ·isually searching there using the nearest town name in the ICOLD
database, then when a dam was located close to nearest town` it was ·eriíied ·isually in
Google Larth where possible by comparing it with the picture oí the dam ií a·ailable. 1his task
was extremely time consuming
and although all possible care was taken there may be errors
present due to lack oí iníormation a·ailable when ·isually recognising the dams.
1his methodology was the same one used by the GRanD project íor geo-reíerencing dams in
their database. \hen sites are geo-reíerenced, the ele·ation AMSL in metres can then be
calculated through ArcGIS using SR1M ele·ation data. 1his option was preíerred to obtaining
ele·ation directly írom Google Larth íor consistency reasons, as the next modelling steps will
be based on SR1M ele·ation data.
2.2.3 Shuttle Radar 1opography Mission ,SR1M, ele·ation data
Remotely sensed ele·ation data were obtained írom the Shuttle Radar 1opography dataset: 1be
CCí.R·C´í CeoPortat i. abte to proriae ´R1M ·0v aigitat eteratiov voaet. ;DíM) for tbe evtire rorta.
1be ´R1M aigitat eteratiov aata proriaea ov tbi. .ite ba. beev proce..ea to fitt aata roia., ava to facititate it.
ea.e of v.e b, a riae grovp of potevtiat v.er.. 1be ´R1M ·0 v DíM´. bare a ;horizontal) re.otvtiov of
·0v at tbe eqvator, ava are proriaea iv vo.aic/ea : aeg · : aeg tite. for ea., aorvtoaa ava v.e. .tt are
proavcea frov a .eavte.. aata.et to attor ea., vo.aic/ivg. 1be.e are araitabte iv botb .rcívfo .´Cíí ava
Ceo1iff forvat to facititate tbeir ea.e of v.e iv a rariet, of ivage proce..ivg ava Cí´ appticatiov.. 2:
SR1M`s ·ertical resolution is approximately 10 m depending on location, and SR1M data are
used to calculate the ele·ation AMSL by importing the shapeíile geo-reíerenced in Google
Larth into ArcGIS. 1hen, by running the extract tool` in ArcGIS the geo-reíerenced dams
combine with the SR1M ele·ation to calculate the ele·ation oí each dam. SR1M ele·ation was
·alidated with Google Larth data resulting in ·ery consistent íigures
2.2.4 Regulators and transmission system operators
An area where there was a lack oí data in the primary data source was the mean annual energy
,G\h,year, generated írom hydropower plants. 1he electricity regulators and transmission
system operators oí each oí the countries where data were absent were contacted to request
this data, ií publicly a·ailable. lrom the replies we recei·ed some new data which was
pre·iously absent, but many oí the responses coníirmed our initial íeeling that this data are
considered commercially sensiti·e and are not publicly a·ailable. 1he electricity regulators and
transmission system operators oí the countries in the database are listed in 1able ¯.
Country Llectricity Regulator 1ransmission System Operator
Albania Albanian Llectricity Regulatory Authority OS1 sh.a
APG-Austrian Power Grid AG Austria Lnergie-Control Gmbl ,L-Control,
Belgium Commission pour la Régulation de l'Llectricité et du Gaz
Llia System Operator SA
Bosnia-Herzegovina State Llectricity Regulatory Commission ,SLRC, Neza·isni operator susta·a u Bosni i lercego·ini
Bulgaria State Lnergy & \ater Regulatory Commission ,SL\RC, Llectroenergien Sistemen Operator LAD
Croatia Croatian energy regulatory agency lLP-Operator prijenosnog susta·a d.o.o.
Cyprus Cyprus Lnergy Regulatory Authority ,CLRA, Cyprus 1ransmission System Operator
Czech Republic Lnergeticky Regulacní Úrad ,LRÚ, CLPS a.s.
Denmark Lnergitilsynet - Danish Lnergy Regulatory Authority ,DLRA, Lnerginet.dk
Lstonia Lstonian Competition Authority - Lnergy Regulatory Dept
Iinland 1he Lnergy Market Authority ,LMV, lingrid OyJ
Iormer Yug. Rep. Of
Lnergy Regulatory Commission oí the Republic oí Macedonia Macedonian 1ransmission System Operator AD
We estimated at 30 – 40 sites being geo-referenced per day for one person not knowing the native language.
Country Llectricity Regulator 1ransmission System Operator
Irance Commission de Régulation de l'Lnergie ,CRL, Réseau de 1ransport d'Llectricité
LnB\ 1ransportnetze AG
1enne1 1SO Gmbl
Germany lederal Network Agency íor Llectricity
50lertz 1ransmission Gmbl
Greece Regulatory Authority íor Lnergy ,PAL , RAL, lellenic 1ransmission System Operator S.A.
Hungary lungarian Lnergy Oííice ,MLl , lLO, MAVIR Magyar Villamosenergia-ipari Át·iteli
Rendszeriranyító Zartkoruen Mukodo Rész·énytarsasag
Iceland National Lnergy Authority Landsnet hí
Ireland Commission íor Lnergy Regulation ,CLR, LirGrid plc
Italy Autorita per l'Lnergia Llettrica e il Gas ,ALLG, 1erna - Rete Llettrica Nazionale SpA
Lnergy Regulatory Oííice KOS11
Latvia Public Utilities Commission ,PUC, AS Augstsprieguma tIkls
Lithuania National Control Commission íor Prices and Lnergy ,NCC, LI1GRID AB
Luxemburg Institut Luxembourgeois de Régulation ,ILR, Creos Luxembourg S.A.
Malta Malta Resources Authority ,MRA,
Montenegro Lnergy Regulatory Agency Oí Montenegro Crnogorski elektroprenosni sistem AD
Netherlands Dutch Oííice oí Lnergy Regulation 1enne1 1SO B.V.
Norway Norwegian \ater Resources and Lnergy Directorate ,NVL, Statnett Sl
Poland 1he Lnergy Regulatory Oííice oí Poland ,LRO, PSL Operator S.A.
Portugal Lnergy Ser·ices Regulatory Authority ,LRSL, Rede Lléctrica Nacional, S.A.
Romania Romanian Lnergy Regulatory Authority ,ANRL, C.N. 1ranselectrica S.A.
Serbia Lnergy Agency oí the Republic oí Serbia JP Llektromreza Srbije
Slovakia Regulatory Oííice íor Network Industries ,RONI, Slo·enska elektrizacna prenoso·a susta·a, a.s.
Slovenia Lnergy Agency oí the Republic oí Slo·enia Llektro Slo·enija d.o.o.
Spain National Lnergy Commission ,CNL, Red Lléctrica de Lspana: S.A.
Sweden Lnergy Markets Inspectorate,LI, Aííärs·erket S·enska Kraítnät
Switzerland Swiss lederal Llectricity Commission LlCom Swissgrid ag
1urkey 1urkish Llectricity 1LIAS
Ukraine National Llectricty Regulatory Comission oí Ukraine National Lnergy Company Ukrenergo
National Grid Llectricity 1ransmission plc
System Operation Northern Ireland Ltd
Scottish and Southern Lnergy plc
United Kingdom Oííice oí Gas and Llectricity Markets ,Oígem,
Scottish Power 1ransmission plc
1able ¯: electricity regulators and transmission system operators oí countries in database ,hyperlinked,
2.2.S Sources oí country speciíic dam iníormation
\here iníormation was not a·ailable írom the primary data sources or írom the regulators and
the 1SO`s on dams, attempts where then made to access data írom other sources. 1able 8
pro·ides a list oí some oí the other sources that were utilised.
Country Secondary data sources
Albania National Agency oí Natural Resources
Austria Lurope Ln·ironment Agency lydro-Austria
Bulgaria Bulgaria Lnergy lolding
Croatia lLP Proiz·odnja d.o.o.
Czech Republic CLZ
Denmark Lnergy Map
Lstonia INlORSL - Lurope
linland Pamilo lydropower in linland
lormer \ug. Rep. Oí Macedonia Llem
lrance COMPAGNIL NA1IONALL DU RlÔNL
Germany R\L http:,,www.industcards.com,ppworld.htm
Iceland Lands·irkjun Mann·it
Lee Catchment llood Risk Assessment and Management
Koso·o ,under UNSCR 1244, Koso·o Lnergy Corporation J.S.C.
Lithuania http:,,saule.lms.lt,main,hidro_e.html http:,,www.kruoniohae.lt,en
Montenegro Llektropri·reda Crne Gore-LPCG lydropower in Montenegro
Serbia Llectric power industry oí Serbia Serbia Lnergy
1urkey General Directorate oí State lydraulic \orks Art·in - Deriner Baraji ·e lLS
1able 8: list oí secondary sources
2.3 Database íields
1his section will describe each íield in the database and explain any abbre·iations used, ií
Dav vave. 1he dam name íield contains the primary name oí the dam. Names are gi·en in
íorms with Latinised character sets.
.ttervatire aav vave. Ií a dam is known by two names or has an alternati·e name then it will be
listed in this íield. Lxamples include the íollowing dams in Austria
Dam name Other dam name
Zadeje Vau Dejes
Rirer. 1he ri·er íield contains the name oí the ri·er that the dam is constructed on.
^eare.t torv. 1he nearest town íield contains the name oí the town nearest to the dam site. 1his
data were oí particular importance when manually geo-reíerencing the dams using Google
´tate,prorivce,covvt,. 1he state,pro·ince,county íield is the secondary location, i.e. second
administrati·e entity below country le·el. \hiche·er íormat is applicable to the country in
question will be entered in this íield.
íteratiov ;v). 1he ele·ation oí the dam crest abo·e mean sea le·el ,AMSL, in metres, írom the
GRanD database and SR1M data as detailed in section 2.2.
íatitvae ava tovgitvae ;aegree. ava aecivat.). 1he latitude and longitude íields contain the
latitude,longitude, in degrees and decimal, oí approximately the centre oí dam.
Dav beigbt ava tevgtb ;v). 1he dam height and length íields contain the height,length oí the dam
structure in meters.
Dav rotvve. 1he dam ·olume is included in two íields with diííerent units, one in thousands oí
cubic metres, ,1000 m´, -a unit consistent with ICOLD-, and the other in m´ in order to a·oid
any coníusion caused by the ICOLD unit 1000 m´. 1his íield is not
Re.erroir capacit,. 1he reser·oir capacity is as well included in two íields with diííerent units, one
in thousands oí cubic meters, ,1000 m´, -a unit consistent with ICOLD-, and the other in m´ in
order to a·oid any coníusion caused by the ICOLD unit 1000 m´.
Re.erroir area. 1he reser·oir area is as well included in two íields with diííerent units, one in
thousands oí square metres ,1000 m°, -a unit consistent with ICOLD-, and the other in m
order to a·oid any coníusion caused by the ICOLD unit 1000 m
ítectric iv.tattea capacit, ,M\,. 1his íield is populated when the dam in question is operated as a
Meav avvvat everg, ,G\h,year,. 1he mean annual energy ,G\h,year, íield is populated when
the dam in question is operated as a hydropower plant. It was íound that this data are diííicult
to acquire íor indi·idual hydropower plants due to coníidentiality requirements within many oí
the markets. \here publicly a·ailable the data were added.
Dav .tatv.. 1he dam status íield deíines ií the dam structure has been changed o·er its liíetime.
1he abbre·iations used are: A abandoned, l heightened, L lowered, U unchanged, R rebuilt, C
Year of covptetiov. \ear that the dam came into operation whether it is a hydro or no-hydro dam.
Ií the dam was changed or repowered in its liíetime the re-powering date will appear as íollows
íor example, 19xx,xx. 1he 19xx is the original completion date, ,xx is the date oí change or
Re.erroir pvrpo.e. lield deíining how the reser·oir is being used. 1he abbre·iations are: l~
lPP, S~ water supply, C~ ílood control, I~ irrigation, N~ na·igation, R~ recreation, l~ íish
breeding, X~ others. Ií a reser·oir has more than one purpose the letters will be listed on aíter
another without space or comma. 1he priority oí use íor reser·oirs with more than one
purpose is deíined by the order in which they appear. lor example ií a reser·oir has it purpose
listed as lCI` this means that its main use is hydropower, its second main use is ílood
control, then irrigation. 1hese codes íollow the ICOLD con·ention.
Orver. Name oí the company or organisation who is the owner oí the dam, when known.
Cov.vttavt,covtractor. Name oí the company or organisation that designed,built the dam.
A íield íor obser·ations ,Note, was also included.
2.4 Disclaimer oí warranty and citations oí data utilised in database
ICOLD ,International Commission on Large Dams,. 1998-2009. \orld Register oí Dams.
Version updates 1998-2009. Paris: ICOLD. A·ailable online at www.icold-cigb.net. ICOLD
,International Commission on Large Dams,. 1998-2009. \orld Register oí Dams. Version
updates 1998-2009. Paris: ICOLD. A·ailable online at www.icold-cigb.net.
