Overview of pote ntial locations for new Pumped Storage ......Number of deliverable: D4.2 Final...
Transcript of Overview of pote ntial locations for new Pumped Storage ......Number of deliverable: D4.2 Final...
1
S E V E N T H F R A M E W O R K P R O G R A M M E
T H E M E 5 - E N E R G Y
Project acronym : eSTORAGE
Project full title : Solution for cost-effective integration of renewable intermittent generation by
demonstrating the feasibility of flexible large-scale energy storage with
innovative market and grid control approaches
Grant agreement no.: 295367
Collaborative project / Demonstration Project
Number of deliverable: D4.2 Final Report
Dissemination level (PU, PP, RE, CO, RUE, CUE, SUE) : PU
Date of preparation of the deliverable (latest version): 25-11-2015
Date of approval of the deliverable by the Commission: n.a.
Overview of potential locations for new
Pumped Storage Plants
in EU 15, Switzerland and Norway
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 2
R E V I S I O N C H A R T A N D H I S T O R Y L O G
V E R S I O N S
Version number When Organisation name Comments
v0.1 06/12/2013 KEM Interim Report
v1.0 21/10/2015 KEM Draft report for review
v1.1 25/11/2015 KEM Final report
D E L I V E R A B L E Q U A L I T Y R E V I E W
Quality check Status Date Comments
Marko Aunedi (ICL) 16/11/2015
Nathalie Lefebvre (EDF) 12/11/2015
Roberto Lacal Arantegui (JRC) 25/11/2015
Quality Manager 30/11/2015
Project coordinator 30/11/2015
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 3
T A B L E O F C O N T E N T S
1 EXECUTIVE SUMMARY ............................................................................................................................. 9
2 INTRODUCTION ................................................................................................................................... 12
3 GENERAL APPROACH ............................................................................................................................ 14
3.1. OBJECTIVE OF TASK 4.1.2 .......................................................................................................................... 14
3.2. APPROACH .............................................................................................................................................. 15
3.2.1. Key activities ............................................................................................................................ 16
3.2.1.1. Rough criteria ....................................................................................................................... 17
3.2.1.2. Model development ............................................................................................................. 17
3.2.1.3. Dataset collection ................................................................................................................ 17
3.2.1.4. Data quality assessment ...................................................................................................... 17
3.2.1.5. Data preparation .................................................................................................................. 18
3.2.1.6. Review and refinement ........................................................................................................ 18
3.2.1.7. Early results for one country ................................................................................................ 18
3.2.1.8. Final results: theoretical and realisable potential ................................................................ 18
3.2.2. External input and steering ..................................................................................................... 18
3.2.2.1. Steering committee .............................................................................................................. 18
3.2.2.2. Literature ............................................................................................................................. 19
3.2.2.3. Database existing storage (eStorage task 4.1.1) .................................................................. 19
4 CRITERIA ........................................................................................................................................... 20
4.1. INTRODUCTION ......................................................................................................................................... 20
4.2. WATER BODY TYPES .................................................................................................................................. 21
4.3. SELECTION CRITERIA .................................................................................................................................. 21
4.3.1. Energy storage capacity ........................................................................................................... 22
4.3.2. Distance between reservoirs ................................................................................................... 22
4.3.3. Head ......................................................................................................................................... 22
4.3.4. Average slope between water bodies ..................................................................................... 22
4.4. AUTOMATIC PRE-RANKING ......................................................................................................................... 23
4.5. AUTOMATIC ADDITION OF RELEVANT CHARACTERISTICS ................................................................................... 23
4.5.1. Distance to grid ........................................................................................................................ 23
4.5.2. Overlap with restricted zones ................................................................................................. 23
4.6. OVERVIEW ............................................................................................................................................... 24
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 4
5 DATASET SELECTION ............................................................................................................................. 25
5.1. INTRODUCTION ......................................................................................................................................... 25
5.2. COUNTRY BOUNDARIES .............................................................................................................................. 25
5.3. WATER BODIES ......................................................................................................................................... 25
5.3.1. Level of detail .......................................................................................................................... 26
5.3.2. Accuracy................................................................................................................................... 26
5.3.3. Selected dataset ...................................................................................................................... 26
5.4. RESERVOIRS AND DAMS ............................................................................................................................. 26
5.5. ELEVATION .............................................................................................................................................. 27
5.6. GRID DATA .............................................................................................................................................. 28
5.7. SENSITIVE AREAS ....................................................................................................................................... 28
6 METHODOLOGY .................................................................................................................................. 30
6.1. INTRODUCTION ......................................................................................................................................... 30
6.2. GENERAL APPROACH ................................................................................................................................. 30
6.2.1. Coordinate System .................................................................................................................. 30
6.3. DATA PREPARATION .................................................................................................................................. 30
6.4. PRIMARY SELECTION .................................................................................................................................. 31
6.5. RANKING PARAMETERS .............................................................................................................................. 34
6.6. RELEVANT CHARACTERISTICS OF POTENTIAL PAIRS .......................................................................................... 34
6.7. RESULTS PROCESSING AND VISUALISATION .................................................................................................... 34
6.8. VALIDATION ............................................................................................................................................. 35
7 EXPERT SELECTION PROCESS ................................................................................................................... 36
7.1. INTRODUCTION ......................................................................................................................................... 36
7.2. GOAL ...................................................................................................................................................... 36
7.3. DATA FOR EXPERT SELECTION PROCESS ......................................................................................................... 36
7.3.1. Database of pairs ..................................................................................................................... 36
7.3.2. Online interactive map ............................................................................................................ 36
7.4. SELECTION OF EXPERTS .............................................................................................................................. 38
7.5. NON-INCLUDED PARAMETERS ..................................................................................................................... 38
8 RESULTS: THEORETICAL AND REALISABLE POTENTIAL ..................................................................................... 39
8.1. INTRODUCTION ......................................................................................................................................... 39
8.2. RESULTS .................................................................................................................................................. 39
8.2.1. Aggregated results for study area ........................................................................................... 39
8.2.2. Pair locations ........................................................................................................................... 43
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 5
8.2.3. Pair properties ......................................................................................................................... 45
8.2.4. Country-specific observations ................................................................................................. 47
8.2.5. Exclusion causes and observations .......................................................................................... 49
8.3. SENSITIVITY ANALYSIS ................................................................................................................................ 52
8.4. COMPARISON WITH JRC STUDY ................................................................................................................... 53
8.5. DISCUSSION ............................................................................................................................................. 55
9 CONCLUSIONS..................................................................................................................................... 58
10 OUTLOOK .......................................................................................................................................... 60
11 APPENDICES ....................................................................................................................................... 61
11.1. APPENDIX 1: IUCN CATEGORY SYSTEM FOR SENSITIVE SITES .................................................................... 61
11.2. APPENDIX 2: COMMON GLOBAL CONVENTIONS FOR SENSITIVE SITES ......................................................... 62
11.3. APPENDIX 3: JRC WORKSHOP RECOMMENDATIONS ............................................................................... 62
11.4. APPENDIX 4: COUNTRY SHEETS ........................................................................................................... 63
11.4.1. General information ............................................................................................................ 63
11.4.2. Study area: EU-15 + Norway + Switzerland ......................................................................... 64
11.4.3. Austria .................................................................................................................................. 65
11.4.4. Belgium ................................................................................................................................ 66
11.4.5. Finland ................................................................................................................................. 67
11.4.6. France .................................................................................................................................. 68
11.4.7. Germany .............................................................................................................................. 69
11.4.8. Greece .................................................................................................................................. 70
11.4.9. Italy ...................................................................................................................................... 71
11.4.10. Norway ................................................................................................................................ 72
11.4.11. Portugal ............................................................................................................................... 73
11.4.12. Spain .................................................................................................................................... 74
11.4.13. Sweden ................................................................................................................................ 75
11.4.14. Switzerland .......................................................................................................................... 76
11.4.15. United Kingdom ................................................................................................................... 77
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 6
T A B L E O F F I G U R E S
Figure 1: Overview of countries in study area ................................................................................................ 15
Figure 2: Diagram of the two main steps in the site identification and selection process ............................ 16
Figure 3: Schematic diagram of general approach to GIS-based determination of energy storage potential,
in approximate chronological order from left to right. ................................................................................... 16
Figure 4: Conceptual overview of selection process ...................................................................................... 20
Figure 5: Schematic layout of a PSP................................................................................................................ 21
Figure 6 Excerpt from the database of selected pairs, used as input for the expert selection process ....... 36
Figure 7 An online map of Italy, enabling the user to insert queries for specific site selection.................... 37
Figure 8 Relationship between the three types of potential used in this study ........................................... 39
Figure 9: Map of the study area (non-white countries) showing the realisable potential energy storage
capacity per country in GWh. .......................................................................................................................... 41
Figure 10: Theoretical and realisable potential for each country in the study area, ..................................... 42
Figure 11: Theoretical and realisable potential of PSP sites in the study area: locations of all pairs. ........... 44
Figure 12: Overview of key parameters of all PSP sites in the theoretical potential. .................................... 46
Figure 13 Box plots of energy storage capacity in GWh for all realisable pairs in countries with 2 or more
realisable pairs. ................................................................................................................................................ 47
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 7
G L O S S A R Y
ESC Energy Storage Capacity
GIS Geographic Information System
RES Renewable Energy Source(s)
TSO Transmission System Operator
(VS-)PSP (Variable Speed) Pumped Storage Plants
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 8
P A R T I C I P A N T O R G A N I S A T I O N S
Participant organisation name Short name Country
ALSTOM HYDRO FRANCE AHF France
ELECTRICITE DE FRANCE S.A. EDF France
ELIA SYSTEM OPERATOR ELI Belgium
ALSTOM GRID SAS AGR France
IMPERIAL COLLEGE OF SCIENCE, TECHNOLOGY AND MEDICINE ICL United Kingdom
DNV GL KEM Netherlands
ALGOE ALG France
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 9
1 EXECUTIVE SUMMARY
This report comprises a description of the work done in eStorage Task 4.1.2 and estimates the potential of
new pumped storage plant (PSP) locations by using existing water bodies in the EU-15 countries, Norway
and Switzerland. This task is part of the eStorage project that is funded by the European Commission,
working to develop an economically viable solution supporting large-scale integration of intermittent
renewable energy production into the EU electricity grid. This report provides an estimate of the total
theoretical and realisable storage capacity potential (GWh) of PSPs, providing industries and governments
with an overview of capacity and suitable locations and thus informing decisions to develop new PSP
projects. As the study is limited to the study of existing water bodies, it does not deliver the full potential –
which would include projects where 1 or 2 reservoirs need to be created , or where the sea is considered as
an existing waterbody – but is a fair view of the most cost effective one.
The key objectives of this report are:
1. Compile a list of ‘water body pairs’ suitable for (future) development of new PSP sites for each
country in the EU-15, as well as Norway and Switzerland.
2. Determine a ranked version of this list based on criteria in order to provide the hydro community
with more insight on the expected realisable potential of new PSP sites.
3. Provide an overview of total energy storage capacity per country.
Before starting the study, different approaches on how to identify potential new PSP sites were explored,
including the investigation of the key findings of the study done by the JRC “Assessment of the European
potential for pumped hydropower energy storage”1 and the formation of an expert steering group for this
study that included Alstom, EDF and Elia.
One of the key insights of the JRC study was that restrictions for developing new PSP sites vary greatly
between studied countries based on regional/national legislation and permitting and concession granting
procedures. The strategy developed for this study therefore considered this variation when selecting water
body pairs as potential new PSP sites. In close collaboration with the WP4 expert steering group, the
following strategy was established in identifying potential new PSP sites, combining the strength of
computer models and expertise of national hydro experts:
1. Computer model: identification of potential water body pairs using high-level selection criteria
Using a GIS-based computer program to process and analyse geographical information to
establish/identify potential pairs of water reservoirs suitable for PSP development, based on high
level non-country/region specific selection criteria.
2. National hydro experts: Further refinement of initially established potential water body pairs.
The initially established potential water body pairs in each country are further refined with the
support of national hydro experts based on regional or country-specific selection criteria;
Only existing water bodies are taken into account in this study given the breadth of political, economic and
social challenges associated with building new reservoirs. The results from the GIS model are referred to as
1 Joint Research Center, Assessment of the European potential for pumped hydropower energy storage. Marcos
Gimeno-Gutiérrez and Roberto Lacal-Arántegui (2013)
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 10
the ‘theoretical potential’. This theoretical potential is reviewed and a selection is made by industry experts,
this resulting data is referred to as the ‘realisable potential’.
The main outcomes of the study are summarized below:
• The total realisable potential identified in the study area is 2291 GWh, which is only 33% of the
theoretical potential of 6924 GWh.
• The total number pairs judged by experts to be realisable (117) is only 16% of the theoretical
potential (714 pairs). Expert selection therefore represents a significant element of the study
approach.
• No realisable pairs were found in Denmark, Ireland, Luxembourg or the Netherlands.
• One or two results were found in Belgium, Finland, Germany and Greece, due to lack of sufficient
slope (head divided by distance between water bodies).
• Norway has by far the largest theoretical and realisable potential, both in terms of the number of
pairs as well as energy storage capacity (around 60% of both the theoretical and the realisable
potential in GWh terms is found in Norway).
• Certain regions have a relatively high density of potential PSP sites. Three prominent examples
include the south of Norway, the Alps and the Pyrenees, with about 70% of all theoretical pairs in
the study area and 73% of the realisable capacity. These regions do not seem to coincide with the
locations of major load and generation centres, although this remains to be carefully investigated in
future research.
• Reasons for exclusion differ strongly between countries, presumably due to differences in national
circumstances (type/size/number of nature reserves, regulations, etc.), certain level of subjectivity
of the experts and of course differences in the properties of the theoretical pairs of each country.
