Renewable and Sustainable Energy Reviews 47 (2015) 260–269

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Importance of islands in renewable energy production

and storage: The situation of the French islands
Gilles Notton n
Laboratory Sciences for Environment, University of Corsica Pascal Paoli, UMR CNRS 6134, Route des Sanguinaires, F20000 Ajaccio, France

art ic l e i nf o a b s t r a c t

Article history: The islands are often not or only partially connected to continental electrical networks and have to manage
Received 12 June 2014 their energy supplies in reaching themselves the production/demand balance while ensuring the quality of
Received in revised form the delivered electricity. The low consumption and the lack of interconnections induce higher energy costs
13 January 2015
than in other regions. For increasing their security, the islands prefer to use indigenous sources as renewable
Accepted 8 March 2015
Available online 1 April 2015
ones; unfortunately, the intermittence and stochastic character of these “fatal” energy sources make them
more difficult to manage and it is aggravated in the case of small island networks. The islands are good
Keywords: locations for using and testing new technologies of energy production and storage. Most islands have a good
Islands renewable energy (RE) potential often underused. The difficulties of electricity management in island grids
Renewable energy
are first shown particularly with the impact on the energy production cost. Then, the problem of the
Integration rate
integration of renewable energy sources (RES) in island electrical grids is highlighted. The energy situation of
Stochastic production
French islands is presented with a particular focus on the part of renewable sources. A higher share of
renewable energy in the energy mix can be reached only with the development of the energy production
efficient prediction, with the development of energy storage means and with an optimal management of the
energy flux via the utilization of smart grids.
& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
2. Electricity supply/demand balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
3. Case of island territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
4. Intermittent and stochastic characters of renewable energy sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
5. The islands: Territories suitable for the renewable energy development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268

1. Introduction In the Chapter 17 of Agenda 21 adopted at the Rio Conference [3],

In Europe, there are about 300 islands (6% of the territory) for
14 M-inhabitants, i.e. higher than the population of some European
countries. More than 100.000 islands of all sizes are scattered in
the World in all the latitudes and longitudes with almost 500
M-inhabitants [1,2]. The total islands area is 1/6 of the Earth area.
n
Corresponding author. Tel.: þ 33 495524152; fax: þ33 495524142.
E-mail addresses: gilles.notton@univ-corse.fr, gilles.notton@yahoo.fr
it is written that “the islands are specific from both environment and
development point of view; they are very fragile and vulnerable and
in the context of sustainable development, energy is the cornerstone
of their planning strategies”.
In the final document of the First European Conference on Sust-
ainable Island Development [4], is stated: “Non-renewable energy
sources must be considered as provisional solutions, unsuitable as a
long-term solution to the energy problem in islands“.
The aim of this article is to stress that the difficult energy context
of islands has led electricity distributors to develop, for a long time,
more efficient solutions for managing the network and integrating