1his database incorporates data írom the GRanD database which is copyright oí the Global
\ater System Project ,2011,. GRanD is described in íurther detail by Lehner et al. |26[
3 Potential for transformation to PHS
1his section applies the methodology detailed in section 1 to the data described in section 2 íor
Croatia and 1urkey, by means oí a geographical iníormation system ,GIS,, ArcGIS, to describe
and quantiíy two potentials aíter topology A ,1A, and topology B ,1B,. 1he section details the
data and data íiles required to carry out the analysis, the design oí the GIS model, the design oí
the diííerent scenarios and the results oí the scenarios.
A transíormation site is the original dam under examination with a potential reser·oir site
which will create a new PlS plant.
3.J Data required and data processing
1o build up a GIS map the data must be íirst gathered and then con·erted into a usable íormat.
In order to describe and quantiíy the potential íor transíormation to PlS a number oí types oí
data are required as detailed below:
3.J.J Coordinate system
A geographic coordinate system is a reíerence system that uses latitude and longitude to deíine
the locations oí points on the suríace oí a sphere or spheroid. A geographic coordinate system
deíinition includes a datum, prime meridian, and angular unit
Unlike a geographic coordinate system, a projected coordinate system is deíined on a ílat, two-
dimensional suríace with constant lengths, angles, and areas across the two dimensions - it is
always based on a geographic coordinate system that is based on a sphere or spheroid.
1he modelling process oí describing and quantiíying the potential íor transíormation to PlS
requires the GIS data to be in projected coordinate system íormat. 1his is because the analysis
is based around the ArcGIS slope tool, which in order to calculate the slope, requires all inputs
to be in the same íormat. In this case the slope tool requires all data to be in metres and thus a
projected coordinate system is required. All data are downloaded in GCS_\GS_1984
coordinate system. 1hese data are then con·erted to projected coordinate system using
ArcGIS` project` tool ,Data management tools,Projections and transíormations,.
See Wikipedia and ESRI (www.esri.com) for further definitions
Figure 4: SRTM elevation data download map
Lach country may cross many U1M zones so when projecting one U1M zone must be
selected. 1he projected coordinate system selected íor Croatia is
\GS_1984_U1M_Zone_34N, and íor 1urkey is \GS_1984_U1M_Zone_36N
3.J.2 Shuttle Radar 1opography Mission ,SR1M, ele·ation data
1he ele·ation data used íor the analysis is processed SR1M 90m digital ele·ation data. 1hese
data were obtained írom the Luropean Commission`s Joint Research Centre website,
http:,,srtm.jrc.ec.europa.eu, and can be downloaded directly írom
1he SR1M data are a·ailable íor the whole world, which is broken down into 1¯28 blocks. Due
to the data being a·ailable in blocks, the country oí interest may co·er more than one oí these
blocks. Lach block is downloaded and then merged into one raster layer using ArcGIS` Mosaic
tool ,Raster Dataset,Mosaic to New Raster,, to íorm one ele·ation íile íor each country.
ligure 5: methodology íor creating a single SR1M raster ele·ation íile
1hen the mosaicked geographic coordinate system raster íile must be con·erted to projected
coordinate system using the project raster tool` which is in the ArcGIS toolbox Data
Management,Projections and 1ransíormations,Raster,Project Raster`.
1his new mosaicked layer will not detail the political boundaries oí the country oí interest. 1his
step uses a layer with the political borders oí the country as a cookie cutter` to select only the
SR1M data that in within the political borders. 1his process is executed using the ArcGIS
extract tool ,Spatial Analysis 1ools,Lxtraction,Lxtract by Mask,.
3.J.3 Political borders
1he political borders layer illustrates the shape oí the border oí the country in question. 1he
layer íile was obtained írom DIVA-GIS |2¯[ in geographic coordinate system íormat. 1he layer
was con·erted to projected coordinate system using the project tool` which is in the ArcGIS
toolbox Data Management,Projections and 1ransíormations,leature,Project`.
3.J.4 Dam locations
As explained abo·e, the methodology íor geo-reíerencing dam locations was based on creating
a database oí all dams with a capacity abo·e one million m
or 1 M\ oí installed hydropower
capacity. 1he database was composed oí ICOLD data which were then added geographical
Mosaic SRTM files
Extract by country
Final SRTM Raster
Convert to projected
coordinate system ArcGIS
reíerences ,latitude, longitude, írom the GRanD database íor 15 - 20 ° oí the Croatian and
1urkish dams. 1he rest oí the latitude and longitude data were obtained manually by locating
the dams in Google Larth based on the ICOLD name oí the dam - the same methodology
used to populate the GRanD database with geo-reíerences. 1o put into context the extent oí
the manual work around 32 dams in Croatia and 5¯0 dams in 1urkey were geo-reíerenced
manually. 1his task was extremely time consuming
and although all possible care was taken to
reíerence each dam correctly there may be errors present, due to lack oí iníormation a·ailable
when ·isually recognising the dams. 1he methodology íor locating these dams in Google Larth
was as íollows.
Lach dam was searched íor based on the nearest town` íield
in the database using Google Larth.
\hen a dam was located close to nearest town`, it was
·eriíied ·isually in Google Larth where possible, by comparing
it with the picture oí the dam ií a·ailable on the 1urkish DSI
All the dam locations were recorded in a Google Larth KML íile.
In order to use the dam locations in ArcGIS the Google Larth
KML íile was con·erted to an ArcGIS shapeíile. lor this, a script
that con·erts Google Larth KML íiles to shape íiles was obtained
írom LSRI |28[. 1his script is added to the ArcGIS toolbox and the
íile is con·erted to a shapeíile
Reser·oir capacity data were imported írom the database through
the import íeatures oí ArcGIS. Latitude and longitude coordinates
were added to ArcGIS shapeíile attributes table using the add X\
coordinates` tool, which is in the ArcGIS toolbox Data
Management, leature,Add X\ Coordinates`.
1he ele·ation írom the SR1M raster was extracted and added to
the dam location attributes table using the Lxtract Values to
Points` tool, which is in the ArcGIS toolbox Spatial Analysis
1ools,Lxtraction,Lxtract Values to Points`.
1he dam locations geographic coordinate system shapeíile íile was
then con·erted to projected coordinate system using the project
raster tool` which is in the ArcGIS toolbox Data
Management,Projections and 1ransíormations,Raster,Project
3.J.S CORINL land co·er ,CLC, 2006 100m ·ersion 13 ,02,2010,
CLC is a map oí the Luropean en·ironmental landscape based on interpretation oí satellite
images. It pro·ides comparable digital maps oí land co·er íor each country íor much oí
Lurope. 1his is useíul íor en·ironmental analysis and íor policy makers. CORINL stands íor
Cooraivatiov of ívforvatiov ov tbe ívrirovvevt. 1he LU established CORINL in 1985 to create pan-
Luropean databases on land co·er, biotopes ,habitats,, soil maps and acid rain.
1he Luropean Ln·ironment Agency, in conjunction with the Luropean Space Agency, the
Luropean Commission and Member States produced an update oí the Luropean CLC database
as part oí a last 1rack Ser·ice on Land as part oí the Ctobat Movitorivg for ívrirovvevt ava
´ecvrit, ,GMLS, initiati·e. 1his update in·ol·ed:
Locate remaining dams
using Google Earth
Convert KML to SHP
Add XY coordinates
Add dam elevations
Final dam shape file
Add reservoir capacity
Add the geo-referenced
dams from GRanD
Convert to projected
Figure 6: geo-referencing dam
1he creation oí a change dataset íor the period 2000-2006 with local interpretation oí
satellite imagery, and
1he application oí this change dataset to the CLC2000 dataset to produce an update oí
the íull in·entory íor 2006 ,the snapshot database,.
1he CLC data íor Croatia and 1urkey were extracted írom the Luropean data site and are used
íor the inhabited areas and road,rail constraint. 1able 9 pro·ides the CORINL grid codes
which identiíy the categories the data are di·ided into.
No. Label J Label 2 Label 3
1 Artiíicial suríaces Urban íabric Continuous urban íabric
2 Artiíicial suríaces Urban íabric Discontinuous urban íabric
3 Artiíicial suríaces Industrial, commercial and transport units Industrial or commercial units
4 Artiíicial suríaces Industrial, commercial and transport units Road and rail networks and associated land
5 Artiíicial suríaces Industrial, commercial and transport units Port areas
6 Artiíicial suríaces Industrial, commercial and transport units Airports
¯ Artiíicial suríaces Mine, dump and construction sites Mineral extraction sites
8 Artiíicial suríaces Mine, dump and construction sites Dump sites
9 Artiíicial suríaces Mine, dump and construction sites Construction sites
10 Artiíicial suríaces Artiíicial, non-agricultural ·egetated areas Green urban areas
11 Artiíicial suríaces Artiíicial, non-agricultural ·egetated areas Sport and leisure íacilities
12 Agricultural areas Arable land Non-irrigated arable land
13 Agricultural areas Arable land Permanently irrigated land
14 Agricultural areas Arable land Rice íields
15 Agricultural areas Permanent crops Vineyards
16 Agricultural areas Permanent crops lruit trees and berry plantations
1¯ Agricultural areas Permanent crops Oli·e gro·es
18 Agricultural areas Pastures Pastures
19 Agricultural areas leterogeneous agricultural areas Annual crops associated with permanent crops
20 Agricultural areas leterogeneous agricultural areas Complex culti·ation patterns
21 Agricultural areas leterogeneous agricultural areas Land principally occupied by agriculture, with
signiíicant areas oí natural ·egetation
22 Agricultural areas leterogeneous agricultural areas Agro-íorestry areas
23 lorest and semi natural areas lorests Broad-lea·ed íorest
24 lorest and semi natural areas lorests Coniíerous íorest
25 lorest and semi natural areas lorests Mixed íorest
26 lorest and semi natural areas Scrub and,or herbaceous ·egetation Natural grasslands
2¯ lorest and semi natural areas Scrub and,or herbaceous ·egetation Moors and heathland
28 lorest and semi natural areas Scrub and,or herbaceous ·egetation Sclerophyllous ·egetation
29 lorest and semi natural areas Scrub and,or herbaceous ·egetation 1ransitional woodland-shrub
30 lorest and semi natural areas Open spaces with little or no ·egetation Beaches, dunes, sands
31 lorest and semi natural areas Open spaces with little or no ·egetation Bare rocks
32 lorest and semi natural areas Open spaces with little or no ·egetation Sparsely ·egetated areas
33 lorest and semi natural areas Open spaces with little or no ·egetation Burnt areas
34 lorest and semi natural areas Open spaces with little or no ·egetation Glaciers and perpetual snow
35 \etlands Inland wetlands Inland marshes
36 \etlands Inland wetlands Peat bogs
3¯ \etlands Maritime wetlands Salt marshes
38 \etlands Maritime wetlands Salines
39 \etlands Maritime wetlands Intertidal ílats
40 \ater bodies Inland waters \ater courses
41 \ater bodies Inland waters \ater bodies
42 \ater bodies Marine waters Coastal lagoons
43 \ater bodies Marine waters Lstuaries
44 \ater bodies Marine waters Sea and ocean
48 NODA1A NODA1A NODA1A
49 Unclassiíied Unclassiíied land suríace Unclassiíied land suríace
50 Unclassiíied Unclassiíied water bodies Unclassiíied water bodies
255 Unclassiíied Unclassiíied Unclassiíied
1able 9: CORINL grid codes
\ater courses ,grid code 40, and water bodies ,grid code 41, íorm the ri·ers and lakes layer
and it was added íor mapping and aesthetics and do not ha·e any input to the model.
Inhabited areas and industrial and commercial units are included írom: continuous urban íabric
,grid code 1,, discontinuous urban íabric ,grid code 2,, and industrial and commercial units
,grid code 3,.
3.J.6 UNLSCO Sites
1be |vitea ^atiov. íavcatiovat, ´cievtific ava Cvttvrat Orgavi¸atiov ;|^í´CO) .ee/ to evcovrage tbe
iaevtificatiov, protectiov ava pre.erratiov of cvttvrat ava vatvrat beritage arovva tbe rorta cov.iaerea to be of
ovt.tavaivg ratve to bvvavit,. 1bi. i. evboaiea iv av ivtervatiovat treat, cattea tbe Covrevtiov covcervivg tbe
Protectiov of tbe !orta Cvttvrat ava ^atvrat íeritage, aaoptea b, |^í´CO iv 1·¨2 1¨.
Croatian cultural sites included in this analysis are: Lpiscopal Complex oí the Luphrasian
Basilica in the listoric Centre oí Porec ,199¯,, listoric City oí 1rogir ,199¯,, listorical
Complex oí Split with the Palace oí Diocletian ,19¯9,, Old City oí Dubro·nik ,19¯9,, Stari
Grad Plain ,2008,, and the Cathedral oí St James in Sibenik ,2000,. 1he only natural reser·e site
included is the Plit·ice Lakes National Park ,19¯9,.