• Few existing PSPs have not been identified by the GIS model, first of all because some of the
existing PSPs do not fall within the boundaries of our select ion criteria (e.g. the energy storage
capacity is below 1GWh), secondly, because the data used as input for the GIS model is not always
accurate or complete.
In the light of these findings and based on feedback from the international hydropower community,2 the
following recommendations for future work are formulated:
• To evaluate the true realisable potential for new PSP sites, each water pair should be studied
separately from a business case perspective. Obviously, this was not part of this study given the
limitations in terms of budget and time. However this study was performed with the understanding
that site by site business case analyses by national/regional governments and/or 'PSP utilities' is the
next step. This study aims to provide these parties with an overview and starting point, when they
have the regional/national need/ambition for a new PSP site.
• The study area should be extended to include other parts of Europe to provide a more complete
assessment of the PSP potential in the entire Europe.
• The identified PSP potential should be compared with the need for energy storage in power
generation and demand scenarios predictions for the next decades, and the role of PSP relative to
other flexibility solutions should be determined.
• Because several instances of imperfect data (e.g. duplicates or omissions in water body databases)
have been found to impact the study results, improved GIS source data quality would be highly
recommendable.
• Future studies in this area would benefit from including new reservoirs (still to be constructed) in
the evaluation of PSP potential.
2 The preliminary findings of this study were discussed in a public workshop on October 22, 2014, in London (UK). Final
results were discussed in a public workshop on October 15, 2015, in Birr (CH).
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 11
• It could be considered to apply different selection criteria for each country in order to
accommodate the minimum or maximum need for (additional) PSP capacity for that country. To
define this need, it should be noted that in an interconnected European market one country could
install more PSP than it needs and sell the flexible services to neighbouring countries or vice versa,
in case that is more economical.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 12
2 INTRODUCTION
The transition towards low-carbon Renewable Energy Sources (RES) is a key component of the EC
commitment to reduce greenhouse gas emissions by 80-95% by 2050 compared to 1990 levels. The key
driving forces behind this transition are the mitigation of climate change as well as the reduction of
dependence of EU countries on imported fossil fuels.
Even if low-carbon power generation technologies such as biomass, Carbon Capture and Storage (CCS) and
nuclear power are widely used to realize this ambitions, a large share of electricity would still have to come
from non-dispatchable RES such as wind and solar. However, efficient large-scale integration of RES into
the EU electricity system will require additional sources of flexibility in the system.
Pumped Storage Plants (PSP) are able to store energy when there is a surplus of energy at transmission the
level (e.g. due to strong wind during the night). However, a conventional PSP plant has a limited capability
to regulate the level of its electricity consumption due to fixed-speed motors used for pumping. A PSP plant
with variable speed (VS) motors/generators does not have this limitation and therefore provides additional
regulation capacity when pumping (i.e. storing water in the upper reservoir).
“eStorage” is a project funded by the European Commission, working to develop an economically viable
solution supporting large-scale integration of intermittent renewable energy production into the EU
electricity grid. Its consortium has been awarded a €13.3 million grant by the European Commission
(Directorate-General for Research, within the « FP7 Cooperation: Energy » program). The objectives of the
eStorage project are:
• Demonstrate technical and economic feasibility of upgrading an existing fixed speed pumped hydro
storage to variable speed technology.
• Enhance the functionality of IT systems to develop grid management solutions in line with real-time
market systems.
• Quantify the benefits of an EU-wide rollout of variable speed pumped hydro storage systems under
alternative scenarios.
• Propose changes to the market and regulatory frameworks, to support appropriate business
models for flexible energy storage in the EU.
• Develop and assess technology solutions allowing the upgrade of 75% of European pumped hydro
storage to variable speed to obtain additional capacity for flexible load balancing.
Work package 4 (WP4) of the eStorage project concerns the exploitation of the VS PSP technology and has
the following objectives:
• Survey of possible upgrade of existing PSP into VS.
• Inventory of suitable locations for new PSP sites.
• PSP survey allowing to perform a technology benchmark and allowing to define the potential
market for this technology.
With the overall aim to assess the replication potential of the results of the VS PSP demo from WP1
throughout EU-15, Norway and Switzerland, Task 4.1.2 makes an inventory of potential new PSP locations.
In this study we focus on identifying new potential PSP sites; the decision whether the respective site
should be equipped with VS or a fixed speed motor/generator is left to the PSP developer.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 13
The specific objectives of T4.1.2 are to:
1. Compile a list of ‘water body pairs’ suitable for (future) development of new PSP sites for each
country in the EU-15, Norway and Switzerland.
2. Determine a ranked version of this list based on criteria in order to provide the hydro community
with more insight on the expected realisable potential of new PSP sites. This list will provide
industries and governments an overview of suitable locations, enabling political and business
decisions for development of new PSP projects.
3. Provide an overview of total energy storage capacity per country.
In order to achieve these goals, the specific conditions necessary for installing a PSP need to be taken into
account. Consortium partners participating in WP4 are Alstom, EDF and ELIA, whose experience with
conventional PSP has provided valuable input for this study. DNV GL is the Task leader for task 4.1.2.
In 2013, the Joint Research Center (JRC) has published the results of a similar study, titled "Assessment of
the European potential for pumped hydropower energy storage"1. Although it was not originally planned to
do so (as the study was not available at the time of drafting the eStorage project proposal), this report
draws upon certain useful elements of the JRC study such as selection criteria, expert decisions and
suggestions for datasets. An important distinction between the JRC study and this report is that this study
is divided into two key stages: first, a rough site selection is made by means of a Geographic Information
System (GIS) software model, which is followed by a more refined selection process undertaken by local
expert stakeholders. This approach resulted in a list of potential pairs per selected country that is
potentially very useful for informing industry stakeholders and policy makers.
In this document, we first describe the general approach for T4.1.2 in Chapter 3, explaining the process and
emphasizing the added value with respect to similar studies in the past. Then in Chapter 4 the selection
criteria are described, taking into account the primary selection criteria as well as ranking criteria. In
Chapter 5 the selection of appropriate datasets is discussed and in Chapter 6 the methodology applied in
the GIS model is described. The expert selection process is described in detail in Chapter 7. The study
results are given in Chapter 8.
Chapter 9 presents conclusions and Chapter 10 provides an outlook for future work on this research topic.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 14
3 GENERAL APPROACH
3.1. Objective of task 4.1.2
Prior to devising a general approach to reach the intended results, the required end results of the study
need to be carefully defined. The overall goal of T4.1.2 is to produce a data map consisting of potential new
PSP locations in EU-15, Switzerland and Norway (the study area, see Table 1 and Figure 1), including an
estimate of the total theoretical and realisable storage potential of the PSP technology. This data map will
provide industries and governments an overview of suitable locations, informing political and commercial
decisions to develop new PSP projects.
Based on this objective and further input from the Steering Committee, the objective was more precisely
defined as follows:
• to provide a ranked list of potential PSP (variable speed or otherwise) sites for the study area:
o with a site defined as a pair of suitable existing water bodies.
o with at least geographical location and energy storage potential for each site.
o taking into account criteria for selection or exclusion.
o taking into account ranking criteria.
• to provide an overview of total energy storage capacity potential per country:
EU-15
Austria Belgium Denmark Finland
France Germany Greece Ireland
Italy Luxembourg Netherlands Portugal
Spain Sweden United Kingdom
Other Norway Switzerland
Table 1 Countries in study area
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 15
Figure 1: Overview of countries in study area
3.2. Approach
Based on the expected results, the following analytical approach has been developed. First, because of the
geographical nature of the problem and the volume of data involved (e.g. altitude differences, distances,
disfavoured areas, number of water bodies to be processed, etc.), it was evident that a GIS model would be
required for maximum efficiency.
In addition, from the discussions with the Steering Group it was clear that the GIS model was not sufficient
in itself to capture the complex and sometimes arbitrary process of potential site selection. It was therefore
decided to add an expert site selection stage to the process as an important follow-up step. In this stage
industry experts are asked to provide expert judgement on whether the model-selected pairs (theoretical
potential) could actually be implemented in reality, or they need to be excluded for various reasons related
to e.g. environmental restrictions or concessions.
The potential PSP locations resulting from the GIS model are referred to as the theoretical potential.
Subsequently this theoretical potential was presented to industry experts to come to a more realistic
potential, i.e. the realisable potential. The approach can thus roughly be divided into two steps, this is
also schematically depicted in Figure 2.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 16
Figure 2: Diagram of the two main steps in the site identification and selection process
During the execution of the study, experience gained led to revisions and adjustments, ultimately
culminating in the process depicted in Figure 3 below, representing the final version of how activities were
structured within Task 4.1.2.
Four parallel sub-processes can be identified in the approach:
• Criteria development: drawing up explicit and quantitative criteria for selection and ranking of
potential sites.
• Model development: design, programming, application and fine-tuning of GIS model.
• Data processing: selecting, checking and preparing GIS source data.
• External input and steering: collecting and integrating outside information and feedback.
Review of
criteria / model / data
Early
results
(1 country)
Rough selection criteria
Theoretical
potential
(study area)
Improved
selection
criteria
Final
selection
criteria
Model development
Steering
Committee
Steering
CommitteeProject plan
Model
finetuning
Finalisation of datasets
Model
execution
Database
existing storage
(4.1.1)
LiteratureSteering
Committee
Realisable
potential
(study area)
National
experts
Dataset
collection
Data
quality
assessment
Data
preparation
Figure 3: Schematic diagram of general approach to GIS-based determination of energy storage potential, in
approximate chronological order from left to right.
Four parallel sub-processes form the overall process: selection and ranking criteria development (dark orange), GIS
model development (brown-yellow), data processing (dark blue), and external input and steering (top row). Green
boxes represent interactive phases
3.2.1. Key activities
In the paragraphs below, the key activities constituting the general approach used in Task 4.1.2 (as depicted
in Figure 3) are briefly explained. The ordering of the activities is mostly chronological, although many
activities in different sub-processes take place simultaneously (e.g. criteria development and dataset
collection). External input and steering-related activities are discussed separately.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 17
3.2.1.1. ROUGH CRITERIA
At the start of the general approach a rough set of criteria was generated for the selection of potential PSP
sites. For this purpose, quantitative technical criteria needed to be defined to be used by the GIS model and
possibly by the industry experts performing the selection subsequently. Two types of criteria were devised:
• Selection criteria to include suitable water bodies (individual or groups) for the final list of potential
sites (or conversely, to exclude unsuitable ones).
• Ranking criteria were used to arrange the list of selected sites in order of suitability to facilitate the
expert selection process.
Inputs for these criteria came from the Steering Committee, relevant literature, as well as the task objective
defined in the project plan.
3.2.1.2. MODEL DEVELOPMENT
As soon as an early draft of the criteria had been developed, the model development commenced. A GIS
software tool was selected in which a model is built that automatically applies the criteria to source data,
thereby generating the results. In other words, model steps and functionality are the translation of the
criteria into computerized steps, and were therefore developed and refined in parallel with the
development of the criteria.
3.2.1.3. DATASET COLLECTION
Also based on early draft of the criteria, the collection of source data for the analysis was initiated.
Available (GIS) datasets for all aspects relevant for the adopted criteria needed to be identified and
collected. For example, a criterion on reservoir storage capacity means that one or more datasets
containing information on reservoirs, their altitude and volume needed to be acquired. The data collection
process was carried out in parallel to criteria refinement, and was only finalised when no more
fundamental changes (i.e. new aspects taken into account) were made to the criteria and the data quality
was considered sufficient.
3.2.1.4. DATA QUALITY ASSESSMENT
Data quality needed to be checked for all datasets collected, to be able to assess the value of the final
results of the investigation. Also, for some types of data like altitude data, several datasets were available
and the most suitable one needed to be selected. If only a single source was available and the quality was
considered insufficient, alternative actions might be taken such as criteria adjustment or generating new
data based on other sources, depending on the specific case. Quality-determining factors included spatial
resolution, coverage of our area of interest, presence of voids and anomalies, inclusion requirements (e.g.
reservoirs with minimum volume), completeness, metadata quality, etc.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 18
3.2.1.5. DATA PREPARATION
Given that the acquired GIS data was not directly usable in the GIS model developed for this study, it
required certain effort to reformat, merge separate datasets, create geographical subsets (i.e. organise
data across investigated countries), etc.
3.2.1.6. REVIEW AND REFINEMENT
After constructing a working model, based on sound criteria and using suitable data of sufficient quality, a
model review round was necessary. Questions and issues that arose during this first stage were evaluated
and decided upon, and the preliminary model output was compared against the objective and the data on
existing storage sites gathered within Task 4.1.1. This led to refinements made to the model, criteria and
data collection. The Steering Committee had an active role in this general review stage.
3.2.1.7. EARLY RESULTS FOR ONE COUNTRY
The next key step was the generation and review of early results for one particular country (France) in the
area of investigation. The analysis of early results has led to final fine-tuning of the criteria and the model.
The impact of ranking criteria, in particular their weighting factors, was carefully studied and adjusted
where necessary, with the support of the Steering Committee.
3.2.1.8. FINAL RESULTS: THEORETICAL AND REALISABLE POTENTIAL
When generating the final results, applying the ranking criteria required certain manual adjustment, using
the Steering Committee’s knowledge and experience from existing sites to fine-tune intermediate rankings
to more realistically reflect suitability. The unranked output of the finalised GIS-model (the theoretical
potential) is a list of potential PSP sites with a number of properties associated to each site. The ranking
process assigned a value and a weighting factor to each property.
The initial established potential water body pairs for each country are further refined by national hydro
experts using region/country specific selection criteria.
3.2.2. External input and steering
The factors that are mentioned below provide valuable input from outside the influence of eStorage task
4.1.2 that improves the process and its results.
3.2.2.1. STEERING COMMITTEE
The Steering Committee plays an invaluable advisory and steering role in various phases of the process.