http://dx.doi.org/10.1016/j.rser.2015.03.053
1364-0321/& 2015 Elsevier Ltd. All rights reserved.
 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
intermittent renewable energy. Our objective is to show, as said by
Marin [5] and Duic et al. [6], that islands have become genuine
laboratories of the future of energy sustainability. To reach this
objective, this paper is divided into 4 parts:
 In a first part, it will be shown that it is a difficult task for an
electrical system operator to maintain a precise balance
between the electricity production and the electrical consump-
tion in a conventional electrical network. Since the power
installed in the island grid is finite and small, any perturbation
(e.g. a sudden production decrease or an increase of the power
of a motor working) weakens this energy balance and disturbs
the frequency and the electricity quality. Some examples will
be given in view to illustrate this difficulty.
 Then, the energy mix of French islands will be presented and it
will highlight that fuel generators are mainly used inducing
electricity production costs much higher than in inland; which
has led each state to adopt specific measures to take into
account this specificities.
 In a third part, it will be underlined that the influence of the
variability and unpredictability of wind and solar on the
electricity management is all the more important that the
penetration levels is high; then, other generation units have
to compensate the difference and must be operated with a
sufficient reserve margin.
 Despite the fact that the intermittent character of the production
raises more complex problems for islands than for large inter-
connected electrical networks, it will be shown that the part of
“fatal” renewable energy system in islands is more important that
in mainland; it is due to the high cost of the other energy means
which pushes islands to use their natural resources; but this
objective can be reached only with the development of efficient
prediction of the energy production, through the development of
energy storage means and an optimal management of the energy
flux via the utilization of smart grids.
2. Electricity supply/demand balance
The electricity must be used immediately after being produced!
The energy manager must dispatch the electricity and ensure a precise
balance between electricity generation and demand. In a power sys-
tem, the power balance must be maintained at any moment, i.e. the
electricity demand must be balanced by the electricity generation at
all times. The electrical network must work at a stabilized frequency
(50 Hz, sometimes 60 Hz), thus, the production systems must perma-
nently adapt their production to the power consumption either in the
electrical stations or in the dispatching centre.
In normal mode, there is a situation of balance: the production is
always equal to the consumption and the electricity frequency weaves
very little around the reference frequency. If the production decreases
suddenly due to the loss of a production means (or a cloud passage
Table 1
Characteristics of electricity production plants [6].
Type Size (MW)
Nuclear power plant 400–1300 per reactor
Steam thermal plant 200–800 per turbine
Fossil-fired power plants 1–200
Combined-cycle plant 100–400
Hydro power plant 50–1300
Combustion turbine (light fuel) 25
Internal combustion engine 20
261
above a PV plant), this balance is weakened and the frequency falls
below the reference frequency. In this case, a rapid increase of the
delivered electrical power by a connected production means has to
occur (for example, an increase of the produced power of an engine
working at part-load) or the start-up of a new production means is
needed. But the rise speed in power of an energy plant and its starting
time is not instantaneous as seen in Table 1 [7]. As the run-up time is
long, the activation of a new production system must be anticipated;
as seen in Table 1, only a hydraulic plant starts rapidly, followed by a
light fuel turbine; internal combustion engines (fuel or gas) need
about 45 min before producing electricity.
If an increase of the consumption and consequently the start-up of
a new energy means are not anticipated, the imbalance continues and
it will need to shed a fraction of the load. If this load shedding is too
slow or insufficient, there is a risk of blackout. Similarly, if the load is
lower than the production, the frequency increases and this presents
dangers for the electrical machines, the security system stops the
electrical machine and leads to a power failure in the network.
3. Case of island territories
The insular electrical systems are governed by the same electro
technical rules as those that are applied in mainland networks but
the absence or the limitation of an interconnection to a large net-
work provide it some specificities recognized by the competent
European and French authorities [8].
As the islands are not or only partially interconnected, the energy
manager has to reach the supply/demand balance without the
assistance of external production means located on neighbouring
areas. Islands have a structural fragility: a short circuit in the electrical
system will generate a voltage drop in all the island [9,10], the low
inertia implies a high frequency variability with consequences on the
voltage [9,10]; the previous problems are compounded by the high
unit size of an electrical generator in comparison with the peak power
in the network. The default probabilities in an insular network are
very high compared with an interconnected network [10]. Voltage and
frequency drops are more numerous and deeper in islands than in
mainland (dozens each year) [9]. In Corsica, before the partial AC
interconnection with Sardinia, more than 200 failures per year
occurred on the transmission network with voltage and frequency
dips (less than 46 Hz) [11].
Islands have often a small population, a low and variable energy
consumption (in Corsica, a minimum of 130 MW in May and a
maximum of 500 MW in December) that prohibits the use of high