1urkish cultural sites included the Archaeological Site oí 1roy ,1998,, City oí Saíranbolu
,1994,, Great Mosque and lospital oí Di·rigi ,1985,, lattusha: the littite Capital ,1986,,
lierapolis-Pamukkale ,1988,, listoric Areas oí Istanbul ,1985,, Nemrut Dag ,198¯,, and
Xanthos-Letoon ,1988,. 1wo world heritage mixed nature,cultural sites were also included, the
Goreme National Park and the Rock Sites oí Cappadocia ,1985,, and the lierapolis-Pamukkale
Shapeíiles oí UNLSCO sites were not a·ailable íor Croatia or 1urkey. lowe·er, coordinates oí
the sites were a·ailable on the UNLSCO website |1¯[. 1hese coordinates were added to Google
Larth and then con·erted írom KML to an ArcGIS shapeíile.
In order to use these data as a constraint in the model a buííer oí 5 km is applied to each point.
No transíormation site is permitted within this 5 km area.
3.J.7 Ln·ironmental sensiti·ity
1he en·ironmental aspects were based around Natura 2000, an LU wide network oí nature
protection areas. It is comprised oí Special Areas oí Conser·ation ,SAC,, and oí Special
Protection Areas ,SPAs,. |31[. L·en when Croatia and 1urkey are not yet in the LU-2¯, and
thus are not required to ha·e Natura 2000 designated conser·ation areas, as LU candidate
countries they are required to being establishing Natura 2000 areas. Croatian Natura 2000 data
were obtained írom http:,,Natura2000.dzzp.hr,Natura2000,, howe·er, the authors were
unsuccessíul in íinding Natura 2000 data íor 1urkey.
1he prospecti·e sites should not be in a Natura 2000 area.
3.J.8 1ransport inírastructure
1he rail and road GIS shapeíiles were obtained írom DIVA-GIS |2¯[ in geographic coordinate
system íormat. 1he shapeíiles were con·erted to projected coordinate system using the project
tool` which is in the ArcGIS toolbox Data Management,Projections and
3.J.9 Llectricity grid inírastructure
\e were unable to obtain GIS shapeíiles oí the electricity grid inírastructure oí Croatia or
1urkey írom the public domain. Maps oí the Croatian and 1urkish electricity grid inírastructure
were obtained írom the Global Lnergy Network Institute ,GLNI, |30[. 1here is limited
accuracy with these maps but they are the best source oí data a·ailable.
In order to use the iníormation in these maps they had to be digitised using ArcMap. 1his is a
manual process were the map is íirst laid o·er an existing GIS map using the geo-reíerencing
toolbar oí ArcMap. Once the maps are aligned as accurately as possible the electricity
transmission lines are manually digitised by tracing them using the sketch tool. 1his digitised
data are then sa·ed in shapeíile íormat íor use in the model.
3.2 Design oí ArcGIS model
1ransformation topography & physical characteristics
Distance between dams 5 km -· 1km
Minimum head 150 m
1opology A, assumed minimum new reser·oir size ¯0 000 m°
Minimum distance írom new reser·oir to inhabited sites 500 m
Minimum distance írom new reser·oir to existing transportation inírastructure 200 m
Minimum distance írom new reser·oir to UNLSCO site 5 km
Maximum distance írom new reser·oir to electricity transmission network 50 km
New reser·oir should not be within a Natura 2000 conser·ation area
1able 10: o·er·iew table oí the model parameters íor 1A and 1B
3.2.J 1opology A design
1he model parameters are used to íorm ·arious
scenarios with which to analyse how diííerent
transíormation characteristics will aííect the
1he buííer distance parameter is used to deíine
the search distance írom the existing dam to
potential reser·oir sites. A ·alue oí 5 km is
chosen íor the base scenario, but íurther
scenarios were modelled reducing the distance
oí the buííer in 1-km steps down to a minimum
buííer oí 1 km. 1his will result in the íollowing
parameters íor the model scenarios:
Scenario 5 ~· 5 km
Scenario 4 ~· 4 km
Scenario 3 ~· 3 km
Scenario 2 ~· 2 km
Scenario 1 ~· 1 km
1hen the ArcGIS slope íunction was used to
analyse the topography to ascertain the ílatness
oí the potential transíormation site.
Overview of ArcGIS Slope function from its user manual |32[:
íor eacb cett, ´tope catcvtate. tbe va·ivvv rate of cbavge iv ratve frov tbat cett to it. veigbbovr.. ßa.icatt,,
tbe va·ivvv cbavge iv eteratiov orer tbe ai.tavce betreev tbe cett ava it. eigbt veigbbovr. iaevtifie. tbe .teepe.t
aorvbitt ae.cevt frov tbe cett.
Covceptvatt,, tbe ´tope fvvctiov fit. a ptave to tbe ¸·ratve. of a ² · ² cett veigbbovrbooa arovva tbe proce..ivg
or cevtre cett. 1be .tope ratve of tbi. ptave i. catcvtatea v.ivg tbe arerage va·ivvv tecbviqve ;.ee referevce.).
1be airectiov tbe ptave face. i. tbe a.pect for tbe proce..ivg cett. 1be torer tbe .tope ratve, tbe ftatter tbe terraiv;
Calculate slope of areas inside buffer zones
Select areas with slope between 0 and 5 degree
Select areas with average elevation 150 metres
above the dam under test
Buffer placed around dams under test
Calculate average elevation in the area with
slope between 0 and 5 degree with slope
Each dam has a number of potential reservoir
sites. Filter sites with largest energy storage in
Figure 7: TA algorithm flow diagram
tbe bigber tbe .tope ratve, tbe .teeper tbe terraiv. íf tbere i. a cett tocatiov iv tbe veigbbovrbooa ritb a ^oData
¸·ratve, tbe ¸·ratve of tbe cevtre cett ritt be a..igvea to tbe tocatiov. .t tbe eage of tbe ra.ter, at tea.t tbree cett.
;ovt.iae tbe ra.ter´. e·tevt) ritt covtaiv ^oData a. tbeir ¸·ratve.. 1be.e cett. ritt be a..igvea tbe cevtre cett´. ¸·
ratve. 1be re.vtt i. a ftattevivg of tbe ² · ² ptave fittea to tbe.e eage cett., rbicb v.vatt, teaa. to a reavctiov iv
tbe .tope. 1be ovtpvt .tope ra.ter cav be catcvtatea iv tro t,pe. of vvit., aegree. or percevt ;cattea ´percevt ri.e´).
1be percevt ri.e cav be better vvaer.tooa if ,ov cov.iaer it a. tbe ri.e airiaea b, tbe rvv, vvttiptiea b, 100.
Cov.iaer triavgte ß betor. !bev tbe avgte i. 1: aegree., tbe ri.e i. eqvat to tbe rvv, ava tbe percevt ri.e i. 100
percevt. .. tbe .tope avgte approacbe. rerticat ;·0 aegree.), a. iv triavgte C, tbe percevt ri.e begiv. to approacb
ligure 8: example oí how slope is calculated
1he Slope algorithm:
1be rate of cbavge ;aetta) of tbe .vrface iv tbe bori¸ovtat ;a¸,a·) ava rerticat ;a¸,a,) airectiov. frov tbe cevtre
cett aetervive. tbe .tope. 1be ba.ic atgoritbv v.ea to catcvtate tbe .tope i.:
.tope_raaiav. ~ .1.^ ; \ ; a¸,a·2 - a¸,a,2 ) )
´tope i. covvovt, vea.vrea iv aegree., rbicb v.e. tbe atgoritbv:
.tope_aegree. ~ .1.^ ; \ ; a¸,a·2 - a¸,a,2 ) ) · :¨.2·:¨º
1be .tope atgoritbv cav at.o be ivterpretea a.:
.tope_aegree. ~ .1.^ ;ri.e_rvv) · :¨.2·:¨º
ri.e_rvv ~ \ ; a¸,a·2 - a¸,a,2
1be ratve. of tbe cevtre cett ava it. eigbt veigbbovr. aetervive tbe bori¸ovtat ava
rerticat aetta.. 1be veigbbovr. are iaevtifiea a. tetter. frov ´a´ to ´i´, ritb ´e´
repre.evtivg tbe cett for rbicb tbe a.pect i. beivg catcvtatea.
1be rate of cbavge iv tbe · airectiov for cett ´e´ i. catcvtatea ritb tbe atgoritbv:
a¸,a· ~ ;;c - 2f - i) · ;a - 2a - g) , ;º · ·_cett_.i¸e)
1be rate of cbavge iv tbe , airectiov for cett ´e´ i. catcvtatea ritb tbe fottorivg
a¸,a, ~ ;;g - 2b - i) · ;a - 2b - c)) , ;º · ,_cett_.i¸e)
,Slope algorithm description taken írom ArcGIS desktop help,
Figure 9: determine the
horizontal and vertical
A slope ·alue oí S
degrees was chosen as
an acceptable ílatness
oí the topography oí a
Areas that ha·e a slope
between 0 and 5
degrees are íiltered out
using the reclassiíy`
tool and then
polygon areas. 1hese
polygon areas are now
the potential reser·oir
1he a·erage ele·ation
within each polygon is
now tested to see ií it is
greater than 150 metres
abo·e the Dam ID
ele·ation, thus constituting the head oí the scheme. Ií this is the case, the site passes the
criterion and becomes a potential transíormation site. 1he a·erage ele·ation oí the area íor
potential site is used and compared with the ele·ation oí the existing dam. 1his will account íor
the ·olumes oí material to be exca·ated and íilled to make a sloping site ílat beíore
A minimum area
where the slope
satisíied also needs
to be deíined, and
the íigure oí
¯0 000 m
been chosen, see
surface of a new
Some oí the
resulting areas ha·e
a potential ·olume greater than the existing reser·oir, but it is assumed that the new reser·oir -
the one under search-, cannot be larger than the existing ,lower, reser·oir. In these cases the
potential reser·oir ·olume is made equal to the ·olume oí the existing reser·oir.
3.2.2 Lnergy storage potential
1he equation to calculate the energy a·ailable in a body oí water is deíined as íollows:
µ ρ V E h g =
Figure 10: ArcGIS areas with slope between 0 and 5 degrees (in green)
Figure 11: Head calculation
Potential dam site
Existing dam/ Dam ID
Head (min. 150 m)
í ~ everg, araitabte ;]ovte.)
ç ~ aev.it, ;/g,v
) ;101· /g,v
g ~ acceteratiov of grarit, ;·.º1 v,.
b ~ fattivg beigbt, beaa ;v)
1 ~ rotvve ;v
a~ geveratiov efficievc, of ;·0º)
3.2.3 1opology B design
In a similar way to 1A
one model parameter,
buííer distance, is used
to íorm ·arious
scenarios with which to
analyse how diííerent
characteristics will aííect
the íinal results.
1his parameter is used to
deíine the search
distance between existing
dams. lor the base
scenario two existing
dams must be within 5 km ,scenario 5, oí each other. 1o implement this each dam location has
a 2.5 km buííer around it and where·er buííers intersect, this represents a potential
Scenario 5 ~· 5 km ,2.5 km -2.5 km,
Scenario 4 ~· 4 km ,2.0 km -2.0 km,
Scenario 3 ~· 3 km ,1.5 km -1.5 km,
Scenario 2 ~· 2 km ,1.0 km -1.0 km,
Scenario 1 ~· 1 km ,0.5 km -0.5 km,
3.3 Scenario design 1A & 1B
1he buííer distance írom the existing reser·oir site will be the parameter used to create the 1A
and 1B scenarios. 1here will be a total oí íi·e scenarios. 1he results íor each scenario will
return the number oí suitable transíormation sites oí each scenario. A sensiti·ity analysis will be
prepared to e·aluate the results.
1he buííer scenarios will ·ary the ·alue oí the buííer distance írom 5 to 1 km in 1-km steps.
Scenarios 1A buffer distance (km) 1B buffer distance (km)
Scenario 5 5 2.5 - 2.5
Scenario 4 4 2.0 - 2.0
Scenario 3 3 1.5 - 1.5
Scenario 2 2 1.0 - 1.0
Scenario 1 1 0.5 - 0.5
1able 11: buííer scenarios íor 1A and 1B
A reser·oir has a capacity oí 10 000 000
cubic metres with a 300-metre head.
L ~ ,1 019 9.81 300 10 000 000 0.9,
As 1 \h ~ 3 600 Joules, the stored energy
in reser·oir ~ ¯.5 G\h
Figure 12: Buffer distance from existing dam to potential dam site or existing
In all cases the minimum area oí the reser·oirs ,potential and existing, is set at ¯0 000 m
the minimum head at 150 m.
3.3.2 Constraint analysis
1he constraint analysis will be applied to both uníiltered ,1A only, and the íiltered results. 1he
parameters íor each constraint are detailed below in 1able 12.
1ransformation physical constraints
Minimum distance írom centre oí new reser·oir to inhabited sites 500 m
Minimum distance írom centre oí new reser·oir to existing transportation inírastructure 200 m
Minimum distance írom centre oí new reser·oir to UNLSCO site 5 km
New reser·oir should not be within a Natura 2000 conser·ation area
Maximum distance írom centre oí new reser·oir to electricity transmission network 50 km
1able 12: physical constraints model parameters
1he results section will present the íollowing:
1. Physical potential ,no constraints, íiltered, ,1A & 1B,
2. Inírastructural potential ,constraints, íiltered, ,1A &1B,
In the result charts íor both uníiltered and íiltered results, íor both Croatia and 1urkey the
head reíers to the height diííerence between the existing dam and the potential transíormation
site. Mean head reíers to the mean head oí the total number oí sites íor each scenario result.