Their experience, knowledge, and databases on existing sites ultimately increase relevance and realism of
the final results. The Steering Committee consists of the following parties:
• Alstom Hydro (France): PSP technology provider.
• Elia (Belgium): TSO.
• EDF (France): utility and PSP end user.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 19
3.2.2.2. LITERATURE
Naturally, scientific and professional literature provided input to the work performed, mainly on technical
matters such as the implementation of criteria as a model step, or suitable values for key parameters.
Metadata of datasets was an important source of information too. Concerning scientific reports, special
mention has to be made of the report1 undertaken by the JRC on this topic which describes the approach
and results of a similar study.
3.2.2.3. DATABASE EXISTING STORAGE (ESTORAGE TASK 4.1.1)
Within eStorage task 4.1.1, a database was built containing, among other information, properties of PSPs
installed based in Europe like capacity and number. This information, especially when combined with know-
how of the Steering Committee, served as validation for intermediate results and supported data quality
assessment and model development.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 20
4 CRITERIA
4.1. Introduction
In order to meet the objective of Task 4.1.2, to develop a data map consisting of theoretical and realisable
energy storage potential of new PSP locations in EU-15, Switzerland and Norway, the first objective was to
design criteria for the selection of potential locations for new PSPs. Criteria for the selection and exclusion
of potential locations for new PSPs have been selected based on interviews, literature review and input
from the WP4 Steering Group (consisting of Alstom, EDF and ELIA).
The criteria for selecting appropriate sites for PSP are divided in two parts: selection criteria and ranking
criteria. Whereas the application of selection criteria leads to selection of combinations of water bodies and
exclusion of non-conforming ones, ranking criteria give insight in the suitability of selected combinations.
The ranking criteria are not directly used for selection/exclusion, but rather to inform industry experts in
their selection process to judge which pairs of each country’s theoretical potential are actually realisable.
Theoretical
potential
Realisable
potentialPrimary
selection
Expert
selection
Pre-ranking
Additional
information
Figure 4: Conceptual overview of selection process
The development of selection and ranking criteria has proven to be an iterative process where discussions
of project outcomes with the steering group led to adjustments of existing criteria and development of new
ones. A crucial ingredient for a successful outcome of this task has been the involvement of industrial PSP
parties to ensure that the right selection steps and selection criteria are used. PSP experts from EDF (owner
and operator of PSP plants), Alstom (supplier of PSP plants) and Elia (TSO) have participated in a dedicated
steering group for this task. This steering group has actively been requested to provide feedback and input
to the work process as well as selection criteria in order to increase the adequacy and usefulness of the
task’s outcomes.
The outcome of this exercise is a verified list of criteria for the selection of suitable new (VS) PSP locations
that can be used for dataset selection (chapter 4) and the analysis in the software program ArcGIS (chapter
5).
The following sections provide an overview of water body types, selection criteria and the weighting
methodology.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 21
4.2. Water body types
At the kick-off meeting for WP4 of the eStorage project, the participants decided that only those options
for new PSP locations are taken into account where two water bodies (i.e. reservoirs and/or lakes) already
exist and are within suitable distance and difference in elevation. Water bodies that have not yet been
constructed are not taken into account in this study. The rationale behind this is that circumstances for
building new reservoirs are politically, economically and socially challenging. A sufficient potential based on
existing water bodies is thought to exist and only this proverbial low-hanging fruit for PSPs is considered
relevant for this study.
Based on recommendations from the steering group it has been decided to limit this study to existing water
bodies. This included the following categories:
• Reservoirs for electricity generation purposes.
• Reservoirs not for electricity generation purposes.
• Other water bodies, including lakes3.
Any combination of two of these types can form a potential PSP location.
4.3. Selection criteria
Selection criteria are used to select pairs of water bodies with a GIS model that are potentially eligible to be
developed as PSP. Criteria are in the following categories (see also Figure 5):
1 Energy storage capacity.
2 Distance between water bodies.
3 Head.
4 Inverse average slope between water bodies: distance divided by head.
Penstock
Lower water body
Upper water body
Powerhouse
Distance
Figure 5: Schematic layout of a PSP
3 In this report, PSP locations are frequently loosely referred to and treated as ‘reservoirs’, but strictly speaking this
should be ‘water bodies’, as in some instances one or both locations are a lake.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 22
4.3.1. Energy storage capacity
The energy storage capacity depends on the volume of the reservoir and on the head of the pair. The
calculation of the capacity is discussed in chapter 6.4.
With the steering group it was decided to only select potential pairs which have an energy storage capacity
greater than 1 GWh. The rationale is that these are more interesting for PSP development than sites with
smaller capacities.
For reservoirs, as a general rule, a maximum of 80% of the total volume was allocated as usable volume, in
line with expert opinions4 and literature
1. For pairs with a distance between the reservoirs greater than
300m though, a usable volume of 100% is taken for the upper reservoir5. For lakes (natural water bodies
without dams), only the top meter was considered usable for pumped storage.
To minimize the environmental impact for the third category (other water bodies, such as lakes) a water
level fluctuation restriction is applied. Depending on the country of research, the size of this water level
fluctuation is defined (e.g. for France the maximum water level fluctuation for ‘other water bodies’ is 1
meter).
4.3.2. Distance between reservoirs
As the penstock is one of the most expensive parts of a PSP, the distance between the reservoirs is an
important contributor to the cost of the development of the PSP. In discussion with the Steering Group, a
maximum distance of 10 km was selected between the water bodies.
4.3.3. Head
The head is the difference in elevation between two water bodies. In line with the JRC study1, it was first
suggested to use a minimum head of 150 m. The Steering Group has recommended using 80 m instead,
considering the multitude of PSPs developed in Portugal with heads below 100 m. In another analysis of
existing PSP potential, Alstom found that 23 out of 342 units (7%) had a head lower than 80 m, further
supporting the validity of the chosen value.
4.3.4. Average slope between water bodies
For a PSP to be technically viable, the distance between the water bodies should be in relation to the head
to limit friction losses. Together with the steering group a minimum average slope of 5 % was adopted.
4 Interview with EDF, Alstom Hydro
5 Interview with EDF
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 23
4.4. Automatic pre-ranking
In order to facilitate the expert selection process, a weighting of the most important parameters has been
integrated into the model. Weighting is done for:
• Energy storage capacity: the amount of energy to be stored is most critical to decide whether a pair
is economically viable or not. Therefore this parameter is weighted with a factor of 2
• Inverse slope: distance between water bodies divided by the head is the inverse slope. The higher
the inverse slope, the more attractive the pair is to develop as a PSP. Nevertheless, this parameter
is less important than energy storage capacity and is therefore weighted with a factor of 1.
In discussion with the Steering Group it has been decided not to add grid availability to the parameters to
be weighted, since the value for this parameter is depending on various other factors such as the length of
the period for delivering power as well as the capacity of the grid to incorporate the PSP’s power.
4.5. Automatic addition of relevant characteristics
In order to facilitate the expert selection process, some important characteristics of the selected pairs are
made visible. These characteristics are defined in collaboration with the Steering Group.
4.5.1. Distance to grid
The availability of an appropriate grid connection is an important factor influencing the suitability of a
potential pair. To facilitate the expert selection process, the distance to three different voltage levels is
included in the results. Based on the steering group suggestion, the following voltage levels are used:
• 150-220 kV
• 220-400kV
• > 400kV.
4.5.2. Overlap with restricted zones
Certain areas are less- or unsuitable for the construction of new PSP infrastructure. This can be because of
already existing constructions or because regulations restrict certain activities. Together with the steering
group the following restriction factors are identified which can exclude or complicate the construction
process of new PSP infrastructure:
• Global or European conventions:
o World Heritage Sites (UNESCO),
o Man and the Biosphere (UNESCO),
o Wetlands of International Importance (Ramsar),
o Natura2000.
• National protected areas (classified according to IUCN category system):
o Strict nature reserve,
o Wilderness area,
o National park,
o Natural monument or feature,
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 24
o Habitat/species management area,
o Protected landscape/seascape,
o Protected area with sustainable use of natural resources.
• Urban land use:
o Industry,
o Residential,
o Cemeteries.
4.6. Overview
Below an overview is presented of a) selection criteria (Table 2), b) pre-ranking parameters (Table 3) and c)
relevant characteristics of potential pairs (Table 4).
Criterion Value
Energy storage capacity > 1 GWh
Distance between water bodies < 10 km
Head > 80 m
Average slope > 5%
Table 2 Selection criteria
Parameter Weight factor
Energy storage capacity 2
Inverse slope 1
Table 3 Pre-ranking parameters
Characteristic Information type
Distance to grid value in km
Overlap with restricted zones list of zones, map (geographical data)
Table 4 Relevant characteristics of potential pairs
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 25
5 DATASET SELECTION
5.1. Introduction
Based on the criteria discussed in the previous chapter, the following categories of datasets are required.
Each category will be discussed in more detail in subsequent paragraphs:
• Country boundaries,
• Water bodies,
• Reservoirs and dams,
• Elevation,
• Grid data,
• Sensitive areas.
5.2. Country boundaries
This dataset will be used to clip all datasets to the required geographic area. Different datasets are
available and for this study we selected the Global level Country Boundaries (2011) from the diva GIS
website6.
5.3. Water bodies
The water body dataset is an important input for this study. As it contains the collection of water bodies
from which the PSP sites (water body pairs) will be selected, it will be the foundation for the final result.
Ideally it should contain only still water bodies and no rivers. Multiple water body datasets are available,
including the following:
• Global Lakes and Wetlands Database7,
• European Catchments and Rivers Network System (ECRINS)8,
• SRTM Water Body Dataset9,
• World Water Bodies (WWB)10
,
• OpenStreetMap water layer11
,
• Corine Land Cover (CLC)12
,
Many of these have similar characteristics; key distinguishing features are described in the paragraphs
below.
6 http://www.diva-gis.org/data/misc/countries_shp.zip
7 http://worldwildlife.org/pages/global-lakes-and-wetlands-database
8 http://www.eea.europa.eu/data-and-maps/data/european-catchments-and-rivers-network
9 ftp://xftp.jrc.it/pub/srtmV4/SRTM_Mask_ArcAscii/
10 http://www.arcgis.com/home/item.html?id=3c659b5a29334e54a4887b7531a9a858
11 http://osmdata.thinkgeo.com/
12 http://www.eea.europa.eu/data-and-maps/data/clc-2006-vector-data-version-2
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 26
5.3.1. Level of detail
The level of detail of the water bodies in the OpenStreetMap dataset, and to a lesser extent in the ECRINS
dataset, is higher than in the others. However, especially in highly detailed datasets, many of the larger
water bodies are split into smaller features to improve loading time for (online) viewing. Because water
body areas are required for calculating their volume it is important to work with the complete water bodies
and not fragments. The Corine Land Cover dataset is in that sense ideal, as it contains only single water
body features.
5.3.2. Accuracy
No difference in accuracy, i.e. the extent to which the features match reality, is detectable for these water
body datasets. It must be noted that the source and generation method of OpenStreetMap water layer
data is not known, and that some rivers are also included which could influence the result.
5.3.3. Selected dataset
The best compromise between detail and usability was reached by using the Corine Land Cover
supplemented by the OpenStreetMap water layer. If at a later stage the conclusion is drawn that the high
level of detail is not required or complications are occurring with the accuracy of OpenStreetMap data, the
best alternative would be ECRINS combined with a method to merge water body segments of single water
bodies.
5.4. Reservoirs and dams
For the purpose of this study, reservoirs are defined as dammed water bodies. Because in general the
limitations for PSP are different (less restricting) for reservoirs than for other water bodies it is important to
know the location and purpose of the reservoirs and dams. The reservoir purpose is required to estimate
the usable volume for the different types of water bodies. Several reservoir and dam databases are
available including:
• Global Reservoir and Dam (GRanD) Database 13
,
• DAMPOS database14
,
• ICOLD dam register 15
,
• European Catchments and RIvers Network System (ECRINS).
The ICOLD dam register is the most extensive global database on dams. Unfortunately this dataset does not
include the location of its dams, so for this study it is not suitable by itself. The GRanD databse is the result
of an international effort to collate existing dam and reservoir datasets with the aim of providing a single,
13
http://sedac.ciesin.columbia.edu/data/set/grand-v1-reservoirs-rev01 14
http://dampos.eionet.europa.eu/ 15
http://www.icold-cigb.org/
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 27
geographically explicit and reliable database. However, in comparison to the ICOLD dam register, many
dams are not included16
. DAMPOS is a dam database and web tool included in the European Environmental
Information and Observation Network (EIONET). It is based on the ELDRED2 database, which has been
coupled to the ECRINS database. It was determined that the DAMPOS database in combination with the
GRanD databaseto would provide the most extensive foundation to identify reservoirs.
5.5. Elevation
The altitude difference between water bodies is an important factor influencing the suitability of a
potential PHS site. A plethora of elevation maps is available for this, but only few cover a significant part of
the study area and have a suitable resolution. The following ones are seen as most promising:
• Shuttle Radar Topography Mission (SRTM), version 417
,
• Enhanced SRTM by Ferranti18
,
• Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), version 219
,
• Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010)20
,
• World 30 v2.0 DEM21
.
The SRTM is a 90m resolution elevation map with coverage between 60 degrees north and 56 degrees
south. Different versions are available including the one enhanced by Ferranti, which has an extended
coverage by combining elevation from different sources, mainly digitised Russian military contour maps.
ASTER has a resolution of 30m and near global coverage. The accuracy of both ASTER and SRTM are in the
same range. Although ASTER data seems more favourable on paper, SRTM is in many cases preferred, due
to the amount of voids (holes) and spikes (errors) still present in the final ASTER product. GMTED2010 is
currently the most detailed freely available elevation map with global coverage. However, the resolution of
GMTED2010 is 250m and this is too coarse for this study. World 30 DEM is a commercial product which
combines ASTER, SRTM and GMTED2010, resulting in what the supplier claims to be the most accurate
global terrain model. The price of this product is considerable; data for an area the size of Austria costs
around EUR 1200.