rated power production means and requires the utilization of electrical
plants with a small rated power for a better adaptation to the load and
for limiting the disturbances due to the loss of an electrical plant [9];
few electrical systems exist with a low rated power (Table 1). An
internal combustion engine operates between 65% and 100% of its
peak power (Table 1), using three 20 MW engines allows to have a
wider working range (between 13 and 60 MW) than using only one
60 MW engine (between 39 to 60 MW) (Fig. 1).
Minimum power capacity Rise speed in power per min Start-up time (h)
20% 1% 40 (cold)–18 h (hot)
50% 0.5–5% 11–20 h (cold)–5 h (hot)
50% - 80% 10% 10 min–1 h
50% 7% 1–4 h
5 min
15–20 min
65% 45–60 min
262 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
Fig. 1. Comparison between operation with 3  20 MW engines and with one
60 MW engine.
Table 2
Ratio Unit power/Consumed power (peak/troughs)
for various French islands [11].
French mainland (interconnected) 0.3–0.5%
Corsica 9–42%
Guadeloupe 15–22.5%
Guyana 23–40%
Martinique 17–29%
Reunion 12–24%
An electrical unit peak power must not exceed 25% of the average
power in the network [8] (thus, in the 300 GW European grid, the loss
of a 1.3 GW nuclear unit results in a ratio of 0.4%; in a 200 MW insular
grid, the loss of a most powerful unit of 40 MW results in a ratio of
20%). The ratios (unit power of the largest unit on the consumed
power) for the French mainland and island networks are shown in
Table 2. The ratio is very high in the islands and the loss of a power
unit has very negative consequences: the loss of a 1300 MW nuclear
unit on the French territory leads to a speed of the frequency variation
of only 6 mHz/s, but, in Corsica, the loss of the 50 MW DC/AC
conversion unit causes a variation of 2.8 Hz/s [12].
In view to have an available margin of power able to react
rapidly to an increase of the consumption or to a decrease of the
production by another plant, the production group must operate at
part-load (starting a new plant takes time (Table 1)); then, this it
runs unfortunately with a lower efficiency (the optimal operating
point is generally close to the rated power) and therefore produces
a kWh at a higher cost [10].
Another problem is linked to the over-specialisation of the
economy that forces to install often an over-sized energy capacity to
cover factors such as a high seasonal demand: the economy of some
islands is often based on tourism and then, the consumption increases
during summer when some energy resources are low or not available
as water resources; moreover, in this period, the drinkable water
demand is high making it more delicate to manage.
Then, in an island grid, the electricity generation is heavily
dependent on diesel engines, expensive and polluting but appro-
priate according to their small unit size [13] (see Table 1). The high
contribution of fuel in the electricity production in French islands
is presented in Fig. 2 [14–24] and Tables 3 and 4.
The fuel facilities (diesel engines using heavy or sometimes light
fuel) and the combustion turbines (light fuel) represent 58% of the
installed peak power and produce about 57% of the electricity (reach-
ing 100% in some islands) with about 16% of coal; the distribution of
the energy means differs widely compared to the French mainland
situation [25].
In Corsica, the dependence on external oil energy sources is
very important: 83% of the total energy consumption i.e. 8.5 t of
CO2 per inhabitant against 6.5 for the French average value; the
Fig. 2. (a) Repartition of the installed power in French islands (data from 2012). (C:
Cable—HF: Heavy Fuel—LF: Light Fuel—BC: Bagasse & Coal—BG: Biogas—BM:
Biomass—G: Geothermal—W: Waste—HD: Hydraulic Dam—SH: Small Hydraulic—
WT: Wind Turbine—PV: PV Plant). (b) Repartition of energy production in French
islands (data from 2012). CB: Cable—F: Fuel—C: Coal—B: Bagasse—BG: Biogas—G:
Geothermal—W: Waste—H: Hydraulic—WT: Wind Turbine—PV: PV Plant).
part of these greenhouse gases emissions due to energy produc-
tion is 82% against 65% in the national average [26]. The energy
situation of La Reunion is well described by Praene et al. [27].
The share of renewable energy in the total production is about 21%
in 2012 in the studied islands, with 11.8% for hydraulic, 4.2% for PV,
2.8% for bagasse, 1.3% for wind energy, 0.4% for geothermal and 0.16%
for respectively waste and biogas. The gaps are significant according to
the considered island.
The preponderant fuel utilization in the islands was confirmed
by Liotard [28], Weisser [29,30], Chen et al. [31] and Duic et al. [32]
and induces a strong dependence and an increasing sensibility to
the price of the oil barrel. In Corsica and in the overseas depart-
ments, the cost of electricity is, in the best-case scenario, equal to
at least twice that of the selling price of electricity secured by the
electricity tariff balancing out in all the French territory.
The European Union defined a special status for the “small
isolated networks” for which the states members can adopt specific
measures different from the European continent [33]. The French
law [34] identifies some “area no-interconnected to the continental
network” called ZNI (Zones Non Interconnectées) (Fig. 3): The
French Polynesia (11 islands) with Electricity of Tahiti (Suez) as
electricity supplier, New Caledonia with ENERCAL and EEC, Wallis
and Futuna (EEWF Suez), Corsica, French Overseas Collectivities
(FOC) (Saint Pierre and Miquelon, Saint Barthelemy and Saint
Martin) and French Overseas Departments (FOD) (Guadeloupe,
French Guyana, Martinique, Mayotte, Reunion) where EDF ensures
the public service (production, single buyer, transport, distribution,
marketing). The French Overseas Collectivities and Departments
must produce in totality their electricity (not interconnected) and
largely in Corsica (partially connected).
The French laws impose [34–36] an equalization of electricity tariff
over all the French Territory and took in place the Contribution of the
 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269 263