3.3.3 liltering potential transíormation sites - example
in the model are
recorded at íirst, and
are reíerred to as the
1here will be more
than one potential
íor each existing
dam as illustrated in
ligure 13. 1he best
will be selected by its
section 3.2.2, and
will be reíerred to as
the íiltered result.
lollowing the íiltering process one transíormation site is selected, which is shown graphically
when all blue polygons but one disappear - the one remaining is encircled in green in ligure 13
ligure 13: example oí uníiltered results íor a transíormation site - light blue polygons
3.4 Result oí the analysis
ligure 14 presents a map ·iew oí the data íor Croatia in ArcMap, a component oí ArcGIS.
1he original data shows that a total oí 23 dams ha·e a reser·oir capacity oí greater than
1 000 000 m
in Croatia, and all those dams were analysed in the GIS model. 1he histogram in
ligure 15 shows that in Croatia more dams are at an ele·ation oí between 101 and 200 metres
than at any other range.
L|evat|on of dams |n Croat|a
0-100 101-200 201-300 301-400 401- 300 301-600 601-700 701-800
Figure 15: elevation histogram of the dams in Croatia, above 1 000 000 m
that are analysed in GIS
Figure 14: ArcGIS Croatian map and layers used.
3.4.J.J 1OPOLOG\ A 1RANSlORMA1ION PO1LN1IAL
1he existing Razo·ac dam will ,1.84 million m
, ser·e as an example to illustrate the process oí
íiltering the sites with higher potential. 1his dam íorms part oí the Velebit PlS which uses the
waters írom the catchment area oí the ri·er Zrmanja, near Zadar. \ater resources are the ri·ers
Obsenica, Rieica and Otuæa with the storage basins Obsenica oí 2.¯ million m
and Stikada oí
13.65 million m
|33[, although only the latter is used as upper reser·oir oí the PlS system |34[.
1he íact that Razo·ac is already a PlS system should allow an extra ·alidation oí the model.
By applying the algorithm shown in ligure ¯ the GIS-based initial analysis, including the 5-km
range, slope analysis, potential reser·oir area abo·e ¯0 000 m
, and head abo·e 150 m, results in
9 sites suitable íor a prospecti·e transíormation to PlS, under 1A theoretical potential
scenario 5, as shown in 1able 13.
ligure 16 shows as blue areas the potential transíormation sites that meet the speciíied
parameters. 1he red area shows the site that has been selected as the most suitable
transíormation site, as it oííers the largest energy storage: it is site 1 in the table below.
Site no. Reser·oir
1 1 840 000 85 808 1 ¯16 161 ¯¯8 3.34
2 1 840 000 82 085 1 641 ¯02 610 2.50
3 1 840 000 156 909 1 840 000 338 1.55
4 1 840 000 396 0¯5 1 840 000 335 1.54
5 1 840 000 651 523 1 840 000 300 1.38
6 1 840 000 368 152 1 840 000 299 1.3¯
¯ 1 840 000 352 ¯98 1 840 000 282 1.30
8 1 840 000 5¯¯ 35¯ 1 840 000 241 1.11
9 1 840 000 119 ¯16 1 840 000 23¯ 1.09
1able 13: Croatian sample transíormation site analysis - Razo·ac dam
Figure 16: Croatian sample transformation site analysis - Razovac dam
Other sites studied
Lxisting reser·oir and dam
1he potential reser·oir selected does
not ha·e the largest area in the column
potential reser·oir area. 1his is because
the methodology limits the maximum
·olume oí the potential reser·oir to
that one oí the existing reser·oir ií the
potential reser·oir has a ·olume greater
than the existing reser·oir. 1he single
íactor that had higher iníluence on site
1 being chosen is that it has a higher
head which results in this site ha·ing
the highest potential energy storage oí
1he Stikada reser·oir has, under the
same assumptions as the analysis
abo·e, a storage capacity oí 18.42
1he theoretical potential results, beíore
any natural-spaces related constraint has been applied, are shown íor the diííerent scenarios in
ligure 1¯. 1his íigure shows, against the leít axis, the number oí dams which ha·e at least one
potential site íor creating a new PlS, and against the right axis the total transíormation
potential ,topology A, oí Croatia. Under scenario 5, the least restricti·e, the total physical
theoretical potential shows 14 sites with a total energy storage capacity oí 6¯.56 G\h.
1he application oí other íilters was subject to certain limitations oí the model and íor this
reason íilters were not exactly applied in the order prescribed by the methodology. 1hus, next
the transport inírastructure ,· 200 m away,, inhabited areas ,· 500 m away,, UNLSCO sites ,·
5 km away, and distance to the electricity grid ,· 50 km away, constraints were applied, and the
exclusion oí en·ironmentally sensiti·e areas ,natural spaces íilter, was applied later on.
\hereas the application oí the íormer three íilters results in a limited reduction oí storage
capacity as shown in ligure 18, the natural spaces íilter has a much higher impact.
Box - Validation of model with the reality.
1he current upper reser·oir oí the Velebit PlS
system, Stikada, has a reser·oir capacity oí 13.65
and ele·ation oí 548 m AMSL. Google
Larth shows that the closest distance between the
Stikada and the Razo·ac reser·oirs is 20 km.
Because oí the 5-km limit set up in the model the
Stikada reser·oir was not captured as a possible
site íor a second reser·oir, thus incurring in the
apparent contradiction that the actual upper
reser·oir was not captured by the model.
1hereíore the Velebit PlS case shows that the
analysis assumptions are on the conser·ati·e side.
\e put the discussion oí this point oíí to the
conclusions and continue the analysis oí country
potential based on the initial assumptions.
ligure 1¯: Croatia 1A theoretical potential: number oí potential sites and total potential storage
Croat|a 1A theoret|ca| potent|a|
Scenarlo 3 Scenarlo 4 Scenarlo 3 Scenarlo 2 Scenarlo 1
No. of dams
number of dams wlLh a poLenLlal slLe 1oLal energy sLorage (CWh)
In eííect, the total number oí scenario 5 physical realisable` potential ,term is used in order to
reílect this particular order oí application oí íilters, sites is 13 with total energy storage oí 59.¯5
G\h. 1his represents the loss oí only one transíormation site due to constraints, with the loss
oí ¯.81 G\h oí energy storage.
1he introduction oí natural spaces as a constraint is run independently to the other constraints.
1he large co·erage oí Natura 2000 areas in Croatia disqualiíies o·er halí ,13 down to 6, oí the
suitable sites when this constraint is applied.
lowe·er, lower-capacity options that were discarded at an earlier stage, might not ha·e the
same natural spaces restrictions as the íiltered sites. 1he model could be run in an iterati·e way
íor those existing dams whose 1A option was discarded at the natural spaces constraints in
order to search whether any other oí the possible sites would pass the natural spaces check.
Croat|a 1A phys|ca| rea||sab|e potent|a|
Scenarl o 3 Scenarl o 4 Scenarl o 3 Scenarl o 2 Scenarl o 1
number of Sl Les 1oLal Lnergy SLorage (CWh)
Croat|a 1A potent|a| after natura| reserves are d|scounted
Scenarl o 3 Scenarl o 4 Scenarl o 3 Scenarl o 2 Scenarl o 1
number of Sl Les 1oLal Lnergy SLorage (CWh)
ligure 19: Croatian 1A physical potential
Figure 18: Croatia TA physical “realisable” potential after applying three filters
3.4.J.2 1OPOLOG\ B ANAL\SIS
1here are no solutions íor 1B in Croatia. Any dams that are within 5km oí each other do not
ha·e an ele·ation diííerence oí 150 m or greater to pro·ide the suííicient head required by the
In order to íind a 1B transíormation site we need to exceed the methodology parameters by
extending the buííer distance to 11 km beíore we íind a solution to 1B. 1he decision taken at
the modelling stage was that this distance between two existing dams would not be a ·iable
transíormation. lowe·er, the experience oí the Velebit and other PlS shows that this
assumption can be conser·ati·e in particular when a high head diííerence is possible.
3.4.J.3 CONCLUSIONS: COUN1R\ PO1LN1IAL
1he country potential is assumed to be without including the natural spaces constraint. As a
result oí not being any 1B potential site, in the case oí Croatia the country potential is the
scenario 5 1A physical realisable` potential íiltered íor other constraints. 1his yields 13 sites
with 60 G\h oí energy storage. 1his íigure can only roughly be compared to the storage oí the
current PlS ,20 G\h, because the latter corresponds to a major mixed-PlS where natural
inílow plays an important role. lor reíerence, peak Croatian demand is approximately 3.2 G\
and annual electricity consumption abo·e 18 000 G\h |63[.
1he diííerent íeatures and uses oí PlS systems, e.g. whether a daily or weekly cycle, whether
pure PlS oí mixed with natural hydropower resources, make it diííicult to deíine whether this
potential íor transíormation is signiíicant. 1he analysis did not deíine the assumptions that
could result in an estimated installed power ,whether pumping or generating, and thus a
comparison cannot be made with the installed PlS capacity oí the country. A íuture
impro·ement oí the model could approach this issue. In the case oí Croatia the current PlS
capacity is 282 M\ generating and 245 M\ pumping, most oí it ,2¯6,240 M\ respecti·ely, at
a single PlS, Velebit. Indeed a comparison with Croatian installed PlS can be misleading
because Velebit is a mixed PlS-con·entional scheme which in 2009 consumed 11¯ G\h
pumping |34[ írom which is estimated to ha·e generated 82 G\h
oí the total 468 G\h
A diííerent approach consists oí comparing the country potential íor storage with the storage
oí the upper reser·oirs in existing PlS in the country. lor Croatia, section 220.127.116.11 un·eiled that
under the same assumptions used to calculate site potential the upper reser·oir oí the Velebit
PlS has a storage capacity oí 18.42 G\h, and contributed to the generation oí 110 and
82 G\h írom 158 and 11¯ G\h pumped in 2008 and 2009 respecti·ely. 1he extrapolation oí
this pattern to the 60 G\h oí country potential yields 3.25 times the current Croatian installed
capacity, i.e. between 266 and 35¯ G\h generated. lowe·er, again this approach can lead to
the wrong íigures because the 18.42 G\h oí storage in the upper reser·oir oí the Velebit PlS
did not only contribute to the PlS system but generates an a·erage 3¯¯ G\h oí pure
hydropower annually |34[.
Gi·en that it cannot be ruled out any new PlS to be a mixed system, possibly the best way to
put the potential into context is to compare the 60 G\h oí potential with the approximately 20
G\h oí currently existing PlS to conclude that under the limitations in this study the country
potential for transformation to PHS in Croatia is at least three times the capacity of
existing PHS plants.
According to LIA |21[ in 2008 the Croatian PlS plants pumped 158 G\h and generated 110 G\h. Gi·en that
Velebit is 98° oí the PlS generation, it can be assumed that the cycle eííiciency oí the Velebit PlS plant is ¯0 °.
ligure 20 presents a map ·iew oí the data íor 1urkey in ArcGIS.
1he authors were unable to acquire accurate transmission network data íor 1urkey. lor this
reason the transmission network constraint ha·e been disabled íor the analysis oí 1urkey.
Figure 20: map of Turkey with the layers included in the analysis
Figure 21: elevation histogram of the dams in Turkey, with a reservoir capacity of 1 000 000 m
A total oí 612 reser·oirs larger than 1 000 000 m
in 1urkey were analysed in the GIS model.
1he histogram, ligure 21, shows that there are a large proportion oí dams at ele·ations
between 0 and 400 metres and between 801 and 1 200.
3.4.2.J 1OPOLOG\ A ANAL\SIS
1he physical realisable` potential
, beíore the natural spaces constraints has been applied,
shows íor scenario 5 a total oí 448 potential sites íor a total energy storage oí 4 3¯2 G\h.
1hese potential sites can also be seen in terms oí a·erage head and energy storage per site, as
shown in ligure 23. In the case oí scenario 5 the 448 sites ha·e an a·erage head oí 2¯5 m and
an a·erage energy storage oí 9.¯6 G\h.
1he physical potential aíter en·ironmental constraints ha·e been applied shows a total number
oí 444 potential sites with total energy storage oí 3 81¯ G\h. 1his represents the loss oí 4
potential transíormation sites due to constraints, with a more signiíicant loss oí 555 G\h oí
As explained abo·e we introduce the term realisable` between quotes to reílect that this potential was not in the
methodology but the result oí constrains during its application which can and should be adapted
ligure 22: 1urkey 1A physical realisable` potential
Lrror! Bookmark not defined.