Because the study areas reaches further than 60 degrees north, SRTM alone is not sufficient. The Enhanced
SRTM by Ferranti would be option, but because no independent data is available on the accuracy of the
added data, the reliability of the final results might be compromised. The potential advantages of the
World 30 DEM are considered too little for the cost. It is therefore decided to go for SRTM, complemented
by ASTER for the northern countries.
16
JRC (Joint Research Centre) of the European Commission, 2011: Pumped-hydro energy storage: Potential for
transformation from single dams. Lacal Arantegui, R., Fitzgerald, N., and Leahy, P. (2011). Joint Research Centre —
Institute for Energy and Transport; Petten, The Netherlands. Available at http://setis.ec.europa.eu/setis-
deliverables/studies-and-reports/report-pumped-hydro-energy¬storage-potential-transformation 17
ftp://xftp.jrc.it/pub/srtmV4/tiff/ (v4) 18
http://www.viewfinderpanoramas.org/Coverage%20map%20viewfinderpanoramas_org3.htm 19
http://gdem.ersdac.jspacesystems.or.jp/ 20
http://topotools.cr.usgs.gov/gmted_viewer/gmted2010_global_grids.php 21
https://store.intermap.com/MapShop.aspx?GeoLocation=World#
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 28
5.6. Grid data
The suitability of a potential PSP also depends on the availability of a suitable grid connection the transport
the required/produced energy. Determining factors are distance to and voltage levels of the power lines
and substations.
• European Network of Transmission System Operators for Electricity (ENTSO-E) network data
• Enipedia22
• OpenStreetMap23
• Platts transmission lines and substations24
.
ENTSO-E is an association of Europe's transmission system operators (TSOs) for electricity. For the most
detailed and accurate information on the European grid, their data is the best option. However, due to its
sensitive nature the information is not freely available. To obtain this data, a request was submitted to
ENTSO-E but no decision has been taken yet regarding this matter. Freely accessible grid data is available in
the public domain, but the reliability is uncertain. Freely available sources are Enipedia, which is based on
the ITO electricity distribution map (in turn based on OpenStreetMap). Another option is to directly extract
grid data from OpenStreetMap itself. A problem for free data is that only information on the power lines is
available, not on substations and their voltage levels. Platts energy data is commercial data collected from
different sources, and includes the energy infrastructure for North America and Europe, with unknown
accuracy. Pending ENTSOE’s decision to grant access to their data, the limited power line data from
Enipedia is the selected option.
5.7. Sensitive areas
Within or close to areas of certain land use types, construction or operation activities may be restricted or
forbidden. Relevant land use types for PHS are natural, cultural and urban areas, which therefore must be
taken into account in the GIS-model. The following datasets give information on the location of these areas:
• Natura200025
• World Database on Protected Areas (WDPA)26
• OpenStreetMap Landuse 27
• Corine Land Cover (CLC).
As the name already suggests, the Natura2000 dataset gives the boundaries of all areas included in the
Natura2000 network of nature reserves. The World Database on Protected Areas (WDPA) is the most
comprehensive global spatial dataset on marine and terrestrial protected areas available. The different
national protected areas are classified according to the IUCN category system (Appendix 1). International
protected areas are classified according to respective independent standards. The most common
22
http://enipedia.tudelft.nl/maps/PowerPlants.html 23
http://www.openstreetmap.org/#map=7/49.382/15.309 24
http://www.platts.com/products/gis-data 25
http://www.eea.europa.eu/data-and-maps/data/natura-4#tab-metadata 26
http://protectedplanet.net/ 27
http://osmdata.thinkgeo.com/
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 29
international conventions are mentioned in Appendix 2. OpenStreetMap and CLC both contain a standard
land use classification. While CLC would be a good alternative, in this study OpenStreetMap was selected
because of the higher level of detail.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 30
6 METHODOLOGY
6.1. Introduction
To reach the objective of Task 4.1.2, a GIS model was designed with the objective to systematically and
automatically identify specific locations suitable for PSP application within the EU15 plus Norway and
Switzerland. For this analysis ESRI’s ArcGIS 10.1 software package was chosen. ArcGIS is the globally
accepted standard in GIS software, and its unique model builder functionality allows for relatively easy
model construction and modification.
6.2. General approach
The approach for this analysis is to start with all possible water bodies, and then narrow down site options
(combinations of two water bodies) by applying the criteria as discussed in Chapter 3. The most ‘selective’
criteria are applied first so that those sites are left with a high potential. In subsequent steps, these sites
are then to be investigated in further detail. The analysis is performed on a country-by-country level in
order to keep the amount of data and – subsequently - processing time to an acceptable level.
6.2.1. Coordinate System
Geographic data can be displayed with a geographic or projected coordinate system. Geographic
coordinate systems use latitude and longitude to determine position of a point or element on
mathematical representation of the word. A projected coordinate system refers to data that is defined by a
flat 2-D surface (map) and can be measured in units of e.g. meters or feet. The advantage of a projected
coordinate system is that the length of a unit is constant over the whole map, but depending on the type of
projection the shape and distance between features can be distorted. To minimize the error in a projected
map different projections with different purposes (and for different scales) have been developed. Because
in this study it is important to know the area and spatial relation between different objects, it is required to
project the data. Because the study is focusing on a European scale, and the purpose lays within the
European context it is decided to follow the recommendations of the workshop organized by JRC and the
European Commission in the year 2000 on the implementation of a single geographic system for all
European countries (see appendix 3 for the relevant recommendations of this workshop). In line with this
recommendation, ETRS-LAEA has been selected as projection standard for this study.
6.3. Data preparation
As described in Chapter 4, a broad collection of datasets is used in this study. Before these could be
included in the analysis, preparation steps were required to make the data suitable for use. In general these
pre-processing steps included selection of relevant data and normalisation of names and units. To limit
processing time, only data required for the analysis was retained; other data was discarded. Data
preparation also included clipping to the study area, from where during the main processing step a sub-
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 31
area (country) could be extracted and used for the main analysis, Because further on in the analysis the
model automatically selects and clips the data required for the next step(s) it is very important that all
datasets used the same naming standards. To prevent such issues later on all datasets are therefore in
advance checked for consistency.
For the elevation and water body datasets specific steps were required. Both elevation maps described in
chapter 4 have their disadvantages. SRTM only has coverage up to 60 degrees north, and ASTER has more
inconsistencies. For optimal results both were combined into a full coverage map. The water body dataset
was generated by supplementing the CLC dataset with OpenStreetMap features.
6.4. Primary selection
As described in paragraph 4.3, the primary selection process consists of the application of the criteria listed
in Table 5.
Value
Energy storage capacity > 1 GWh
Distance between water bodies < 10 km
Head > 80 m
Average slope > 5%
Table 5 Primary selection criteria
The criteria above are applied in consecutive steps on the datasets. In each step, the collection of selected
potential PSP sites is further refined. In some steps data is added to each reservoir (pair), such as altitude
difference and mutual distance, required the evaluation of criteria in subsequent steps.
Step 1: Selection of required data
In the data preparation step all different datasets are merged into one base dataset for all the separate
categories (elevation, water bodies etc.). During the primary selection phase, the analysis is conducted for
each country separately, so the first step is to clip the base datasets to the relevant country boundary.
The water body dataset is not always correct about the use of water bodies as reservoirs for electricity
production. Therefore, water bodies are classified as reservoirs when a hydro dam location (from the
DAMPOS or GRAND database) is overlapping with that water body. To accommodate for accuracy errors
also water bodies within a distance smaller than 100 m to a hydro dam point are included.
By definition, watercourses are excluded from the selection. To include reservoirs on a river the following
selection process is included:
• Watercourses which overlap with hydro dam points are split on the intersection.
• The watercourse is clipped by a buffer of 3500 m around the dam location. The value of 3500 m is
based on practical reasoning, based on discussions with EDF.
• Of the two split parts the larger one is selected as the upstream reservoir (based on the assumption
that within the buffer the upstream reservoir is bigger than the downstream river).
Additionally, some dam locations do not overlap (within a distance of 100 m) with any feature in our water
body database. Many of these are very small dams within a small water body or stream, and are therefore
not relevant for the study. However some could potentially be relevant and are therefore included in our
model by the following procedure:
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 32
• Hydro dams that do not overlap with any water body feature with a dam height >15m and known
retention area are selected.
• These selected dam locations are buffered with the buffer size identical to the retention area.
• This buffer is used as the relevant water body feature in the rest of the model.
All types of reservoirs are selected from the DAMPOS database (not just primary and secondary electricity
generation purpose dams).
Step 2: Water body preselection based on usable volume
The water body dataset contains reservoirs and lakes, but also other features not relevant to this study. To
limit the number of water bodies which need to be processed, a first selection is applied. From the
minimum energy storage capacity mentioned in the selection criteria, a minimum required usable volume
can be derived. As a data source for usable volume, the DAMPOS dataset is used (see section 4.4).
In case data is omitted, the volume is calculated. For calculation of this volume, equation 1 can be used for
reservoirs with hydro storage as primary or secondary purpose (see below).28
The usable water body
volume is assumed to be 80% of the volume of the reservoir, except for pairs with a distance between the
reservoirs above 300m in which case 100% of usable volume is taken for the upper reservoir. For lakes a
water level fluctuation restriction is applied. The extent in meters of this water level fluctuation is defined
per country of research.
(1)
(2)
where:
V = volume (106 m
3)
A = water body surface area (m2)
x = acceptable water level fluctuation (m)
Step 3: Extraction of water body altitude from DEM
For each water body, the center point29
is selected and for this point the elevation is extracted from the
Digital Elevation Map (DEM). The information is added to the original water body database.
Step 4: Selection of potential pairs
As the number of possible unique pairs of elements in a data-set increases with the square of the number
of elements, the number of water body pairs can become impractically large if all possible combinations are
generated and processed. Therefore, the number of pairs processed is limited by assuming a maximum of
10 km for the distance between any two water bodies, excluding pairs further apart. For purpose, a buffer
zone of 5 km is created around the features in the water body database. Two water bodies whose buffers
overlap fulfill the requirement of having a mutual distance of less than 10 km, and are then selected by the
model as a potential pair.
28
According to the corresponding technical manual, the reservoir volume formula was derived by
statistical regression analysis on 5824 reservoirs in the GRanD database for which both area and volume are known,
and has a fit quality R2 = 0.8. To reflect the usable volume (as opposed to total volume) of a reservoir, in this study a
correction factor of 0.8 was added. 29
Centerpoint is the center of a feature and is always located within that feature.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 33
Step 5: Path between water bodies (penstock path)
In the primary selection process the shortest distance between the two water bodies has been calculated,
but the actual path was not generated, i.e. the connection was not created and drawn. Because the
connection represents the assumed penstock path which is required for checking overlap with sensitive
areas, it needs to be constructed now, which takes place in a two-step process. First, for each of the two
water bodies, the boundary polygon is converted into a set of equidistant points at every 30m. Then, lines
are drawn from each point of one water body to each point of the other. The shortest of these lines is
assumed to be the most suitable path and is retained as the penstock path.
Step 6: Calculate head and energy storage capacity
The head for each potential pair is calculated by taking the absolute value of their difference in altitude (eq.
3). Subsequently the energy storage capacity of each potential pair can be calculated (eq. 4), using the
smaller of the two usable volumes of the water bodies (data added in step 2) and an assumed overall
power generation efficiency of 90%30
.
(3)
(4)
where:
∆h = altitude difference between reservoirs (m)
hrx = water height reservoir x (m AMSL)
E = energy storage capacity (MWh)
g = gravitational acceleration (m/s2)
ρ = density of water (kg/m3)
V = water body volume (106 m
3)
= power generation efficiency (-)
Step 7: Slope calculation
The slope of the penstock that would connect a pair of water bodies is calculated (eq. 5) from the head and
distance between two water bodies. The head has already been calculated in step 5, but the distance
between the pairs of water bodies is not yet available. A calculation of the shortest distance between each
water body pair is therefore performed (i.e. the length of the penstock path is determined, see step 5),
after which the slope is calculated.
(5)
where:
S = slope
∆h = altitude difference between reservoirs (m)
d = distance between reservoirs (m).
30
Generally accepted average power generation efficiency for PSP
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 34
6.5. Ranking parameters
The ranking phase entails gathering information to enable subsequent manual ranking, of each water body
pair that was selected by application of the selection criteria (step 1 through 7 above). This phase will not
result in changes in the selection of water body pairs, but it will add information and features required for
the ranking (see chapter 3 for more elaborate description of this process).
Step 1: Energy storage capacity weighting
To be added.
Step 2: Inverse slope weighting
This leads to a pre-ranking for the parameters listed in Table 6, as discussed in paragraph 4.4.
Parameter Weight factor
Energy storage capacity 2
Inverse slope 1
Table 6 Ranking parameters
6.6. Relevant characteristics of potential pairs
Step 1: Path between power station and electricity grid (power line)
The path from the powerhouse to the electricity grid (the power line path) is constructed in a similar way as
in step 1. The powerhouse is assumed to be located where the path between the water bodies (see above)
touches lowest one. The grid connection point cannot be any point in the grid, but is assumed to
necessarily be a substation. Similarly as for the penstock path between the water bodies, the power line
path is assumed to be the shortest line between the two, and constructed as such.
Step 2: Overlap with restricting factors
The aforementioned penstock and powerline paths are merged into an infrastructure path. For each pair
this path is overlaid with the sensitive sites (buffered, if applicable) and if overlap occurs, this fact is added
to the data associated with the potential site. If the infrastructure path is overlapping with more than one
restricting zone only the zone with the highest significance is added. This significance is a value added in
advance to the datasets describing the sensitive sites and can be changed manually.