Table 3
Peak power (MW) distributions in the French islands.

C HF LF BC BG BM G W HD SH WT PV Sum

Corsica 150 186.3 105 1.7 194.1 26 18 86 767.1

Guadeloupe 175.8 110.2 93.5 13.5 8.7 27.4 64 493.1
Martinique 240 131.8 6.6 1.1 60 439.5
Reunion 211 140 211.5 3.9 123.4 11.6 14.8 152 868.2
Wallis & Futuna 8.4 0.2 0.11 8.7
Tahiti 149 12 46.5 11 218.5
Mayotte 78.1 8 86.1
St Pierre & Miquelon 26.2 0.6 0 26.8
St Barthelemy 18.6 7 0.1 25.6
St Martin 52.9 1.4 54.3
New Caledonia 376.4 0.2 77.4 0.4 36.9 3 494.4
French islands 150 1522.8 506 305 5.6 0.2 13.5 6.6 441.5 46.7 98.8 385.7 3482.3

C: Cable—HF: Heavy Fuel—LF: Light Fuel—BC: Bagasse & Coal—BG: Biogas—BM: Biomass—G: Geothermal—W: Waste—HD: Hydraulic Dam—SH: Small Hydraulic—WT: Wind
Turbine—PV: PV Plant.

Table 4
Energy (GW h) distributions in the French islands.

CB F C B BG G W H WT PV Sum

Corsica 722.8 999.6 8.8 336.1 28.6 101.1 2,197

Guadeloupe 911.3 545.4 62.1 1.7 50.0 13.8 51.8 91.5 1,727.7
Martinique 1490.8 19.1 1.6 79.6 1,591
Reunion 491.9 1346.5 267.0 8.4 489.1 16.9 191.2 2,811
Wallis & Futuna 19.2 0.5 0.1 19.8
Tahiti 367.2 162.9 7.3 537.4
Mayotte 248.9 13.1 262
St Pierre & Miquelon 42.6 0.9 43.5
St Barthelemy 99.2 99.2
St Martin 194 2 196
New Caledonia 1807.2 381.7 53 3.5 2,245.4
French islands 722.8 6671.9 1891.9 329.1 18.9 50.0 19.1 1384.2 152.8 489.3 11,730

CB: Cable—F: Fuel—C: Coal—B: Bagasse—BG: Biogas—G: Geothermal—W: Waste—H: Hydraulic—WT: Wind Turbine—PV: PV Plant.

Fig. 3. Some French islands all over the World.