, number oí potential sites
and total storage
ligure 23: 1urkey 1A physical realisable` potential, a·erage head and a·erage storage
Scenarlo 3 Scenarlo 4 Scenarlo 3 Scenarlo 2 Scenarlo 1
1urkey 1A phys|ca| "rea||sab|e" potent|a|
Average Pead (m) Average Lnergy SLorage (CWh)
Scenarlo 3 Scenarlo 4 Scenarlo 3 Scenarlo 2 Scenarlo 1
1urkey 1A phys|ca| "rea||sab|e" potent|a|
number of SlLes 1oLal Lnergy SLorage (CWh)
energy storage. lowe·er, it has to be noted that the íilters applied at this stage did not include
the distance to the electricity grid as we could not íind a suitable dataset.
1he sample site íor analysis will be írom the 1A physical realisable potential scenario 5 results.
1he sample transíormation site analysis is carried out on Karacaoren II Dam.
1he blue area shows a potential transíormation site that meets the speciíied parameters, the red
area shows the site that has been selected as the most suitable transíormation site, as it has the
largest energy storage and the red point is the existing dam. 1he details oí this site are
highlighted in blue in 1able 14.
Site no. Reser·oir
lead ,m, Stored energy
1 48 000 000 1 106 519 22 130 381 613 34
2 48 000 000 116 ¯61 2 335 216 549 3
1able 14: 1urkish sample transíormation site analysis
1he transíormation site selected has a potential energy storage oí 34 G\h.
18.104.22.168 1OPOLOG\ B ANAL\SIS
1he theoretical potential under the íi·e scenarios is ·ery small íor 1B. Scenario 5 yields 3
theoretical potential sites írom a total oí 612 dams under analysis, their a·erage head is 294 m
and total energy storage oí 3.36 G\h.
1he physical realisable` potential aíter the application oí the natural spaces íilter results in
only 2 physical realisable` potential sites, in both cases the dams at a distance oí between 4
and 5 km írom each other. 1hey ha·e an a·erage head oí 361 m and a total energy storage oí
3.04 G\h as illustrated in ligure 25. 1his represents the loss oí one transíormation site due to
constraints with the loss oí 0.32 G\h oí energy storage.
ligure 24: 1urkish sample transíormation site analysis
1urkey has not adopted Natura 2000 yet so the natural spaces íilter did not include them.
lowe·er, the model will analyse Natura 2000 as a constraint ií the data becomes a·ailable in
the íuture. 1hereíore the physical realisable` potential becomes the physical potential.
22.214.171.124 COUN1R\ PO1LN1IAL
1he country potential is presented based on the 1A and 1B physical potential íiltered results
under scenario 5, and it is presented in 1able 15. As 1B takes priority o·er 1A any dam that is
included in a 1B transíormation will be excluded írom the 1A transíormation sites. In this case
íour sites are excluded írom 1A with the loss oí 1¯ G\h oí potential storage which are
replaced by 2 1B sites adding 3 G\h oí potential storage.
Number oí Sites 1otal Lnergy Storage ,G\h,
1A Scenario 5 440 3 800
1B Scenario 5 2 3
Country Potential 443 3 803
1able 15: 1urkish country potential
4 Barriers to the realisation of this potential
4.J 1opographical barriers
1he model de·eloped in the pre·ious section has identiíied potential transíormation sites based
on head diííerence, distance between existing and potential sites, ílatness oí the surrounding
topography and reser·oir ·olume. It also implemented constraints relating to the construction
oí new reser·oirs in relation to inhabited sites etc. lowe·er the model is unable to analyse
potential sites based on their geology and hydrology.
1he geological íormation oí the potential site could be a barrier to the realisation oí a potential
transíormation site. A detailed geological analysis oí each potential transíormation site would
ligure 25: 1urkish 1B physical potential, number oí potential sites and total storage
Scenarlo 3 Scenarlo 4 Scenarlo 3 Scenarlo 2 Scenarlo 1
1urkey 18 phys|ca| potent|a|
number of SlLes 1oLal Lnergy SLorage (CWh)
need to be períormed to assess its íeasibility íor transíormation to PlS. Porous bedrock is one
potential barrier, íor example, as water losses due to seepage may be larger in porous karst
,limestone, areas |63[. 1he construction oí underground penstocks may also be hindered by
In temperate zones many upland sites suitable íor constructing new upper reser·oirs are on
peat-co·ered slopes. Peat soils may pose a barrier due to their unstable nature when disturbed.
1he disturbance could arise either írom construction oí the reser·oir itselí, or írom the
construction oí access roads íor equipment. Large-scale peat mo·ements ha·e occurred in areas
adjacent to wind íarm construction in Ireland |35[. Large cut-and-íill operations on peat are
problematic as saíe storage oí exca·ated peat is diííicult.
linally, the earthquake risk oí the potential site also needs to be assessed.
A lack oí suríace water at or near to the potential transíormation site could be a potential
barrier to the realisation oí the potential. Ií the potential reser·oir site has an inílow this would
make the site more suitable íor the construction oí a new reser·oir and the creation oí a mixed
PlS plant. An analysis oí the hydrology oí the existing reser·oir is needed to identiíy ií there
are seasonal ·ariations in the supply and le·el oí water. 1he incoming sediment loads to any
existing or new reser·oir also must be assessed, as silting may pose a íurther barrier to
transíormation by reducing the usable reser·oir ·olumes o·er time.
L·aporation is not expected to be a signiíicant problem with PlS |59[.
1he analysis oí transport and grid inírastructure should go one step íurther than was possible
with the model. In eííect, a complete analysis oí the road inírastructure in the region oí the
potential transíormation site would be needed to e·aluate ií it can support high ·olumes oí
and,or hea·y construction machinery. A detailed analysis oí the local electricity grid
inírastructure would also be required. In the case oí existing con·entional hydropower sites,
grid inírastructure will be in place. lowe·er, it may need to be upgraded to pro·ide two-way
power ílows to íacilitate pumping as well as generation. \hether the grid has to be extended or
reiníorced this impro·ement might add ·alue to the stability oí the grid and this could help
o·ercome this barrier.
lor non-hydro dams the local grid inírastructure has to be examined in detail. Issues to
consider include the proximity to the distribution and,or transmission network, proximity to
the nearest substation, the a·ailability oí spare capacity at the substation, the íeasibility oí
upgrading existing substations where they are inadequate. lor sites to be connected to the
distribution network the presence oí other large, ·ariable loads and generators ,such as wind
íarms or energy-intensi·e industries, on the local network may be a potential barrier as well as
an opportunity: íor example, a PlS plant and a wind íarm can be associated so that the latter
pro·ides pumping power with minimum transmission losses.
4.2 Lconomic barriers
4.2.J Llectricity market analysis
1his section períorms a preliminary analysis oí the electricity markets oí Croatia and 1urkey.
1his includes the identiíication oí market signals that would justiíy in·estment in new electricity
storage íacilities within each market ,in·estment in renewable technology,.
4.2.2 Croatia |36[
1he electricity market in Croatia is based on electricity trading through bilateral contracts
concluded between the supplier, the trader and,or the generator. In addition, a contract íor
using the network must be signed with the transmission or distribution system operators -ííP·
Operator pri;evo.vog .v.tara ,lLP-1SO, and ííP·Operator ai.tribvci;./og .v.tara ,lLP-DSO,
respecti·ely- depending on the ·oltage le·el the customer is connected to. During the
realisation oí contracts de·iations in supply and demand occur and thereíore the need íor
system balancing. Real time system balancing is the responsibility oí the 1SO. In order to co·er
power system de·iations in each hour, lLP-1SO oííers balancing energy íor sale or purchase
to market participants.
In 200¯, Croatia adopted a íeed-in tariíí legislation based on the tariíí system íor the
production oí electricity írom renewable energy sources and cogeneration |3¯[ and the
regulation on incenti·e íees íor promoting electricity production írom renewable energy
sources and cogeneration |38[. 1ariíís íor wind power plants reach 90 t,M\h.
Croatia is directly interconnected to Slo·enia, Bosnia and lerzego·ina, Serbia and lungary,
and this creates the potential ,ií the amount oí PlS is increased, to store surplus wind
generation írom these neighbouring countries. It has been estimated that Slo·enia has a
potential to install 600 M\ oí wind generation |39[, Serbia has a potential to install 1 300 M\
oí wind generation |40[. lungary must meet the 13 ° renewable target as part oí the LU 2020
targets and wind generation will be the main contributor to this target |41[.
Pro·iding more storage in this region could add ·alue to increasing renewable penetration not
only in Croatia, but in all interconnected countries.
1he Llectricity Market Regulatory Agency was introduced in 2001 to liberalise the electricity
market in an attempt to enhance competition. Since then, liberalisation oí the market is still
undergoing as the reíorms are not yet completed |42[. 1he 1urkish renewable energy act
considers renewable all non-íossil based energy sources. \ind power, run oí ri·er hydropower
plants and reser·oir hydropower plants with reser·oir areas smaller than 15 km
identiíied as renewable energy sources |43[.
1urkey has ·ast untapped hydropower and wind potential. According to UNLSCO 1urkey
technically íeasible hydropower potential is 213 000 G\h |44[. Aíter its General Directorate oí
State lydraulic \orks currently 1urkey has 1¯2 hydroelectric power plants in operation with
total installed capacity oí 13 ¯00 M\ generating an a·erage oí 48 000 G\h,year, which is
35 ° oí the economically ·iable hydroelectric potential |12[. 1he large quantity oí untapped
hydropower potential in 1urkey could make the de·elopment oí PlS unattracti·e, unless the
penetration oí wind power becomes ·ery large indeed.
In December 2006, the Ministry oí Lnergy published the wind map oí 1urkey, which has
stimulated wind power in·estments írom 1¯2 M\ at the end oí 200¯ to 1 329 M\ at the end
oí 2010. Alone this last year, 528 M\ oí new wind energy capacity was added in 1urkey, on a
year-on-year growth rate oí 66 °. 1urkey hopes to install up to 20 G\ oí wind by 2023,
helping the country to obtain 30 ° oí its electricity generation írom renewable sources |45[.
1urkey has adopted a hybrid system within which renewable power plants built beíore 2012 are
eligible íor the 50 - 55 t,M\h RLli1 íor íirst 10 years oí operation, pro·iding a hedge against
íoreign exchange risk. lurthermore retail licence owners are required to allocate a portion oí
the electricity purchases to renewable power.
4.2.4 Capital cost
ligure 26 capacity ·s. project and speciíic capital cost ,ex transmission line, íor proposed PlS in Lurope,US. \-
axis is íull CapLx cost. 1he size oí bubble is indicati·e oí relati·e cost per M\. Plants in Switzerland and US were
con·erted to Luro using the íollowing exchange rates. ,1 Cll ~ 0.6515t, 1 USD ~ 0.¯0¯15t,. Source: Deane et
Project costs íor PlS are ·ery site speciíic with some quoted costs ·arying írom oí 600-3 000
t,k\ |46[. In the lower end oí the price range a íigure oí 500 LUR,k\ íor power generation-
related costs and 0-16 LUR,M\h íor storage capacity-related costs has been quoted based on
pre-2004 estimates |4¯[. 1he use oí re·ersible pump-turbines in·ol·es that a single penstock
can be constructed, which can reduce construction costs by up to 30 ° with a small increase in
the plant cost |48[.
ligure 26 shows the large ·ariation in capital cost íor two similarly-sized projects, LLAPS and
Limberg II ,500 M\,. LLAPS |49[ is an example oí a 1A transíormation, it uses an existing
lower lake and the project will build an upper reser·oir and penstock and powerhouse. Limberg
II |5[, an example oí a 1B transíormation, uses two existing reser·oirs and builds penstock and
generation equipment. 1hese cases highlight the potential capital cost sa·ings oí de·eloping 1B
transíormation sites o·er 1A sites.
1he breakdown oí
costs into their
etc., is also highly
is diííicult to obtain. A sample oí these data pro·ided by Krajacic et al. |64[ íor a PlS project
in the island oí Krk ,Croatia, is presented in ligure 2¯. 1he total cost íor a 10 M\, 2 pumps
and turbines system is 16.¯8Mt.
Figure 27: Split of cost for a specific PHS project
Variability in capital costs is inherent in PlS projects. 1he construction cost is site and country
speciíic due to the high labour and material intensity oí this type oí construction projects. 1hus
the uncertainty oí capital costs can be a barrier to transíormation.
4.3 Social barriers
Due to the nature oí hydropower projects there are many social barriers that arise. 1hey are
discussed in more detail below.
4.3.J Inhabited sites
1he building oí large dams to create reser·oirs oíten leads to the resettlement oí local residents
as íarmland and dwellings are submerged, íor which large hydropower projects normally
recei·e ad·erse publicity. In some cases when an existing site, like an abandoned quarry, is
utilised as a reser·oir íor a PlS transíormation the public acceptance may be greater íor such a
de·elopment. 1he construction oí new power line inírastructure to transport electricity to and
írom PlS plant can aííect dwellings and settlements in its close proximity.
It is diííicult to gauge social acceptance beíore a site has gone through the planning process.