Step 3: Exclusivity of a water body
In the list of potential pairs it is possible that one water body is part of two or more pairs, i.e. non-exclusive.
The output list of the model shows which pairs are part of a system of pairs and therefore may not all
realisable simultaneously, even if a pair may be realisable on its own. The national experts involved in the
expert selection process check which pairs of a system are realisable and subsequently which of these pairs
must be discarded because their shared water body / bodies are already in use.
6.7. Results processing and visualisation
The model will create a list of potential PSP sites, showing all suitable water body pairs. Ranking parameters
and subsequent weighting factors lead to an overall suitability score per pair, and therefore a ranking of all
potential sites. The ideal location for further development however depends not only on the calculated
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 35
suitability score, but also on e.g. the availability of logistical infrastructure, energy storage demand in that
specific location, the resources of the investor and some other factors like geology for the penstock.
To facilitate the expert selection process, the following information has been developed:
• overview maps of the suitable location including the location of the water bodies, constructed
penstock and power line paths, key properties, and restricting factors (indicating proximity to
sensitive sites),if any.
• graph (per country) of the energy storage capacity of all potential sites, including the relative
position of the specific pair.
6.8. Validation
To (partially) validate the combination of GIS-model and dataset for the purpose of this study, it is useful to
compare the obtained theoretical potential with actual data on PSPs and water bodies. For this purpose,
the theoretical potential of France was taken as an example. Modeled values for all 35 pairs were
compared to the national expert’s accurate data (unavailable for a few water bodies), leading to the
following analysis:
• average absolute difference in elevation: 2.3%. One outlier not taken into account with modeled
elevation 66% higher than actual elevation
• average absolute difference in head: 5.9%
• average absolute difference in usable volume: 34%. Many modeled volumes either quite accurate
or around 20% lower than actual volumes. One outlier not taken into account with modeled
volume 2207% (!) higher than actual volume
• number of existing PSPs found: 4 out of 5 matching the selection criteria (one additional PSP exists
which does not match the minimum capacity criterion). The cause for not finding the remaining PSP
is one of its reservoirs being absent from the water body database.
From the analysis above it can concluded that:
• elevation and head differences are quite low,
• moderate differences occur in usable volume (and therefore energy storage capacity),
• disproportionally deviating values occur (2 out of 35 for France) for unknown reasons,
• suitable pairs might be incorrectly excluded due to missing water bodies in the database.
It should be noted that properties of water bodies not included in the theoretical potential were not
compared. However, there is no reason why these water bodies would have properties with a different
accuracy than what was found above.
Ultimately, it was concluded that the theoretical potential has a sufficient quality for the purposes of this
study. It is recommended to look into quality of the input datasets to try to improve the accuracy of model
outcomes (see also Outlook), especially regarding the usable volumes and completeness of the water body
database.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 36
7 EXPERT SELECTION PROCESS
7.1. Introduction
As the model is based on generic criteria and does not take into account regional differences, the next step
in the PSP selection process is the invitation of PSP experts to perform the so-called ‘expert selection
process’. Here, the expert will support the selection process by comparing the outcomes of the model with
actual regional characteristics of the country under investigation. This ultimately leads to a refined list of
potential PSPs, or the realisable potential.
7.2. Goal
The goal of the expert selection process is to validate the feasibility of the automatically selected pairs, by
utilizing the experience of PSP experts. The end-result of the assessment is a further refinement of the list
of potential pairs, primarily by discarding those pairs which are not feasible and secondarily by ranking
those that are.
7.3. Data for expert selection process
7.3.1. Database of pairs
The database generated by the GIS model contains key characteristics of each pair found by the model
based on the selection criteria, i.e. each pair within the theoretical potential (see Figure 6 for an example of
the representation).
Figure 6: Excerpt from the database of selected pairs, used as input for the expert selection process
• Potential pairs are sorted and ranked by Energy Storage Capacity and by Length divided by Head
(inverse slope), with weights of 2 and 1, respectively. The function of this ranking is to provide the
expert with an indication of what theoretically would be the most interesting pairs.
• If known, the concession area of the respective pair is shown (indicating who has the right to build /
operate hydro and or PSP sites)
• Sensitive areas are displayed such as Natura 2000 and local protection areas; in case the
powerhouse, penstock and/or water body or bodies is/are located in a sensitive area (both national
as well as international), this is indicated in the table.
7.3.2. Online interactive map
For each country in the study area, an online map is created based on geo-referenced data from the GIS
model (see Figure 7). The online map has several functions. It presents an overview of the location of all the
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 37
pairs in that country. The user can also easily take location-based considerations into account when judging
potential pairs, such as proximity to infrastructure and the topology of systems of potential pairs, i.e. how
pairs are linked in a system where one or more water bodies are part of several selected pairs, see example
in Figure 8).
Figure 7: An online map of Italy, enabling the user to insert queries for specific site selection
Figure 8: Screenshot of an online map showing a single
PSP site.
Figure 9: Screenshot of an online map showing a system
of different interconnected sites
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 38
7.4. Selection of experts
As noted earlier in this report, the outcome is a data map will that provide industries and governments an
overview of suitable locations, enabling political and business decisions for development of new PSP
projects. The experts were selected based on the following competences for executing the expert selection
process:
• A track record of PSP-related activities in the country of investigation.
• Expertise not only on PSP in general, but preferably on technical aspects of PSP development
and/or exploitation.
• Preferably employed by a company that develops PSPs, in order to understand business
considerations for choosing a potential site.
The network of consortium partners in the eStorage consortium was actively involved in either direct
assistance in the expert selection process or leveraging their contacts in the sector.
7.5. Non-included parameters
The model and expert selection process were not able to take into account all parameters that could play a
role in determining the realisable PSP potential. Key parameters not included were for example:
• Power output of the PSP. It could be interesting for the expert to see what would be the expected
power output of the potential pair. However, no unambiguous value can be determined, because
this will depend on the application of the PSP and therefore with what charge/discharge durations
the PSP will be used. It has therefore been decided not to add this information to the table. The
topic will be covered by the expert selection process though
• Inflow of water in upper reservoir. In order to understand the usability of the upper reservoir, it is
interesting to know what the inflow characteristics are. However, this information is dynamic and
depends on complicated seasonal and climatic conditions, which would require a significant
amount of data and time to take into account. Moreover, it was considered out of scope for the
objectives of this study. Therefore it has been decided not to add this to the model
• Environmental considerations of water use. In order to understand the usability of the upper
reservoir, it is interesting to know what is the environmental impact and sensitivity when utilizing
the reservoirs. The impact is depending on regulatory changes and has not been added to the
model.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 39
8 RESULTS: THEORETICAL AND REALISABLE POTENTIAL
8.1. Introduction
In this chapter, the results of are presented, analysed and discussed. Both the theoretical potential and the
realisable potential (being the ultimate objective of this study) are evaluated for the study area and each of
its 17 countries; the relationship between these two types of potential are illustrated in Figure 8. It is
important to note that pairs that match existing PSPs are considered part of non-realisable potential simply
because they have already been realized; the realisable potential represents what can still be realized in the
future.
For Switzerland, no national expert could be identified in the time available to support the quantification of
the realisable PSP potential in that country. The realisable potential of Switzerland was therefore
determined by the project team based on experience with other countries and by application of only
technical exclusion criteria. It should therefore be regarded as a preliminary estimate, probably slightly
higher than what a Swiss national expert would determine due to identification of additional causes for
exclusion. The total realisable potential of the study area as a whole can be regarded as definitive because
possible additional exclusions from the Swiss realisable potential would have an insignificant effect on it.
Figure 8: Relationship between the three types of potential used in this study
8.2. Results
8.2.1. Aggregated results for study area
Table 7 and Figure 9 and Figure 10 below provide numerical and visual overviews of the main aggregated
results for the study area: number of pairs and energy storage capacities of theoretical, realisable and non-
realisable results.
In 11.4 Appendix 4: Country sheets, numbers and graphs are provided showing the results for each of the
individual countries.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 40
Parameter Unit
Au
stri
a
Be
lgiu
m
Fin
lan
d
Fra
nce
Ge
rma
ny
Gre
ece
Ita
ly
No
rwa
y
Po
rtu
ga
l
Sp
ain
Sw
ed
en
Sw
itze
rla
nd
*
Un
ite
d
Kin
gd
om
Stu
dy
are
a
Theoretical
potential
GWh
154
2
1
249
9
160
435
4377
278
721
217
457
85
6924
Non-
realisable
potential
GWh 145 2 0 132 2 0 281 3020 201 525 136 291* 39 4634
Realisable
potential
GWh 8 0 1 117 7 160 154 1356 77 196 81 166* 46 2291
Theoretical
potential
Pairs 25 1 1 35 2 1 64 381 10 95 62 37 19 714
Non-
realisable
potential
Pairs 22 1 0 19 1 0 43 349 8 78 57 20 11 597
Realisable
potential
Pairs 3 0 1 16 1 1 21 32 2 17 5 17 8 117
* Preliminary values (see 8.1)
Table 7 : Overview of theoretical, realisable and non-realisable PSP potential
within the study area by listing the number of pairs and capacity. Percentages apply to non-realisable respectively
realisable potential (pairs or capacity) as a faction of the theoretical potential of a country. Note that countries
where no pairs have been identified have been left out of the table: Denmark, Ireland, Luxembourg and the
Netherlands. Because cross-border pairs are counted for each country, the sum of the individual potentials (number
of pairs and capacity) per country is not equal to the aggregated number for the study area.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 41
*Preliminary values (see 8.1)
Figure 9: Map of the study area (non-white countries) showing the realisable potential energy storage capacity per
country in GWh.
Cross-border pairs are included in both totals of the corresponding countries, so the totals of each country do not
add up to the total of the study area.
A number of additional observations can be made from the table and charts Figure 10:
• The total realisable potential in the study area is 2291 GWh, which is only 33% of the theoretical
potential of 6924 GWh. This means that many pairs were excluded by experts even though their
key parameters (as taken into account by the model through the selection criteria) appeared
favourable, emphasising the importance and value of involving experts in the process rather than
relying on modelling only.
• The number of realisable pairs (117) is only 16% of the theoretical potential (714 pairs), which
combined with the aforementioned observation indicates that realisable pairs on average have
more than double the storage capacity of theoretical pairs. In part this is a direct consequence of
pair exclusion specifically due to small storage capacity (see 8.2.5), so not an independent effect.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 42
Figure 10: Theoretical and realisable potential for each country in the study area,
in GWh (top) and as percentages of the total theoretical potential (bottom left) respectively the total realisable
potential (bottom right). Denmark, Ireland, Luxembourg and The Netherlands are not included in these charts
because no pairs have been found for these countries
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 43
• For 4 countries, no pairs have been found: Denmark, Ireland, Luxembourg and the Netherlands
• For 4 further countries, the theoretical potential consists of only 1 or 2 pairs: Belgium, Finland,
Germany and Greece. From these countries, only Belgium has no realisable pairs. It is interesting to
note that Greece’s single pair represents 7% of Europe’s realisable potential on its own.
• By far the largest number of pairs and largest capacity belongs to one country: around 60% of both
the theoretical and the realisable potential (in GWh) is found in Norway
• The remaining 8 countries have a few percent share of the total potential each, with mostly
moderate differences.
8.2.2. Pair locations
Observations (see Figure 11):
• A number of regions have a relatively dense collection of sites. Three prominent examples, also
indicated in Figure 11 have about 70% of all theoretical pairs in the study area and 73% of the
realisable capacity:
o South Norway, with 43% of all theoretical pairs and 54% of the realisable capacity.
o The Alps, with 17% of all theoretical pairs and 13% of the realisable capacity.
o The Pyrenees, with 11% of all theoretical pairs and 5% of the realisable capacity.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 44
Figure 11: Theoretical and realisable potential of PSP sites in the study area: locations of all pairs.
The study area is indicated in light grey, other countries are shaded. Realisable pairs are indicated by orange dots,
non-realisable pairs by dark grey dots; not all different pairs may be distinguishable in the figure above because
they may be so close together that dots partially or fully overlap. Three high-potential areas are indicated with
green boxes, with text indicating realisable (orange) and non-realisable (dark grey) total numbers of pairs and their
capacity, respectively
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 45
• Other regions in Europe are almost completely devoid of sites, such as the east of Scandinavia plus
Finland, most of Germany, most of the UK except the North, most of Italy except the very North,
the east of Spain etc. This observation and the previous one are of course strongly correlated with
the presence or absence of significant elevation differences in the area.
A number of selected pairs are cross-border pairs, of which an overview table is given below.
Theoretical Realisable
Countries Pairs Total
capacity
(GWh)
Pairs Total
capacity
(GWh)
Austria / Germany 1 7 0 0
Spain / France 3 41 0 0
Spain / Portugal 5 110 1 29
Italy / Switzerland 7 77 0 0
Norway / Sweden 5 11 0 0
Total 21 246 1 29
Table 8 Overview of cross-border pairs
Although it represents only a few percent of the total capacity, there is some cross-border PSP potential in
the study area. Interestingly, only 1 out of the 21 cross-border pairs is considered realisable. Out of the 20
non-realisable pairs, in 10 cases the cause for exclusion was the very fact that they are located in two
countries with all the consequences associated with it. However, the other 10 pairs were excluded for other
reasons, showing together with the single realisable pair that cross-border pairs could be realisable in
principle and hence should not be excluded a priori.
8.2.3. Pair properties
The following Figure 12 and Figure 13 illustrate the key properties head, distance and energy storage
capacity of each pair as well as a statistical analysis of pair sizes for countries with significant potential.
Key observations:
• A visual check seems to indicate that the non-realisable and realisable potential both include
medium-sized and (very) large systems. The realisable potential seems to contain fewer (very) small
systems, which makes sense because an inferior size can be an exclusion criterion.