Social acceptance is a barrier that has a lot oí uncertainty associated with it íor hydro and PlS
de·elopments. In the long term, well-managed, suitably landscaped sites may become
appreciated by their local communities as ·isually attracti·e leisure areas.
\hen damming an existing waterway to create a reser·oir, reduction in water le·els may aííect
na·igation or recreational users ,e.g. canoeists,. 1hese could be potential barriers to the
de·elopment oí PlS but due to the close-system nature oí a PlS and the íact that the
reser·oir that is the basis íor transíormation was already built, it is unlikely to aííect PlS
4.3.3 1rans-boundary issues
Ií regions up and down the ri·er are not in the same country, placing a dam in one region may
aííect ílood risks or water supply issues in another country. 1his could be a potential barrier to
the de·elopment oí PlS due to political sensiti·ity.
4.4 Ln·ironmental and planning barriers
4.4.J Conser·ation issues
Ií the potential transíormation site is within or in close proximity to a Natura 2000 designated
site, a LUROPARC lederation designated site |15[, a UNLSCO designated site |16, 1¯[, a site
oí special scientiíic interest ,SSSI,, a special area oí conser·ation ,SAC,, a special protected area
,SPA,, a national park or aííect the catchment home to protected species, then de·elopment oí
this transíormation may be diííicult. In the case that a de·elopment is allowed to proceed in
one oí the listed protected areas the de·eloper may be asked to replace any habitat that has
been remo·ed or damaged due to the de·elopment, resulting in an additional cost.
Ií the potential transíormation site is on or aííects a ri·er that supports migratory íish or other
animals, spawning grounds or ií the ri·er is used íor angling, de·elopment oí this
transíormation may be diííicult. 1his could be a potential barrier to the de·elopment.
4.4.3 Ln·ironmental beneíits
In some cases, a properly designed PlS system can e·en be used to impro·e water quality
through aeration, pre·enting algal growth and íish kills |4[ |5[.
4.S \ater supply barriers
4.S.J \ater resources
Ií the potential transíormation site is on a watercourse that supplies drinking water or water íor
irrigation, this could be a barrier to de·elopment, as the operational requirements íor multiple
uses will ha·e to be managed together. lowe·er a reser·oir that is constructed íor a PlS
de·elopment could also ha·e a secondary íunction as a storage reser·oir íor irrigation or
drinking water supply, which may make a proposed project more economically attracti·e.
4.S.2 Chemical and physical water quality |51[
An analysis would need to be períormed to assess ií the de·elopment oí the PlS has the
potential to aííect the quality oí water oí the watercourse or ií pollutants could be discharged
during construction. Also the potential that the de·elopment could cause signiíicant algal
growth would need to be assessed. 1hese studies would be carried out as part oí a planning
process and ií the results were negati·e it may become a barrier to the de·elopment.
4.S.3 Biological water quality
An analysis must be completed to assess ií changes in ri·er ílow are likely to cause a signiíicant
change in the in·ertebrate community.
4.6 llood protection
An analysis must be completed to assess ií changes to the ri·er result in reduced ílow capacity
oí the ri·er or ií any alterations are needed that they do not increase the potential to cause
ílooding in the surrounding area. 1he de·elopment oí the site must not aííect any a·ailable
íloodplain area or block potential o·erland ílood ílow that would result in increased e·ents oí
PlS has the potential to curtail ílooding by scheduling pumping during ílood risk periods. 1his
could be used to oííset any other negati·e eííects oí the de·elopment oí the site.
4.7 Conclusions oí the barriers analysis
It is possible to broadly classiíy the barriers identiíied in this section as hard and soít barriers.
lard barriers are those imposed by site conditions or by the absence oí suitable inírastructure.
In general, they may be addressed by technological solutions, but the costs may be prohibiti·e.
Oíten such barriers are highly site-speciíic. In the case oí geological and hydrological barriers,
íurther work would need to be carried out on sites that are identiíied in order to identiíy
possible solutions and the associated costs. lor example, seepage losses can be reduced by
lining reser·oirs with imper·ious materials such as clay or synthetic membranes.
A clear operational strategy íor a proposed transíormation scheme should be identiíied in order
to íully assess site íeasibility. It should incorporate the wider operating en·ironment: ílood
protection, other reser·oir uses, renewable penetration, and proposed renewable de·elopment
to 2020 and beyond.
1he capital costs íor de·eloping PlS can be prohibiti·e, depending on the topography oí the
site. lowe·er, by utilising existing reser·oirs capital cost can be reduced dramatically. New
technological de·elopments may allow some other hard barriers to be o·ercome. Variable
speed, re·ersible pump-turbines will increase the operational ílexibility oí planned PlS
íacilities, and will better equip them to support the integration oí ·ariable renewable generation.
New concepts such as coastal seawater PlS, where the sea acts as a lower reser·oir, may open
up a greater number oí potential sites. lowe·er, there is only one such plant in operation in the
world, a 30 M\ demonstration íacility located in Okinawa, Japan |52[ and it is unclear why no
íurther such de·elopments took place. PlS using an underground ca·ern as the lower reser·oir
has also been proposed and ií successíul, would eliminate many oí the en·ironmental problems
associated with constructing reser·oirs on the suríace |53[.
Soít barriers relate to societal acceptance and the regulatory and market en·ironments íor PlS
and general energy inírastructure de·elopment. 1hese can oíten be addressed by non-technical
measures but may pro·e to be diííicult to resol·e. Societal acceptance oí projects can be
impro·ed through campaigns oí public iníormation, by consultation and communication with
local communities, and by reíerencing successíully completed ,and attracti·e, projects. Ií the
regulatory en·ironment poses barriers ,e.g. through long delays in obtaining planning
permission,, this can be addressed through legal measures, but these will oíten require a
concerted eííort in order to be eííected. 1his usually takes the íorm oí lobbying the responsible
agencies at a national or LU le·el. Similarly, ií barriers are imposed through the existing rules oí
electricity markets, it may be possible to make changes through submissions to national
regulators. lowe·er, this may meet with resistance as regulators preíer not to make írequent
changes to market rules, as the resulting uncertainty may deter íuture in·estments.
Se·eral international research projects íocus or recently íocused on barriers to electricity
storage. 1hose include S1ORILS ,".aare..ivg barrier. to ´1ORage tecbvotogie. for ivcrea.ivg tbe
pevetratiov of ívtervittevt íverg, ´ovrce.`, 200¯-10, and stoRL ,íacititativg everg, .torage to attor bigb
pevetratiov of ivtervittevt reverabte everg,`, starting May 2011, with íunding írom the LU
programme Intelligent Lnergy Lurope |54[, |65[. 1he íormer addressed island systems and
promotion measures whereas the latter aims to identiíy the best practices in Lurope íor
o·ercoming non-technical barriers to the de·elopment oí energy storage íacilities.
S 1opics for future research
1he topics íor íurther research can be broken down into 3 distinct areas: post modelling site
analysis, íuture model de·elopment, and íurther related research.
S.J Post modelling site analysis
Post modelling site analysis would entail íurther more detailed analysis oí the global country
potential, based on detailed knowledge oí speciíic sites, írom either measurements or non-GIS
sources such as operator reports or en·ironmental assessments.
Geological assessment. 1he model de·eloped here assesses sites based on their slope. In
order to íully assess the suitability íor transíormation to PlS íurther analysis on the geological
makeup oí the site and surrounding area would be required. 1his could be used e.g. to more
accurately identiíy costs -depending on the type oí rock a new reser·oir might need an
waterprooíing layer, or is costlier to dig |63[.
Hydrological assessment. A hydrological assessment oí the global country potential would be
needed to e·aluate the inílow into existing reser·oirs or the inílow into the potential reser·oir
sites. It might be that existing GIS water models could pro·ide the input íor this assessment.
Ilood risk assessment. 1he analysis oí barriers details and discusses, in general, a list oí
potential cases which could cause ílood risks. luture research would require an analysis oí each
ílood risk, using the points mentioned in section 4.6 íor the country potential.
Additionally, PlS may be used íor ílood protection, the model could present the capacity oí
water storage in the proposed PlS schemes and relate it to other parameters that identiíy the
role oí PlS in ílood protection. 1hose parameters could be hydraulic, population,
S.2 luture model de·elopment
S.2.J In·estigate the a·ailability oí additional data layers
1he incorporation oí geology and hydrology ,e.g. precipitation and e·aporation, data would
impro·e the íunctionality oí the model de·eloped in this work. A layer oí geological data would
allow some unsuitable sites to be either ílagged or eliminated, íor example sites located on peat
soils or sites with porous bedrock such as limestone ,see the discussion in section 4.1.1,. A layer
containing e·aporation or potential e·apotranspiration would be useíul in order to iníer the
likely e·aporati·e losses írom storage reser·oirs. lowe·er, e·aporation is highly spatially
·ariable, especially in upland regions where orography and local winds may ha·e large eííects,
and large-scale model datasets may not be capable oí íully representing this ·ariability.
Precipitation data would help to quantiíy inílows to reser·oirs.
S.2.2 Other model impro·ements
Alternati·e, higher-resolution, terrain datasets may be a·ailable, on a commercial basis. 1hese
datasets, ií suitable, may impro·e the accuracy oí the site selection process. \e recommend
that the model be tested in a small region with a subset oí any new terrain dataset beíore
proceeding to recalculating transíormation potentials on a country-wide basis. In this way, the
eííects oí the resolution and quality oí the terrain iníormation on the results can be assessed.
A potential extension to the model would be to estimate cut and íill ·olumes íor reser·oir
construction. 1his exercise is likely to be computationally intensi·e íor large areas |55[.
\ith a substantial in·estment oí time
, ICOLD reser·oir sites in countries other than Croatia
and 1urkey could be geo-reíerenced by using the manual cross-reíerencing technique with
Google Larth used to update the 1urkish and Croatian databases. Alternati·ely, other sources
oí data could be íound e·en in GIS íormat which could reduce the eííort íor data preparation.
1he indi·idual resulting schemes could be analysed to distinguish between daily- or weekly-
cycle PlS, a way to do this by combining the capacity oí the proposed PlS in G\h with
reasonable assumptions on pumping capacity, some work was already done e.g. by L\I and
energinautics |56[. 1he country potential could be appropriately split between both types oí
\e assumed that when both 1A and 1B transíormations are possible íor the same site the
choice is 1B owning to lower transíormation costs. 1his disregards the possibility oí 1A
yielding much more potential. 1he model could introduce the e·aluation oí cost ad·antage ·s.
higher energy storage potential. Likewise, there are a number oí areas where a combination oí
parameters can be sensible:
- Criteria íor distance between reser·oirs and head may be considered in combination. lor
example a second reser·oir site 5 km away with 150 m head does not seem like a suitable site
- 1opology B is always preíerred o·er 1opology A, but this can be challenged. lor example,
two existing reser·oirs 5 km away with a head oí 150m may not be a desirable de·elopment íor
a PlS ,1B,. lowe·er ií suitable sites with 600m head exists within 1-2 km oí one oí the two
reser·oirs, it may be ía·oured compared to utilizing the second reser·oir. 1he cost oí
constructing the secondary reser·oir is relati·ely small compared to the o·erall cost oí the
Some oí the reser·oirs can be as long as 20 km ,Peruca, in Croatia,, thereíore the 5-km limit íor
the second reser·oir , currently at the centre oí the dam, could be based not any point in the
lake, thus multiplying the explored area and thus the possibilities oí íinding a suitable site |63[.
1he head parameter could be adapted to be net head by taking into account losses, a·erage
head, or a combination oí both elements.
Maximum head could be an additional technical constraint. Lxisting PlS technology limits the
head between the two reser·oirs to ¯00-800 m. In the example íor Croatia, the selected site is at
the ,íeasible, edge oí existing technology |59[.
S.2.3 Lxtension oí scope: site pre-selection.
It has been highlighted that íor the purposes oí identiíying maximum íeasible PlS country
potential, maximum energy storage seems the proper criterion. lowe·er, íor the pre-selection
oí a PlS site this may not be the most suitable criterion. Other criteria, such as maximum head
or minimum distance between the reser·oirs ,both oí these reduce the CapLx, may also be
used. 1hereíore the methodology could extend the scope íor site pre-selection. 1wo possible
approaches to this could be ,a, to combine two or more criteria with diííerent weights íor
íinding the optimum` site or ,b, to select optimum sites based on more than one criteria
separately ,separate runs which could possibly identiíy diííerent sites,.` |59[
S.2.4 Reaching the íinal user
1his model could pro·ide an increasing ·aluable ser·ice to the íinal users ií their needs were
incorporated in the íorm oí layers. Possible layers include:
- lull data on grid capacity is needed to estimate the cost oí grid connection.
- More detailed calculation oí the size and shape oí the new ,second, dam proposed in 1A.
- Cost data íor the diííerent items, e.g. cost per cubic metre oí concrete íor the dam, per km oí
- Building time data íor the diííerent items, data on permitting delays and other project-
- Llectricity interconnection capacity would help determine the possibilities íor increased PlS
to support the electricity system oí neighbour countries.
Possible users include the spatial planning bodies oí regional or national go·ernments, utilities,
and de·eloper oí pumped hydropower schemes. 1hey should íirst and íoremost be consulted
on which kind oí output írom the model, in terms oí speciíications and íormat, would be
needed to let them reduce costs or impro·e their work.
S.3 lurther related research
lurther related research relates to areas that would beneíit the íuture oí PlS, considering
diííerent types oí plants, detailed costs, operation within electricity systems and markets.
S.3.J Analysis oí the types oí existing PlS
PlS has diííerent coníigurations depending on the topography it is sited in.
• 2 existing reser·oirs with natural inílow ,transíormation studied as 1B,
• 1 existing reser·oir and 1 artiíicial, newly-built one, usually the upper one ,studied as
• 2 artiíicial reser·oirs ,closed loop,
1here are also ·ariations on these 3 coníigurations. Research into existing PlS plant
coníiguration would allow íor the categorisation oí each PlS plant under the abo·e headings.
1his knowledge would be useíul íor planned PlS, especially in the area oí capital costs. No·el
technologies such as coastal seawater PlS could also be included in this categorisation.
S.3.2 Analysis oí capital cost
As highlighted in section 4.2.4, the uncertainty oí capital costs íor PlS is a major barrier to the
íuture de·elopment. Research into breaking down the capital costs into its constituent parts
,penstock, generation equipment, reser·oir construction, or e·en a more detailed split, would
pro·ide more certainty íor de·elopers. lollowing on írom the categorisation oí existing PlS, a
capital cost íor each coníiguration could make capital cost estimation more accurate in the
S.3.3 L·aluate the role oí PlS within the electricity markets in Lurope
A barrier íor de·elopers oí PlS is the uncertainty oí income streams írom energy markets.
Lnergy payments írom trading in the wholesale market are normally the main source oí income
íor PlS operators. lowe·er, some market structures also pay capacity payments íor the
a·ailability oí generation and,or pay íor the a·ailability oí generators íor ancillary ser·ices
,reser·e, black start etc.,. Research into how existing PlS operate within existing Luropean
markets would pro·ide some le·el oí income certainty, and an indication oí whether projects
can be íinancially ·iable. lurthermore, stakeholders ha·e highlighted the need to identiíy as
part oí the e·aluation oí the potential the economic aspects including the potential income oí a
PlS plant írom energy
, capacity, and ancillary ser·ices, or whether there are any other
íinancial incenti·es. írev tbovgb var/et avat,.i. i. vot ,ovr focv., a vore aetaitea orerrier of var/et
i..ve. rovta aaa ratve to ,ovr report` |59[.
S.3.4 Connection with the National Renewable Lnergy Action Plans ,NRLAP,.
Under the renewable energy Directi·e ,2009,28,LC, LU Member States ha·e to prepare plans
to meet their respecti·e 2020 goals oí renewable energy contribution. Gi·en that large-scale
energy storage is nowadays only possible with reser·oir-based hydropower or PlS, there is a
clear connection between the implementation path shown in NRLAPs and the need íor energy
storage. 1his connection is shown in the NRLAP which include projections oí PlS as well as
other hydropower installation.
lurther research could look at how PlS ,and the transíormation to PlS, could enable grids to
accommodate a higher ·ariable-RL component.
S.4 Potential co-operation with GRanD
1his project has beneíitted írom the data pro·ided by the GRanD project. Ií íurther work is to
be carried out in this area, then íormal co-operation with GRanD may be mutually beneíicial.
As  suggested “what is important is the difference between off-peak and peak electricity prices,
multiplied by efficiency”
6 Methodological remarks and conclusions
In this analysis oí potential íor transíormation the authors were obliged to take decisions based
on empirical analysis as well as on their own experience, with the limitations imposed by the
model and with a·ailability oí data being a key iníluencing íactor. Because oí the latter those
decisions at times had to be arbitrary and not necessarily matched the reality. A good example
to illustrate this point is the 1A case íor Croatia which was contrasted with the reality. 1his
one, the Razo·ac dam, is part oí the Velebit PlS system where the Stikada reser·oir is the
upper one. lere the reality challenged two key assumptions oí the design oí the model, namely
that the size oí the potential new upper reser·oir should not be bigger than the lower ,existing,
reser·oir, and that the economic distance between the two reser·oirs should be lower than
5 km. Indeed the ·alidity oí the latter assumption was íurther challenged by other example, the
PlS project "Atdorí" ,1.4 G\, 13 G\h, in Baden-\ürttemberg ,South Germany, with a
distance between the two ,new, reser·oirs oí 8.5 km |62[.
L·ery scenario íor high penetration oí renewable energy in electricity systems highlights the
need íor electricity storage ,|56[, |5¯[, and puts storage as a key íactor íor reducing the cost oí
energy ií the renewable electricity is oí a ·ariable nature. 1his modelling exercise is, to the
knowledge oí the authors, the íirst approach to identiíying and quantiíying the potential íor
transíormation to pumped hydropower storage in Luropean countries based on one or two
existing dams. lowe·er, this exercise belongs to the íield oí research and, as the reality check
has hinted, its results might be some stages away írom the accuracy and deíinition required íor
an actual project íeasibility study. 1his is important because ,we belie·e, the ultimate goal oí an
exercise to quantiíy the potential íor increasing PlS should be dual: to íeed the decision-
making process with sound science and to reduce the costs oí transíormation íor all actors
in·ol·ed: go·ernmental spatial planning agencies, engineering companies and PlS de·elopers.
Re·iewers ha·e highlighted that the parameters used to restrict the search íor suitable sites were
too restricti·e regarding reality. 1here are inhabited areas less than 200 m írom a reser·oir,
penstock can be buried so they can cross transport inírastructure and thus the 100-m distance
to the latter might be a unnecessary restriction |63[. 1he restriction oí 5 km to UNLSCO sites
may be excessi·e when, e.g. this is an isolated chapel in the middle oí the mountains
1hroughout this report the primary íocus was on the storage capacity, which is necessary since
we are in·estigating the a·ailability oí potential reser·oirs. lowe·er, power capacity is
somewhat independent oí the storage capacity i.e. to increase the power capacity more
penstocks can be constructed at existing PlS sites. Some research has indicated that increasing
the power capacity could enable higher wind penetrations without any increase in the storage
capacity` |58[ |65[
lydrology is a critical criterion especially íor areas,countries in dry climates, such as those in
Southern Lurope. 1he assumption that the existing reser·oir ·olume can be potentially used
íor PlS is not necessarily ·alid in dry areas. lor example in Cyprus dams are rarely íull or e·en
near íull, they are o·ersized in order to maximize the water collection in rainy years and use it
íor storage in dry years. In these cases the addition oí a second reser·oir would increase the
·olume oí water stored at the peak rainy season but during those period the reser·oir could not
be used as PlS but as permanent water storage. lurthermore, a PlS plant cannot use all the
existing ·olume oí the reser·oir, otherwise debris and silt would be drawn up the pump-
turbine. Large le·el diííerences are not technically ía·ourable íor the mechanical equipment,
depending on the shapes,areas oí the reser·oirs large head diííerences may de·elop írom the
start oí the pumping cycle to the end. As a íirst step, the model could limit the ·olume oí the
reser·oir by a certain percentage ,say, 80°, to account íor all these íactors |59[.
1he next steps could include the opening oí a dialogue with these stakeholders that would
result in a more ílexible model able to pro·ide more accurate results that are closer to reality
and thereíore start to be useíul íor at least some oí those actors. L·entually, the process started
with this work could ,some would say should`, be expanded to the whole oí Lurope.
1he country potential íor transíormation to PlS in Croatia is oí 60 G\h, which compares
with the current installed PlS storage oí 20 G\h. lowe·er, the latter is mostly ,98°, in one
single, mix-PlS installation which generates 80° oí its electricity írom pure hydropower
resources. Ií a ·olume íactor
is applied as proposed by |56[, 60 G\h would correspond to
turbine capacity oí 2.3 G\
1he realisable potential íor transíormation in 1urkey shown by this analysis is 3 800 G\h. 1his
íigure corresponds to 146 G\ oí turbine capacity at a ·olume íactor oí 26, and can be
compared with the estimated 35 G\ oí peak demand and 230 000 G\h oí generation in 2010.
Uníortunately there is not an option to compare this transíormation potential with the existing
PlS capacity because currently there are no PlS plants in 1urkey.
\e need some insight on how the prospecti·e new PlS could be used to help stabilise the grid
and increase the uptake oí renewable energy. In eííect, some oí the potential PlS could be
used íor intra-day balancing, i.e. pumping at night when there is excess electricity írom baseload
,coal or nuclear, plant, and generating during the day. Some other PlS with higher storage
capacity could be used íor weekly or monthly storage ií economically íeasible. A PlS
transíormation based on the Karacaoren II Dam in 1urkey, with 34 G\h oí storage capacity,
could be used to store electricity írom excess wind rather than curtailing wind production.
In eííect, wind energy cycles may last hours but most írequently last 3-4 days depending on the
local climatology. In electricity systems with high wind penetration and low export capacity,
islands or electrical peninsulas` such as Ireland or Inner Mongolia in China, wind would need
to be curtailed whereas a PlS plant with large storage capacity can absorb and then release
these wind energy during peak demand and thus ha·ing the additional en·ironmental eííect oí
a·oiding the use oí the peaking plant íuelled by íossil íuels whether natural-gas or coal ,e.g. in
China,, and beneíiting írom a subsequent reduction in greenhouse gas reductions.
1. lernando Peran Montero and Juan J. Pérez, \ind-hydro integration: pumped storage to
support wind.` |Online[. A·ailable: http:,,www.hydroworld.com,index,display,article-
2. B. Dursun and B. Alboyaci, 1he contribution oí wind-hydro pumped storage systems in
meeting 1urkey's electric energy demand,` Renewable and Sustainable Lnergy Re·iews,
·ol. 14, no. ¯, pp. 19¯9-1988, Sep. 2010.
3. LLA ,Luropean Ln·ironmental Agency, L1C,ACC 1echnical Paper 2010,1¯,
December 2010: Small-scale hydropower: a methodology to estimate Lurope`s
en·ironmentally compatible potential. A·ailable at
According to |56[ 1he ·olume íactor represents the ratio oí storage size and the turbine capacity ,e.g. 1 G\
pump storage turbine has a 26 G\h storage basin,`
4. C. \ang and R. B. Jackson, Opportunities and barriers to pumped-hydro energy storage
in the United States,` Renewable and Sustainable Lnergy Re·iews, ·ol. 15, no. 1, pp. 839-
844, Jan. 2011. A·ailable at http:,,www.duke.edu,~cy42,PlLS.pdí
5. Verbund is an Austrian utility responsible íor the Rei|eck II, Limberg II and other PlS
projects. Online at http:,,www.·erbund.com,cc,en,about-us,our-power-plants,power-
plant-types,pumped-storage-plants, accessed 8,11,11.
6. Lurostat, Luropean oííice oí statistics, a·ailable at
¯. lLP Proiz·odnja d.o.o. web page íor the Velebit PlS plant
8. lLP Proiz·odnja d.o.o. web page íor the Vinodol hydropower system
9. Llectricity renewable energy target potential írom Screening Report Croatia, Luropean
hr_internet_en.pdí. See also the Croatian targets írom http:,,narodne-
10. Summaries oí LU legislation
11. International Lnergy Agency`s Reverabte. ívforvatiov 2011, page 455
12. DSI: General Directorate oí State lydraulic \orks oí 1urkey, consulted on 10,10,11 at
13. M. Ozturk, N. C. Bezir, and N. Ozek, lydropower-water and renewable energy in
1urkey: Sources and policy,` Renewable and Sustainable Lnergy Re·iews, ·ol. 13, no. 3,
pp. 605-615, Apr. 2009.
14. Global \ind Lnergy Council, http:,,www.gwec.net,index.php·id~133 & \ind Summit
15. LUROPARC lederation: http:,,www.europarc.org,home,
16. UNLSCO Biosphere Reser·e sites: http:,,www.unesco.org,new,en,natural-
1¯. UNLSCO \orld leritage list: http:,,whc.unesco.org,en,list
18. LU member countries, http:,,europa.eu,abc,european_countries,index_en.htm
19. Luropean lree 1rade Area, http:,,www.eíta.int,
20. LU candidate and proposed candidate,\estern Balkan countries,
21. Lnergy Iníormation Administration, DoL, US, a·ailable at
gions&syid~2005&eyid~2009&unit~BK\l, consulted 2¯,05,11.
22. International Commission on Large Dams ,ICOLD, http:,,www.icold-cigb.net
23. 1he Global Reser·oir and Dam ,GRanD, database,
24. Digital \ater Atlas at G\SP, http:,,www.gwsp.org,85.html
25. 1he CGIAR Consortium íor Spatial Iníormation,
26. Lehner, B., Reidy Liermann, C., Re·enga, C., Vorosmarty, C., lekete, B., Crouzet, P.,
Doll, P., Lndejan, M., lrenken, K., Magome, J., Nilsson, C., Robertson, J., Rodel, R.,
Sindorí, N., \isser, D. ,in press,: íigb re.otvtiov vappivg of tbe rorta`. re.erroir. ava aav. for
.v.taivabte rirer ftor vavagevevt. lrontiers in Lcology and the Ln·ironment.