• For the pairs in the chart, i.e. those meeting the selection criteria, pairs with a pair distance below 7
km occur more frequently (and with a higher ESC), even though it would seem more logical to find
increasingly more pairs with increasing pair distance.
• Similarly it can be observed that pairs above roughly 800 m head are significantly less in number
and smaller, while pairs above roughly 1200 m head are both rare and very small, probably because
there are not many regions in the study area with such altitude differences within short distances.
• No clear trend is visible between energy storage capacity and head. Although energy storage
capacity is proportional with the head of a pair (see Equation 4) which could be expected to show,
this is probably approximately counteracted by the reduced upper pair volume of pairs with a
higher head.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 46
• Unsurprisingly, a triangular empty area is clearly visible in bottom right of the chart, corresponding
to the selection criterion of a 5% average slope between the two water bodies in a pair.
Figure 12: Overview of key parameters of all PSP sites in the theoretical potential.
Each bubble represents a theoretical pair; the yellow bubbles form the realisable potential while the grey bubbles
form the non-realisable potential. The vertical position of a bubble in the chart indicates the head of the pair and
the horizontal position indicates the distance between the water bodies. Bubble size indicates the energy storage
capacity of the pair, with a virtual 10 GWh pair (black bubble) located in the origin of the chart for scale
To show statistical information on energy storage capacity per country, a boxplot31
is shown in Figure 13.
Only countries with 2 or more realisable pairs were included in the plot. Observations are:
• It is immediately apparent that Norwegian realisable pairs are much larger in maximum, average
and spread (in GWh).
• The Portuguese realisable pairs have a very high capacity, although statistically relevant
observations can hardly be made based on only 2 pairs.
31
A boxplot is a graphical representation of the minimum, first quartile, median (or second quartile), third quartile and
maximum of data. The median and quartiles divide the data in four equally sized groups, so each of the groups
contains 25% of the data. In this way, a boxplot shows the distribution of data. * Preliminary values (see 8.1).
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 47
• While there are some differences between the other countries, their interquartile ranges (between
the first and third quartiles) are much more similar, between a few GWh and around 10-15 GWh.
Figure 13 Box plots of energy storage capacity in GWh for all realisable pairs in countries with 2 or more realisable
pairs.
Note that Norway was plotted in a separate diagram because its much higher values require a significantly different
scale for meaningful representation
8.2.4. Country-specific observations
In this section, the details and particularities of the theoretical and realisable potential (as judged by local
experts) for each country are briefly mentioned in Table 9.
In 11.4 Appendix 4: Country sheets, numbers and graphs are provided showing the results for each of the
individual countries.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 48
Country
Reviewing party Country-specific observations
Austria
Verbund / Vienna
University of
Technology
A large number of the pairs that have been found by the GIS model were already under
consideration by the experts from Verbund and from Vienna University of Technology.
Belgium
N/A
Only one pair is found by the model with a capacity of 1.6 GWh, which is identified as an
existing PSP. Not more pairs were found, nevertheless, the existing capacity of COO (COO I
& COO II) is 5 GWh and should have been found. Also, another project “COO III” is
considered to be built, a 3 GWh PSP unit.
Denmark
N/A
No theoretical potential was found, primarily due to insufficient altitude differences
between water bodies in Denmark.
Finland
EDF
No country-specific expert has been requested for validation, since only one pair has been
found. The PSP expert at EDF has been requested to provide his expert judgements on
Finland as well. One pair has been found, which has a size of 1.3 GWh.
France
EDF
The PSP expert for France is employed at EDF, which is consortium member of the
eStorage project, and Steering Group member for WP4. For France, a relatively high
amount of theoretical pairs was considered realisable as well (47%). A possible reason for
this is the partaking of this in the steering group of WP4, being the co-developer of the
primary selection criteria for the GIS model. Therefore, the outcomes of the model are
better aligned with the preferences of the expert for this country.
Germany
E.ON
Two potential pairs were found in Germany, one of which is a cross-border pair with
Austria. One of the pairs is excluded due to difference in water quality between the water
bodies, i.e. the water of the water bodies are not allowed to be mixed.
Greece
EDF
No country-specific expert has been requested for validation, since only one pair has been
found. The PSP expert at EDF has been requested to provide his expert judgements on
Greece as well. Only one potential pair in Greece according to the model. Strikingly, this
pair has an ESC of 160 GWh.
Ireland
N/A
No theoretical potential was found, primarily due to insufficient altitude differences
between water bodies in Ireland.
Italy
ENEL
The exclusion criterion Different river systems was considered as important by the Italian
expert from Enel. In Italy, no shifting of water is allowed between two river systems. It is
therefore one of the few countries where an expert has indicated that ‘sensitive areas’
was a factor of exclusion.
Luxembourg
N/A
No theoretical potential was found, primarily due to insufficient altitude differences
between water bodies in Luxembourg.
Netherlands
N/A/
No theoretical potential was found, primarily due to insufficient altitude differences
between water bodies in The Netherlands.
Norway
Norsk Hydro
Norway has the greatest potential ESC. Due to the large amount of pairs, a first step in the
expert selection process was to exclude all pairs with a capacity below 2.5 GWh. There are
243 pairs with such a low capacity, leaving 139 pairs for extensive individual review. Many
of the river systems in Norway are protected from further hydropower development by
national law. Thus, as a first step in the expert selection process, all pairs involving
protected river systems were excluded. Most of the remaining pairs are located close to
large existing hydropower plants in the South. Many of the highest ranked pairs identified
in the model were well known by national hydropower experts.
Portugal
Gas Natural Fenosa
No country-specific expert has been requested for validation, since only a small number of
pairs have been found and the expert for Spain is well aware of the specific country
conditions for Portugal. The PSP expert at Gas Natural Fenosa has been requested to
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 49
provide his expert judgements on Portugal as well.
Spain
Gas Natural Fenosa
Although the primary selection criteria as used for the GIS model did not take into account
the volume of the reservoirs explicitly, the PSP expert from Gas Natural Fenosa indicated
that a number of the reservoirs as part of the potential sites were too small to take into
account. One reason for this exclusion criterion is that the number of theoretical pairs in
Spain is relatively big (95 pairs).
Sweden
Vattenfall
Many of the river systems in Sweden are protected from further hydropower
development by national law. As a first step of the expert selection process, all pairs in
these rivers (Pite, Kalix, Torne and Vindel) were excluded. Most potential locations were
found in the North of Sweden, in rivers already heavily developed with hydropower, as
elevations differences in the South are smaller.
Switzerland
Technical exclusion
criteria only
For Switzerland, no national expert could be involved in the determination of its realisable
potential. The realisable potential of Switzerland was determined by the project team
based on experience with other countries and by application of only technical exclusion
criteria. It should therefore be regarded as a preliminary estimate, probably slightly higher
than what a Swiss national expert would determine due to identification of additional
exclusion causes.
United Kingdom
Quarry Battery
For the United Kingdom, several pairs were excluded because of the distance to the grid
being too high, specifically for potential pairs in the northern part of the UK. Also, a
number of pairs in sensitive areas were excluded, and because of water usage (i.e.
drinking water).
Table 9 Overview of country-specific observations
8.2.5. Exclusion causes and observations
In cooperation with the national experts involved in the expert selection process, a limited set of exclusion
causes was identified, i.e. a set of reasons why certain pairs in the theoretical potential are excluded and
considered non-realisable, the remaining pairs forming the realisable potential. The following Table 10 and
* Preliminary values (see 8.1)
Table 11 Table 11 highlights these exclusion causes and how often and where they apply.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 50
General category
Specific category Explanation
Technical Is the site technically interesting?
Criteria not met No pair: none or only one appropriate water body, duplicate with other
pair or water body.
Water management High inflow compared to storage capacity (upper reservoir), low volume
and seasonally low inflow (lower reservoir), snow/ice or evaporation
issues.
Other Other technical issues, e.g. slope, topology, geology, etc.
Environmental Are there any environmental or water use restrictions?
Environmental
restrictions
Protected areas, two different river systems, sensitive area or other
environmental issues.
Water use Water body has other purpose, e.g. drinking water, irrigation, etc.
Competing hydro / PSP Are any of the water bodies already in use by competing PSP or hydro
power plants?
Existing PSP Pair is an already existing PSP.
Competing PSP Water body already in use by existing PSP and does not have sufficient
capacity left.
Competing
hydropower
Water body already in use by existing hydro power plant and does not
have sufficient capacity left.
Ownership / border Are there any ownership or border issues?
Different owners The ownership rights of the two pairs are in different hands
Cross-border One pair is in another country than the other pair, which in this case
leads to issues causing exclusion
Relative interest Is the site interesting given other opportunities?
Competing pair more
promising
A competing found pair is more promising and the shared water body
does not have sufficient capacity to provide both pairs.
Small storage capacity In comparison with other pairs the pair is not interesting due to its small
storage capacity.
Table 10 Exclusion causes for non-realisable pairs and an explanation of their meaning
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 51
Exclusion cause
Au
stri
a
Be
lgiu
m
Fin
lan
d
Fra
nce
Ge
rma
ny
Gre
ece
Ita
ly
No
rwa
y
Po
rtu
ga
l
Sp
ain
Sw
ed
en
Sw
itze
rla
nd
*
Un
ite
d K
ing
do
m
Stu
dy
are
a
Existing PSP 9 1 - 5 - - 10 2 5 5 - 9 4 49
Competing PSP 3 - - - - - 1 - - 3 - 6 - 13
Environmental
restrictions
- - - - - - 7 128 - 2 29 - 1 166
Water use - - - - - - - - 1 10 - - 2 12
Cross-border difficulties 1 - - 4 - - 6 - 1 5 - 2 - 13
Competing pair more
promising
4 - - 10 - - 11 22 - 9 - 3 2 61
Small storage capacity 1 - - - - - 1 161 - 32 17 - - 214
Criteria not met 4 - - - - - 4 36 1 11 11 - - 62
Other technical reasons - - - - 1 - 1 - - 1 - - 1 4
Miscellaneous reasons - - - - - - 2 - - - - - 1 3
Total 22 1 - 19 1 - 43 349 8 78 57 20 11 597
* Preliminary values (see 8.1)
Table 11 Quantification of exclusion causes for non-realisable pairs in each country and the total study area.
Key observations:
• In Norway, the country with by far the largest non-realisable potential (excluded pairs), the most
common exclusion cause is “small storage capacity” (161 cases), probably due to the sheer number
of better options available. Note that many pairs excluded due to small storage capacity have sizes
that would not lead to exclusion in other countries. This is a consequence of the expert selection
process; the sum of realisable potentials as determined by national experts may therefore not be
the realisable potential of the study area as it would be judged by European experts not bound to
national considerations.
• In Norway, the second prominent exclusion cause is environmental restrictions (128 cases).
• In the other countries together, the three most prominent exclusion causes are:
o Small storage capacity (53 cases)
o Existing PSP (47 cases)
o Competing pair more promising (39 cases).
• Exclusion causes differ strongly between countries, probably due to differences in national
circumstances (type/size/number of nature reserves, regulations, etc.) unavoidable subjectivity of
experts and of course differences in the properties of the theoretical pairs of each country.
• Although potentially relevant, the following exclusion causes were ultimately not judged to apply:
water management, different owners and competing hydropower.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 52
8.3. Sensitivity analysis
Sensitivity analysis aims to quantitatively chart the effect of input parameters on model output. It enables
understanding how changes in the selection criteria for potential sites influence model results, and which
changes in parameters have a large or small effect. For the sensitivity analysis, France has been taken as
target country.32
Five scenarios were studied with key selection criteria changed between them to
investigate the impact of these changes on the model results (see Table 12 below).
Scenario Parameters Number of results
1 (standard parameters) Head >80 m 35
ESC > 1 GWh
Distance < 10 km
2 Head > 150 m 37
ESC > 2 GWh
Distance < 20 km
3 Head > 100 m 48
ESC > 1 GWh
Distance < 16 km
4 Head > 80 m 40
ESC > 0.8 GWh
Distance < 10 km
5 Head > 40 m 92
ESC > 0.5 GWh
Distance < 20 km
Table 12 Sensitivity analysis parameters
Sensitivity analysis 2 and 3 focus on the impact of the distance between the water bodies and the energy
storage capacity. This distance seemed to be important for the amount of suitable results. By enlarging the
area by 10 kilometers, a number of 19 additional pairs have been found. At the same time, 17 pairs were
removed due to a doubled energy storage capacity minimum of 2 GWh. The head plays a minor role in
these changes.
In sensitivity analysis 4 the energy storage capacity was changed from 1 GWh to 0.8 GWh. This change led
to five additional pairs.
In sensitivity analysis 5 all criteria were broadened with a factor of 2. The energy storage capacity is most
sensitive to scaling; lowering this parameter led to finding 28 extra pairs by itself. Enlarging the maximum
distance led to 19 extra pairs and 2 extra water pairs were selected due to decreasing the minimum head.
An additional 8 extra pairs had both a higher distance and a lower energy storage capacity.
32
Like for the model validation, the choice of France was mostly because of extensive expertise on hydropower in
France in the Steering Group. Time permitting, it would be interesting to perform this analysis for all countries in the
study area.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 53
In conclusion, we can say that especially the energy storage capacity and the distance between the water
bodies are sensitive selection criteria. Changing these parameters leads to relatively large differences in
theoretical potential.
The expanded parameters could be used when the standard parameters show insufficient results for a
country. Also, the sensitivity analysis can be different for other countries, because of the high dependency
of these parameters on local factors such as geology and elevation.
8.4. Comparison with JRC study
As mentioned before, the JRC has executed a GIS-based energy storage study1 with similarities to this study.
It is interesting and relevant to compare the results of the two studies to the extent possible. To this end,
the theoretical potential as identified by JRC (their so-called “topology T1”) was compared with the
theoretical potential identified in this eStorage study. Filtering was applied to make results more
comparable (see Table 13): JRC results with pair distances above 10 km were excluded, while eStorage
results with a head of under 150 m were also excluded. Only countries in the smaller eStorage study area
were compared. It has to be noted that differences still remain, such as:
• Environmental restrictions could cause pairs to be excluded from the JRC theoretical potential
unlike for the eStorage study.
• A minimum reservoir volume of 100,000 m3 was only used in the JRC study.
• A minimum energy storage capacity of 1 GWh was only used in the eStorage study.
• A minimum slope of 5% was only used in the eStorage study.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 54
Parameter →→→→
Number
of sites
Average
head
Average
capacity
Total
capacity
Unit →→→→ pairs m GWh/pair GWh
Austria
eStorage 24 815 6 152
JRC 40 557 5 199
Belgium
eStorage 1 249 2 2
JRC 2 252 3 5
Finland
eStorage 1 209 1 1
JRC - - - -
France
eStorage 33 558 7 243
JRC 91 384 4 409
Germany
eStorage 2 497 4 9
JRC 9 190 2 19
Greece
eStorage - - - -
JRC 1 151 28 28
Italy
eStorage 63 607 7 432
JRC 179 470 4 661
Norway
eStorage 373 524 12 4,335
JRC 29 301 11 332
Portugal
eStorage 7 336 37 258
JRC 15 260 8 118
Spain
eStorage 86 617 8 650
JRC 652 371 2 1182
Sweden
eStorage 50 338 3 154
JRC 1 222 22 22
Switzerland
eStorage 35 572 15 445
JRC 66 502 7 431
United
Kingdom
eStorage 16 305 5 80
JRC 561 189 - 199
TOTAL eStorage 691 - 10 6,761
JRC 1,646 - 2 3,605
Table 13 Comparison of JRC and eStorage study results: theoretical potentials for identical head and distance
criteria
Observations:
• For most countries, JRC has found significantly more pairs, however with a total energy storage
capacity that is relatively only slightly higher. In other words, the additional pairs JRC identified
necessarily have a much smaller energy storage capacity, which can also be derived from
comparing average capacities.
• For Norway and Sweden, eStorage has identified a significantly higher potential (13 and 7 times
higher, respectively) than JRC, probably due to better water body data for Norway and possibly for
Sweden as well.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 55
• If it is assumed that, for Austria, France and Italy where JRC has found more results, the pairs
identified by eStorage are a subset of the JRC results, the average storage capacity of those
additional pairs is between 2 and 3 GWh, which is much lower than the pairs found by eStorage.
However, consequently in those countries some pairs with a capacity above 1 GWh must exist but
were not found by eStorage. Possible reasons could be flaws in the eStorage GIS-dataset and/or the
fact that these pairs have a slope below 5% (eStorage selection criterion).
• For Greece, the one pair eStorage has found is comparable to the one JRC pair but has a head
slightly lower than 150 m according to the eStorage model, so it was not included in the
comparison table. It is likely the results can be unified by updating the elevation data and
recalculating heads.
• For Finland, eStorage has found one pair but JRC has not, which might be due to environmental
restrictions.
• For Portugal, eStorage has found fewer pairs, but with a significantly higher total capacity.
• For the United Kingdom, the number of additional pairs JRC found compared to eStorage is much
higher than for other countries. A probable explanation is the very low average capacity of these
additional pairs of around 0.2 GWh which is below the eStorage minimum of 1 GWh.
These diverse and interesting observations suggest that a comparison of JRC and eStorage results in a more
detailed manner (i.e. not using only aggregated data but their full numerical and GIS datasets) would be
useful if not essential, to look into the differences above and ultimately to be able to improve upon both
study results.
8.5. Discussion
A large share of pairs selected by the GIS model have been excluded by the expert selection process,
indicating the importance of including country-specific expert opinions into the study
About two thirds of the theoretical potential is excluded by means of the expert selection process,
indicating that realisable potential can’t be drawn from the model, but expert opinions are required. The
total realisable potential in the study area was 2291 GWh, which is only 33% of the theoretical potential of
6924 GWh. Therefore, many pairs were excluded by experts even though their key parameters (as taken
into account by the model through the selection criteria) appeared favourable, emphasising the importance
and value of involving experts in the process and not only modelling.
The model identified many promising sites and the ranking seemed to provide a good indication on
which sites are the most promising, with the expert selection process providing additional essential
insights
Many of the highest ranked pairs were well known by the local experts. The suggested ranking based on
storage capacity (GWh) and slope provided a good indication on which pairs where most promising and
could often also be used to identify which pair was most promising in systems with several competing pairs.
However, several experts highlighted water management as reasons to alter ranking or even rule locations
out. Some locations were regarded as particularly beneficiary because of contribution to flood control
and/or better use of existing hydropower systems. On the other hand pairs involving pumping to reservoirs
with high inflow compared to storage capacity or pairs with low (seasonal) inflow to lowest water body
were ruled out or considered less promising.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 56
Most of the high-ranked pairs are located in already developed hydropower systems
Several of the highest ranked pairs often corresponded to aired existing PSPs or HPPs or the pairs are
located in vicinity to existing plants. This is natural since the criteria used to identify potential PSPs sites
also are highly relevant for many HPP sites. Many of the excluded and low ranked pairs are in areas which
are not yet developed for hydropower purposes because of low storage potential, environmental
restrictions, road and grid accessibility etc.
For a number of countries, many pairs are ruled out because of environmental restrictions
The model does not automatically take environmental restrictions into considerations; thus many pairs are
ruled out on based on environmental issues identified in the expert selection process. In addition to Natura
2000, national regulations (e.g. protected rivers) and local environmental conditions (e.g. minimum flow
requirements) were considered in the expert selection process. As a general rule, pairs involving 2 different
river systems were ruled out due to environmental considerations (in many cases these pairs also include
different ownership). However, there might be cases where a PSP involving 2 different river systems could
be feasible, but according to expert opinion environmental issues will generally make this difficult and the
implantation of the EU Water Framework Directive will make it even less likely that such pairs will be
developed.
Many pairs are excluded for several reasons, for instance a combination of environmental concerns, small
storage capacity and inadequate fit with existing hydro power systems.
Even after the expert selection process, there may still be environmental and technical issues which are
not yet identified and GWh potential needs to be investigated further
Although a large number of pairs were ruled out in the expert selection process, there might also be
unidentified environmental and technical restrictions for the remaining pairs. This is especially true for
locations in areas which the national experts do not know that well.
The GWh potential as well as placement of the penstock and power house are roughly estimated, and more
detailed business case studies are required in order to determine this and other design parameters for the
identified PSP sites. This is outside the scope of this project.
The model could be improved to take into account more accurate data
For some countries there exist detailed publicly available data on reservoir volumes, environmental
restrictions, existing PSPs and HPPs, inflow etc. which could be used to improve the model. The availability
and quality of the information varies significantly from country to country, and to make the model as
consistent as possible it was decided not to include country specific data in the first step. Such
considerations were rather taken into account in the expert selection process.
Countries not included in the model are also interesting in the European perspective
The study cover EU-15 + Norway and Switzerland, thus countries in Eastern Europe (including Turkey) are
not included. When assessing the potential to provide flexibility in the European system, many of these
countries are interesting.
Discussion and pair ranking / exclusion input from national experts was highly useful, even though
inevitably introducing some subjectivity
The expert selection process is inherently subjective, i.e. involvement of different experts in filtering out
the same theoretical potential might not lead to the same realisable potential. Exclusion causes differ
strongly between countries, probably due to differences in national circumstances (type/size/number of
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 57
nature reserves, regulations, etc.) unavoidable subjectivity of experts themselves and of course differences
in the properties of the theoretical pairs of each country. Still, incorporation of all exclusion criteria into
one model for an unbiased uniform selection process would not have been an attainable goal within the
scope, time and budget of the eStorage project or at all. Such an approach would be extremely time-
intensive and would require costly and rare higher-quality GIS source data of more types (geology,
infrastructure, available network capacity, etc.).
Due to the number of sites, the results for Norway greatly impact the average numbers for the total
study area
The number of realisable pairs (117) is only 16% of the theoretical potential (714 pairs), which combined
with the aforementioned observation indicates that realisable pairs on average have more than double the
storage capacity of theoretical pairs. In Norway, the country with by far the largest non-realisable potential
(excluded pairs), the most common exclusion cause is “small storage capacity” (161 cases), probably due to
the sheer number of better options available. Note that many pairs excluded due to small storage capacity
have sizes that would not lead to exclusion in other countries. This is a consequence of the selection
process with national experts; the sum of realisable potentials as determined by national experts may
therefore not be the potential of the study area as judged by European experts not bound to national
considerations.
About 70% of realisable potential is located in areas that are not ideally located with respect to load and
generation centers in the study area
In the Alps and Pyrenees there are already many existing PSPs and the location is often difficult to reach; in
addition, 43% is located in Norway, where most of the need for energy storage is in mainland Europe. Also,
a number of countries have an existing or future need for flexibility while no potential sites are available
within reasonable distances.
Potential of cross-border pairs
Although only a few percent of the total capacity, there is some cross-border potential in the study area.
Interestingly, only 1 out of the 21 cross-border pairs is considered realisable. Out of the 20 non-realisable
pair, in 10 cases the cause for exclusion was the very fact that they are located in two countries and all
consequences associated with it. However, the other 10 pairs were excluded for other reasons, showing
together with the 1 realisable pair that cross-border pairs could be realisable in principle and hence should
not be excluded a priori.
Matching of the model results with existing PSP
Many but not all existing PSPs have been found by the GIS model. The two main reasons for their absence
from the theoretical potential are non-matching properties and data quality. Some PSPs have properties
not matching the selection criteria. Sufficient relaxation of the selection criteria would lead to their
inclusion in the theoretical potential. However, the resulting additional theoretical potential is unlikely to
lead to a larger realisable potential because it will have less favourable properties than current results and
is likely to be excluded by experts. Data quality is a more problematic cause, because non-identification of
existing PSPs implies nonexistent pairs are probably also left unidentified. Volumes of water bodies may be
incorrect, single water bodies may be fragmented (both potentially leading to incorrect exclusion due to
insufficient energy storage capacity), water bodies may be missing etc. This could be a serious problem with
an impact (a lower realisable potential) that is difficult to quantify.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 58
9 CONCLUSIONS
In this study, the potential for (new) PSP sites in 15 EU countries plus Norway and Switzerland has been
investigated. This study has been carried out with the aid of a GIS model (resulting in theoretical potential)
and subsequently a thorough selection with the support of national industry experts (resulting in realisable
potential). The expert selection has been carried out using different exclusion categories. An exclusion
criterion is either technical, for example if a pair is already in use by another PSP, or regulatory, for example
due to environmental considerations. In the first category, it is considered technically unfeasible to develop
a new PSP; in the latter, it is ultimately a choice by all stakeholders not to develop the pair as a PSP.
A total of 714 sites with the total potential of 6924 GWh have been found by the GIS model, of which 597
pairs with a total energy storage capacity (ESC) of 4634 GWh have been excluded in the expert selection
process. This makes the total realisable potential 117 pairs with a total ESC of 2291 GWh. In other words,
16% of the theoretical potential was considered to be realisable in terms of the number of pairs, while 33%
was considered to be realisable in terms of ESC (GWh).
Four countries were not found to have any theoretical potential: Denmark, Ireland, Luxembourg and the
Netherlands. The reason is that the primary selection criteria from the GIS model were not met. The most
important eliminating criteria were the slope (> 5%) in combination with the distance between water
bodies (< 10km). These criteria also resulted in Finland, Germany, Belgium and Greece having very little
potential, i.e. only one or two potential sites for each of these countries resulting from the model.
Norwegian sites comprise about 60% of both theoretical as well as realisable potential in terms of ESC (in
GWh). Furthermore, three regions have been identified as having a particularly high density of potential
PSP sites: the south of Norway, the Alps and the Pyrenees. These high-density areas include about 70% of
all theoretical pairs in the study area and 73% of the realisable capacity. Countries in these regions have the
most opportunities for PSP development.
The main reasons of the experts to exclude certain sites were: small storage capacity (214 pairs),
environmental restrictions (166 pairs), criteria not met – i.e. no pair: none or only one appropriate water
body or duplicate with other pair or water body (62 pairs) and competing pair more promising (61 pairs).
This latter category means that in a system of pairs, not all pairs are possible due to a competition with
another potential pair (and the water in a water body can only be used once). In this case, only the pair
with the highest rank is selected. Note that this last exclusion criterion has been used only after all the
other exclusion criteria have been applied, meaning that e.g. environmental restrictions may have already
excluded some pairs (one pair can be excluded by more than one exclusion criterion).
For this study, the approach has been adopted to ask for the judgment of national PSP experts to define the
realisable potential (as opposed to relying only on the outcome of a model). This choice has had the
following implications:
• The primary selection criteria as used in the GIS model were not always in line with the preferences
of the experts. The volume of the water bodies was sometimes deemed insufficient, while the
selection criterion in the GIS model was to have ESC of at least 1 GWh, regardless of the volume.
• Interpretation of the exclusion criteria can be different for each expert (and possibly for each
country of examination) so the final results of this study have an element of subjectivity.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 59
• Through application of the expert views on the theoretical potential, the results provide a more
realistic perspective on realisable potential than existing studies based on algorithms.
• The (subjective) judgments of the experts led to different numbers of exclusions (i.e. one expert
may be more rigorous than the other). For example, comparing the exclusions of all countries with
a theoretical potential of more than two pairs, the share of excluded pairs varied between 46%
(United Kingdom) and 95% (Austria).
• For Norway, the expert selection has been executed by starting with the elimination of all pairs
lower than 2.5 GWh due to the large amount of theoretical potential. This resulted in the exclusion
of 161 pairs in Norway.