2¯. DIVA-GIS is de·eloped by Robert lijmans http:,,www.di·a-gis.org,gdata
29. Corine Land Co·er raster is a·ailable at http:,,www.eea.europa.eu,data-and-
30. Global Lnergy Network Institute http:,,www.geni.org,
31. Natura 2000 deíinition taken írom http:,,ec.europa.eu,en·ironment,nature,Natura
32. Burrough, P. A. and McDonell, R.A., 1998. Principles oí Geographical Iníormation
Systems ,Oxíord Uni·ersity Press, New \ork,, p. 190.
33. 2008 annual report oí the Croatian Academy oí Lngineering, a·ailable at
http:,,www.hatz.hr,annuallA1Z_2008.pdí, accessed 3,06,11.
34. Verebit pump storage power plant site by lLP Proiz·odnja d.o.o.,
35. Dykes, A. P. & \arburton, J. Mass mo·ements in peat: A íormal classiíication scheme.
Geomorphology, 200¯, 86, ¯3-93
36. Description oí the Croatian electricity market taken íorm http:,,www.hrote.hr
44. Capabititie. for í,aroporer geveratiov, UNLSCO, a·ailable at
45. Global wind energy council ,G\LC, data on 1urkey, a·ailable at
http:,,www.gwec.net,index.php·id~133, accessed 8,11,11.
46. Gatzen C. 1he economics oí power storage. Munich: Oldenburg Industrie·er, 2008.
4¯. Gonzalez, A., Ó'Gallachoir, B., & McKeogh, L., 2004. ´tva, of etectricat .torage tecbvotogie.
ava tbeir potevtiat to aaare.. riva everg, ivtervittevc, iv íretava. Cork: UCC Sustainable Lnergy
48. Bogenrieder, \. í,aroetectric porer. In Renewable Lnergy Vol 3C, Springer Berlin
leidelberg, 2006, pp. 165-196
49. LVM\D LLAPS website, online at
50. J. Deane, B. Ó Gallachóir, and L. McKeogh, 1ecbvo·ecovovic rerier of e·i.tivg ava ver pvvpea
b,aro everg, .torage ptavt, Renewable and Sustainable Lnergy Re·iews, ·ol. 14, no. 4, pp.
1293-1302, May 2010.
51. Description oí each constraint is írom the UK Ln·ironmental Agency, Good practice
guidelines to the en·ironmental agency handbook, 1be evrirovvevtat a..e..vevt of propo.ea
tor beaa b,aroporer aeretopvevt., http:,,publications.en·ironment-
52. Pritchard, S. ]apave.e pvvpea .torage evbrace. tbe oceav rare.. International \ater Power and
Dam Construction, 2000, 52,8,, pp. 30-34
53. Martin, G. D. & Barnes, l. S. .qvifer |vaergrovva Pvvpea í,aroetectric íverg, ´torage. In
Proceedings oí the Llectrical Lnergy Storage Applications & 1echnologies Coníerence,
San lrancisco, USA, April 200¯.
55. Deretopvevt of a covpvter prograv to tocate potevtiat .ite. for pvvpea b,aroetectric everg, .torage, D.
Connolly, S. MacLaughlin, M. Leahy, Lnergy, 35 ,1,, January 2010, Pages 3¯5-381
56. Roaavap 20:0 - a cto.er too/. Co.t·efficievt Rí´·í pevetratiov ava tbe rote of gria e·tev.iov..
Institute oí Lnergy Lconomics at the Uni·ersity oí Cologne and energynautics, íinal
report October 2011.
5¯. Patbra,. torara. a 100 º reverabte etectricit, .,.tev. SRU - German Ad·isory Council on the
Ln·ironment, special report, October 2011. Luisenstrasse 46, 1011¯ Berlin
58. Dr. Da·id Connolly, co-author oí |55[, re·ision oí the draít oí this paper.
59. Constantinos Varna·as, PhD, Assistant Manager, Generation, to the Llectricity Authority
oí Cyprus, personal communication.
60. Lmmanuel Branche, Senior Lngineer Lconomist, Generation&Lngineering, Sa·oie
1echnolac, Llectricité de lrance
61. 1vr/e, !ater Report 200·. General Directorate oí State lydraulic \orks, 2009. Consulted
on 20,12,11 at http:,,www2.dsi.go·.tr,english,pdí_íiles,1urkey\aterReport.pdí
62. Dr. Klaus Schneider, 1echnische Grundsatzíragen, Schluchseewerk, personal
63. Dr. Ne·en Duic, Uni·ersity oí Zagreb, Croatia, re·ision oí the draít oí this paper.
64. Krajacic G, Loncar D, Duic N, Lacal Arantegui R, Loisel R, Raguzin I. .vat,.i. of fivavciat
vecbavi.v. iv .vpport of ver pvvpea b,aroporer .torage pro;ect. iv Croatia. lorthcoming
65. Dr. Goran Krajacic, Uni·ersity oí Zagreb, Croatia, re·ision oí the draít oí this paper.
1he íollowing documents can contribute to enhance the issues analysed in this scientiíic and
• J. Deane, B. Ó Gallachóir, and L. McKeogh, 1echno-economic re·iew oí existing and
new pumped hydro energy storage plant,` Renewable and Sustainable Lnergy Re·iews, ·ol.
14, no. 4, pp. 1293-1302, May 2010.
• Bogenrieder \. Pumped storage power plant-renewable energy`. Volume 3 'Lnergy
technologies' oí Landolt-Bornstein Group VIII 'ad·anced materials and technologies,
2006. 1he cycle oí pumping and generating can be repeated on a daily, weekly or e·en
seasonal basis. In the daily cycle the reser·oirs can be íilled and emptied within a 24 hour
period while in the weekly cycle the upper reser·oir is partially drawn down and partially
reíilled during weekdays and completely reíilled during weekends when the system load is
normally low. Seasonal pumping is applied in hydropower systems with a large annual
·ariation in inílow oí water to the reser·oirs.`
• Limberg II PlS, Verbund, http:,,reports.·erbund.at,2006,csr,selectedprojects,pumped-
storagepowerplantlimbergii.html Abstract: 1he new pumped-storage power plant,
Limberg II, which is being constructed by Verbund-Austrian lydro Power AG ,AlP, to
supplement the existing Kaprun power plant group, is located at the rear oí the Kaprun
Valley. 1he balancing and backup power plant, Limberg II, will more than double the
output oí the Kaprun power plant group írom 353 M\ to 833 M\ by making optimal use
oí the diííerence in height between the existing Alpine storage lakes, Mooserboden ,2 036
m abo·e sea le·el, and \asseríallboden ,1,6¯2 m abo·e sea le·el,.` 1¯ years construction
to 1995, 166 Mm3 between these íirst two reser·oirs. Very interesting en·ironmental
• LPRI ,Llectric Power Research Institute,, 1990. Pumped Storage Planning and
L·aluation Guide`. Prepared by larza Lngineering Company, Chicago. A
comprehensi·e and stand-alone guide is oííered íor the preliminary e·aluation oí pumped-
storage sites, to help ,a, e·aluate períormance and beneíits oí pumped storage in a utility
system, including dynamic beneíits, ,b, identiíy the physical characteristics oí a site suitable
íor pumped-storage de·elopment, ,c, establish the site's energy storage potential and
installed capacity, ,d, estimate capital cost, and annual operation and maintenance expense,
and ,e, conduct an economic analysis. A PC-based computer program has been written
and is included in the Guide Book to assist in beneíit analysis. 1he concept oí dynamic
beneíits is explained, and how pumped storage contributes to them is described. Six oí the
major power system generation planning models are e·aluated to help utility planners
select the model best suited íor their application. A methodology is pro·ided to help
planners screen and select sites encompassing cost, beneíit, en·ironmental, and regulatory
íactors. Step-by-step procedures are described, one simpliíied and the other more detailed,
to íacilitate the use oí the guide. Background descriptions are pro·ided to assist those
uníamiliar with pumped-storage practice. A series oí cost cur·es are pro·ided to permit the
de·elopment oí a preliminary capital cost estimate íor a site, based on a íew key
parameters which deíine the physical characteristics oí the site.`
• USA Army Corps oí Lngineers. Lngineering and design-hydropower ,e·aluating pumped-
storage hydropower,. Publication number: LM 1110-2-1¯01, December 1985. A·ailable at
Pumped storage operation can be best understood by examining an oíí-stream pumped-
storage project which operates on a daily,weekly cycle ,the most common type oí pumped
storage de·elopment in the United States,. 1he early sections oí this chapter discuss the
analysis oí this type oí project. Later sections are de·oted to pump back, seasonal pumped
storage, and other aspects oí pumped storage de·elopment.`
• Allen A.L, 19¯¯: Potential íor con·entional and underground pumped-storage, ILLL
1ransactions on Power Apparatus and Systems 96,3,, May 19¯¯. 1he purpose oí this
paper is to present a ·ery brieí re·iew oí the current state oí pumped-storage, with a
cataloguing oí its beneíits and problems, and a brieí summary oí íuture potentialities. 1he
potential íor pumped storage is directly related to the public's ·iewpoint.`
\e would like to acknowledge the excellent íeedback pro·ided by the íollowing persons,
íeedback that was crucial to impro·e this document: Dr. Constantinos Varna·as, Assistant
Manager ,Generation, to the Llectricity Authority oí Cyprus, Dr. Da·id Connolly, Assistant
Proíessor, Aalborg Uni·ersity oí Denmark, Alejandro Perea Sanchez, Gerente de Planiíicación
y Oíertas, Iberdrola Generación, Lmmanuel Branche, Senior Lngineer Lconomist,
Generation&Lngineering, Sa·oie 1echnolac, Llectricité de lrance, Dr. Klaus Schneider,
1echnische Grundsatzíragen, Schluchseewerk, Dr. Ne·en Duic and Dr. Goran Krajacic,
Uni·ersity oí Zagreb.
EUR 25239 EN – Joint Research Centre – Institute for Energy and Transport
Title: Pumped-hydro energy storage: potential for transformation from single dams
Author(s): Roberto Lacal Arántegui, Institute for Energy and Transport, Joint Research Centre of
the European Commission, Petten, the Netherlands.
Niall Fitzgerald and Paul Leahy, Sustainable Energy Research Group, University
College Cork, Cork, Ireland
Luxembourg: Publications Office of the European Union
2012 – 55 pp. – 21 x 29.7 cm
EUR – Scientific and Technical Research series – ISSN 1831-9424
Cover picture: Dam of Cortes II, part of the pumped-hydropower scheme Cortes – La Muela, in
Spain. Courtesy of Iberdrola
Electricity storage is one of the main ways to enable a higher share of variable renewable electricity
such as wind and solar, the other being improved interconnections, flexible conventional generation
plant, and demand-side management.
Pumped hydropower storage (PHS) is currently the only electricity storage technology able to offer
large-scale storage as that needed for accommodating renewable electricity under the 2020 EU
Compared with the high environmental and social impact of most new hydropower plant in Europe,
the transformation of an existing reservoir into a PHS system offers the prospects of a much smaller
environmental and social impact.
The authors developed a geographical information systems (GIS) -based methodology and model to
identify the potential for transforming single reservoirs into PHS systems, and to assess the
additional energy storage which these new PHS could contribute to the electricity systems. The
methodology was applied as case studies to Croatia and Turkey.
GIS-based tools have the potential for effective and efficient identification of both national/EU
potentials (of policy and scientific-interest) and individual site candidates for transformation (pre-
feasibility, project-level). Once the model is set up, improvements to such tools, e.g. allowing better
sensitivity analysis, can be effectively applied to the whole of the EU with minimum effort.
This paper first summarises the methodology and tool used and then exposes the results of its
application to two countries as case studies. These results limit the assessment to potential sites
within 5 km of one existing reservoir (TA) or of one another (TB), and a minimum 150 m of head. In
the case of Croatia, it was found that at least a potential of 60 GWh is possible for which can be
compared with the existing 20 GWh of storage capacity at its PHS plants. In the case of Turkey a
potential of 3 800 GWh was assessed mostly under TA, with 2 potential TB sites providing three
additional GWh of storage potential.
How to obtain EU publications
Our priced publications are available from EU Bookshop (http://bookshop.europa.eu), where
you can place an order with the sales agent of your choice.
The Publications Office has a worldwide network of sales agents. You can obtain their contact
details bysending a fax to (352) 29 29-42758.
The mission of the JRC is to provide customer-driven scientific and technical support
for the conception, development, implementation and monitoring of EU policies. As a
service of the European Commission, the JRC functions as a reference centre of
science and technology for the Union. Close to the policy-making process, it serves
the common interest of the Member States, while being independent of special
interests, whether private or national.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.