• The exclusion categories were formulated before the expert selection process and adjusted during
the process. It turned out that three of the initially formulated exclusion categories were never
chosen as reasons for exclusion by the experts, namely: water management, different owners and
competing hydropower.
• A comparison of the results has been made with a related study carried out by the JRC1 given its
similar objectives and a similar approach that included a GIS-based model. Subsets of the results of
both studies have been created by filtering using equalized selection criteria as much as possible.
Comparison of these subsets of results led to the following main observations:
− The JRC has found more pairs for most countries however the additional pairs found by JRC
have a much smaller ESC. On the other hand, eStorage has found a significantly higher
number of pairs for Norway and Sweden. This is presumably due to better quality of water
body data.
− Natura 2000 or other environmental considerations have a lower impact on exclusion in
this study than anticipated by the JRC Study, at least for the experts who have been
consulted in this study.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 60
10 OUTLOOK
Taking into account both the merits and the limitations of the study results as well as feedback from the
international hydropower community,2 the following recommendations for future work are provided here:
• The study area should be extended to include other parts of Europe for a more complete
assessment of the PSP potential in Europe. The scope of this study was EU-15 plus Norway and
Switzerland, while the current and future European grid situation does not have this artificial
limitation, but will rather depend strongly on interactions between different regions, some of
which are outside the study area.
• The identified PSP potential should be compared with the need for energy storage in power
generation and consumption predictions for the next decades, to put the study results in
perspective and determine the role of PSP relative to other flexibility solutions. An additional
insight from the study would be obtained by considering in more detail the locations of the PSP
potential and comparing them against the predicted needs for flexibility from energy storage,
effectively adding another level of realisability to the study results. PSP sites closer to areas in need
of energy storage will be more interesting, those further away less interesting or might even be
considered non-realisable. The final results of the eStorage WP3 scenario studies should provide
some of this information, and conversely, they should make sure to integrate these T4.1.2 results in
their analyses. Lastly, specific and detailed grid connection information should ideally be taken into
account such as available current and future capacity of the line to be used, the feasibility of a new
connection, etc.
• Because several indications on data imperfections have been found to impact the study results,
improved source data quality would be highly recommendable. Through using updated and/or
additional (GIS) datasets of especially water bodies (including water height and volume), dams and
elevation, fewer false positives and more importantly fewer false negatives would be achievable. In
line with this reasoning, a detailed comparison of source data and results of study with the JRC
study1 results would be helpful to identify possible improvements for both studies.
• Apart from improvements in source data or expanding the geographical scope as described above,
the GIS-model developed in this study would benefit from greater stability and speed to allow the
investigation of multiple scenarios (e.g. as required for sensitivity analysis) and enable a more
thorough and broader analysis of results within a shorter period of time.
• It could be considered to include potential new reservoirs (i.e. suitable topology like a closeable
valley) in the PSP potential determination, like in the JRC study. An industry Steering Group should
be consulted to determine suitable criteria for such new reservoirs. Also, again an expert selection
process is recommended for determining the realisable potential from the theoretical potential
(model outcomes) of such PSP sites.
• It could be considered to apply different selection criteria for each country in order to
accommodate the minimum or maximum need for (additional) PSP capacity for that country. For
instance, the expert for Germany was interested in relaxing the criteria, due to low amount of
potential pairs with the general selection criteria. On the other hand, the expert for Norway was
interested in making the criterion for ESC more strict (>2GWh) in order to reduce the amount of
potential pairs.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 61
11 APPENDICES
11.1. Appendix 1: IUCN category system for sensitive sites
Category Ia: Strict nature reserve
Category Ia are strictly protected areas set aside to protect biodiversity and also possibly
geological/geomorphological features, where human visitation, use and impacts are strictly controlled and
limited to ensure protection of the conservation values. Such protected areas can serve as indispensable
reference areas for scientific research and monitoring.
Category Ib: Wilderness area
Category Ib protected areas are usually large unmodified or slightly modified areas, retaining their natural
character and influence, without permanent or significant human habitation, which are protected and
managed so as to preserve their natural condition.
Category II: National park
Category II protected areas are large natural or near natural areas set aside to protect large-scale ecological
processes, along with the complement of species and ecosystems characteristic of the area, which also
provide a foundation for environmentally and culturally compatible spiritual, scientific, educational,
recreational and visitor opportunities.
Category III: Natural monument or feature
Category III protected areas are set aside to protect a specific natural monument, which can be a landform,
sea mount, submarine cavern, geological feature such as a cave or even a living feature such as an ancient
grove. They are generally quite small protected areas and often have high visitor value.
Category IV: Habitat/species management area
Category IV protected areas aim to protect particular species or habitats and management reflects this
priority. Many category IV protected areas will need regular, active interventions to address the
requirements of particular species or to maintain habitats, but this is not a requirement of the category.
Category V: Protected landscape/seascape
A protected area where the interaction of people and nature over time has produced an area of distinct
character with significant ecological, biological, cultural and scenic value: and where safeguarding the
integrity of this interaction is vital to protecting and sustaining the area and its associated nature
conservation and other values.
Category VI: Protected area with sustainable use of natural resources
Category VI protected areas conserve ecosystems and habitats, together with associated cultural values
and traditional natural resource management systems. They are generally large, with most of the area in a
natural condition, where a proportion is under sustainable natural resource management and where low-
level non-industrial use of natural resources compatible with nature conservation is seen as one of the
main aims of the area.
Source: Dudley, N. (Editor) (2008). Guidelines for Applying Protected Area Management Categories. Gland,
Switzerland: IUCN.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 62
11.2. Appendix 2: Common global conventions for sensitive sites
United Nations Educational, Scientific and Cultural Organization (UNESCO) World Heritage Sites:
UNESCO seeks to encourage the identification, protection and preservation of cultural and natural heritage
around the world considered to be of outstanding value to humanity. This is embodied in an international
treaty called the Convention concerning the Protection of the World Cultural and Natural Heritage, adopted
by UNESCO in 1972.
United Nations Educational, Scientific and Cultural Organization (UNESCO) Man and the Biosphere
Programme (MAB):
The Man and the Biosphere Programme (MAB), proposes an interdisciplinary research agenda and capacity
building aiming to improve the relationship of people with their environment globally. Since its launch in
1970 MAB has concentrated on the development of the World Network of Biosphere Reserves (WNBR).
The biosphere reserve concept was developed initially in 1974 and was substantially revised in 1995 with
the adoption by the UNESCO General Conference of the Statutory Framework and the Seville Strategy for
Biosphere Reserves.
The Convention on Wetlands of International Importance especially as Waterfowl Habitat (Ramsar
Convention):
The Convention on Wetlands is an intergovernmental treaty adopted on 2 February 1971 in the Iranian city
of Ramsar, on the southern shore of the Caspian Sea. The Convention entered into force in 1975 and now
(as of August 2007) has 155 Contracting Parties, or member States, in all parts of the world. The mission of
the Ramsar Convention, as adopted by the Parties in 1999 and refined in 2002, is “the conservation and
wise use of all wetlands through local, regional and national actions and international cooperation, as a
contribution towards achieving sustainable development throughout the world”.
Source: http://www.wdpa.org/FAQ.aspx#ctl00_MainContent_Faq3
11.3. Appendix 3: JRC workshop recommendations
1 To reaffirm the recommendations of the November 1999 Workshop on Spatial Reference Systems,
i.e. to adopt ETRS89 as geodetic datum and to express and store positions, as far as possible, in
ellipsoidal coordinates, with the underlying GRS80 ellipsoid [ETRS89]. To further adopt EVRF2000
for expressing practical heights (gravity-related).
2 Recognising that the EC needs cannot be met through usage of the ETRS89 ellipsoidal coordinate
reference system [ETRS89] alone, and map projections are required to supplement the ellipsoidal
system:
• To adopt ETRS89 Lambert Azimuthal Equal Area coordinate reference system of 2001 [ETRS-
LAEA], for statistical analysis and display
• To adopt ETRS89 Lambert Conic Conformal coordinate reference sys-tem of 2001 [ETRS-LCC]
for conformal pan-European mapping at scales smaller or equal to 1:500,000
• To adopt ETRS89 Transverse Mercator coordinate reference systems [ETRS-TMzn], for
conformal pan-European mapping at scales larger than 1:500,000.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 63
11.4. Appendix 4: Country sheets
11.4.1. General information
The following pages contain an overview of results for each of the countries in the study area, as well as for
the whole study area. Denmark, Ireland, Luxembourg and the Netherlands are not included because these
countries have no potential according to the present model. Data on realisable potential for Switzerland is
preliminary (see 8.1).
The results for each country and for the study area are represented in a data sheet with a standardized
format to facilitate comparison. Each data sheet contains the following elements pertaining to its area:
• Table with potentials
o Theoretical, non-realisable and realisable potentials (GWh)
o Number of pairs for each potential
o Percentages of realisable and non-realisable versus theoretical potential
• Bubble chart showing all identified pairs (realisable and non-realisable) and their key properties
distance, head and energy storage capacity. Each bubble represents a theoretical pair; the yellow
bubbles form the realisable potential while the grey bubbles form the non-realisable potential. In
each chart, a virtual 10 GWh pair (black bubble) located in the origin of the chart for scale
• Bar chart showing the occurrence of reasons for exclusion of pairs (number of times)
• Box plot providing statistical information on the identified pairs. No box plot is provided for
Belgium, Finland, Germany and Greece because in these countries the number of pairs is 1 or 2
pairs, so a box plot would not be statistically valuable.
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 64
11.4.2. Study area: EU-15 + Norway + Switzerland
Unit Value
Theoretical
potential
GWh 6925
pairs 714
Non-realisable
potential
GWh 4634
% of GWh 67%
pairs 597
% of pairs 84%
Realisable potential
GWh 2291
% of GWh 33%
pairs 117
% of pairs 16%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 65
11.4.3. Austria
Unit Value
Theoretical
potential
GWh 154
pairs 25
Non-realisable
potential
GWh 145
% of GWh 95%
pairs 22
% of pairs 88%
Realisable potential
GWh 8
% of GWh 5%
pairs 3
% of pairs 12%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 66
11.4.4. Belgium
Unit Value
Theoretical
potential
GWh 2
pairs 1
Non-realisable
potential
GWh 2
% of GWh 100%
pairs 1
% of pairs 100%
Realisable potential
GWh 0
% of GWh 0%
pairs 0
% of pairs 0%
Boxplot N/A
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 67
11.4.5. Finland
Unit Value
Theoretical
potential
GWh 1
pairs 1
Non-realisable
potential
GWh 0
% of GWh 0%
pairs 0
% of pairs 0%
Realisable potential
GWh 1
% of GWh 100%
pairs 1
% of pairs 100%
Boxplot N/A
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 68
11.4.6. France
Unit Value
Theoretical
potential
GWh 249
pairs 35
Non-realisable
potential
GWh 132
% of GWh 53%
pairs 19
% of pairs 54%
Realisable potential
GWh 117
% of GWh 47%
pairs 16
% of pairs 46%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 69
11.4.7. Germany
Unit Value
Theoretical
potential
GWh 9
pairs 2
Non-realisable
potential
GWh 2
% of GWh 27%
pairs 1
% of pairs 50%
Realisable potential
GWh 7
% of GWh 73%
pairs 1
% of pairs 50%
Boxplot N/A
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 70
11.4.8. Greece
Unit Value
Theoretical
potential
GWh 160
pairs 1
Non-realisable
potential
GWh 0
% of GWh 0%
pairs 0
% of pairs 0%
Realisable potential
GWh 160
% of GWh 100%
pairs 1
% of pairs 100%
Boxplot N/A
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 71
11.4.9. Italy
Unit Value
Theoretical
potential
GWh 435
pairs 64
Non-realisable
potential
GWh 281
% of GWh 65%
pairs 43
% of pairs 67%
Realisable potential
GWh 154
% of GWh 35%
pairs 21
% of pairs 33%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 72
11.4.10.Norway
Unit Value
Theoretical
potential
GWh 4377
pairs 381
Non-realisable
potential
GWh 3020
% of GWh 69%
pairs 349
% of pairs 92%
Realisable potential
GWh 1356
% of GWh 31%
pairs 32
% of pairs 8%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 73
11.4.11.Portugal
Unit Value
Theoretical
potential
GWh 278
pairs 10
Non-realisable
potential
GWh 201
% of GWh 72%
pairs 8
% of pairs 80%
Realisable potential
GWh 77
% of GWh 28%
pairs 2
% of pairs 20%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 74
11.4.12.Spain
Unit Value
Theoretical
potential
GWh 721
pairs 95
Non-realisable
potential
GWh 525
% of GWh 73%
pairs 78
% of pairs 82%
Realisable potential
GWh 196
% of GWh 27%
pairs 17
% of pairs 18%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 75
11.4.13.Sweden
Switzerland 2
Unit Value
Theoretical
potential
GWh 217
pairs 62
Non-realisable
potential
GWh 136
% of GWh 63%
pairs 57
% of pairs 92%
Realisable potential
GWh 81
% of GWh 37%
pairs 5
% of pairs 8%
Unit Value
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 76
11.4.14. Switzerland
Unit Value
Theoretical
potential
GWh 457
pairs 37
Non-realisable
potential
GWh 291
% of GWh 64%
pairs 20
% of pairs 54%
Realisable potential
GWh 166
% of GWh 36%
pairs 17
% of pairs 46%
eStorage_D4.2 Overview of potential locations for new variable speed PSP in Europe 77
11.4.15.United Kingdom
Unit Value
Theoretical
potential
GWh 85
pairs 19
Non-realisable
potential
GWh 39
% of GWh 46%
pairs 11
% of pairs 58%
Realisable potential
GWh 46
% of GWh 54%
pairs 8
% of pairs 42%