264 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
Public Service of Electricity (CPSE) which is paid by all the electricity
consumers and serves particularly to compensate the high gap in
insular area between the production cost and the regulated selling
price of electricity. This CPSE is controlled and attributed by an Energy
Regulatory Commission (CRE).
According to the CRE, in 2010, the produced MWh cost was between
122 and 315 € for a regulated selling price at 51.7 €/MW h [37].
In 2011, the compensation of the production extra cost was
equal to 1221 M€ for a turnover of 862 M€ [12].
According to EDF [38], in France, the production cost is around
45€/MW h against 110 €/MW h in Corsica and Reunion, 140€ in
Martinique and Guadeloupe and 160€ in Guyana. Thus, the regulated
selling price is less than 50% of the cost of the electricity produced
locally.
The demographic trends and the consumption tendencies (as
the use of air conditioning …) induce that the ZNI (island areas)
are characterized by a growth of energy production higher than in
continent (þ 3.8% for FOD compared with þ1% for French main-
land). Fig. 4 shows the evolution of the maximum peak power and
of the annual produced energy for some French islands [14–23],
the average percentage of growth for peak power and energy is
written above or under the corresponding curve.
Other technical factors contribute to increase the production cost
as corrosion of the material, more modulated production peak, more
expensive maintenance due to the problems of distance. Another
specificity of non interconnected areas is the unusually high gap
between the produced energy and the energy sold (both technical
and non-technical losses). This gap between interconnected and
non-connected areas was: in 1999, 7% in mainland France, 9% in
Reunion, 10% in Martinique and Guyana, 13% in Guadeloupe [39,40];
in 2003, it reached 10% in Guadeloupe, Martinique and Reunion, 12%
in Guyana and remained close to 15% in Corsica.
In 2012, the technical and no technical losses were:
– on the Corsican network equal to 316 GW h i.e. 14.4% [14];
– 196 GW h i.e.11.3% in Guadeloupe [15];
– 169 GW h (10.6%) in Martinique [16];
– 284 GW h (10.1%) in Reunion [17];
– 3.48 GW h (17.6%) in Wallis and Futuna [18];
– 51.2 GW h (9.5%) in Tahiti [41].
In mainland France, these electrical losses were only 2.5% [23].
Moreover, the number of shedding in an islander grid is high, with
in average 2.1 shedding per month during the five last years [12],
Fig. 4. Evolution of the maximum peak power and annual electrical production in some French islands. (Average percentage of growth for peak power and energy written
above or under the corresponding curve).
between 0.3 in Corsica and 3.5 in Guadeloupe. When an incident
happens in a production means and if the electricity distribution must
be maintained in the island, the loss of a production plant must be
instantaneously compensated by other groups in operation (a stopped
group has a too long starting time) (see paragraph 1); but, if the
variation of the required power is higher than the available margin of
this operational groups, these plants are disconnected in order to
avoid that they operate outside of their operating range. Then, in
cascade, an isolated incident can lead to a general shedding (such an
incident occurred in Corsica in November 2005, 16 after the loss of
only one motor). Such a risk is higher in a ZNI due to the overall power
consumption: the small size of a ZNI implies that each production
means is an important part of the total production and consequently
has a strong impact on the electrical system when a problem occurs
for a production group [42].
Thus, the management of the electrical production and dis-
tribution in the French islands is more complicated than in the
French mainland due to the absence of interconnection and to the
low consumption. Moreover, the electricity production in islands is
accompanied by: a higher increase of the annual consumption and
of the peak power; higher energy losses in the electrical grid, a
limited choice for the type and the size of the usable electrical
production plant and especially a high electricity production cost,
3 to 4 times higher than in mainland.
Managing the electrical production in islands is difficult using
“conventional” plants, but the task is even more complex if “fatal”
renewable energy systems with a stochastic production are inserted
into the electrical grid.
4. Intermittent and stochastic characters of renewable energy
sources
The power system is a dynamic system, continuously subject to
changing conditions, some variations can be anticipated, while others
cannot be predicted [43]. A power system has to meet a continuously
varying consumer load. From a control perspective, the load is the
primary independent variable—the driver to which all the controllable
elements in the power system must be positioned and respond. The
load is continuously changing in amplitude at annual, seasonal, daily,
minute-to-minute and second-to-second time step. The reliability of
the system then becomes dependent on the ability of the system to
accommodate expected and unexpected changes and disturbances
while maintaining quality and continuity of service to the customers.
 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269 265

Wind and sun, as variable and largely unforeseeable generating
resources, impact all of these planning and operation processes.
The wind and solar variability has its own characteristics and time
frames. The variability and unpredictability of wind and solar are
much important than the penetration levels is high [44–46].
When the renewable electrical power is lower or higher than
the power demand, other generation units have to compensate the
difference of power. This implies that a system with a significant
amount of renewable energy systems must operate with a suffi-
cient reserve margin (electrical systems already in operation but in
part-load operation) since most traditional generation units
require a considerable start-up time. This allows the system to
cover up loss of generation when needed [47,48].
The use of solar or wind systems induces two main constraints
[9,48,49]:
1. Constraints related to the source: no guaranteed power due to
the intermittence of the source which induces an obligation to
continuously provide other means in reserve able to compen-
sate the variations of power and to react immediately (in
increasing or in decreasing their production in a short time).
These “fatal” sources are not always available when they are
needed and sometimes are present when the network does not
need them, their production being then in excess.
2. Technical constraints: wind and PV systems have no inertia and
then are sensible to voltage and frequency drops which cut the
production and aggravate the incidents. Moreover they don't
participate to the “system service”, to the electrical network
stability and quality (voltage and frequency regulation, switch-
on in autonomous mode or black-start). These systems behave
like passive generators from an electrical point of view.
distribution network by the network operator when the active power
from these plants reached 30% of the total active power in the network
[50]. The same maximum penetration rate of randomness renewable
electrical systems is used in other countries as Canary Archipelago
(Spain). This power limitation of RES imposes to develop methodol-
ogies for determining the renewable energy absorption capability [51].
Even, if the part of “fatal” energy production in ZNI in France is
still low, this limit of 30% is rapidly reached. For example, in Corsica,
the minimum consumption occurs in May (the needs of building
heating are nil and the touristic period has not yet started), with
about 150 MW (see Fig. 5) [14], thus only 45 MW of fatal renewable
energy (30% of 150 MW) can be sent in the electrical grid. 18 MW of
wind turbines and 86 MWp of PV systems are currently connected
to the network, and their production exceeds the integration limit
during some hours per year. This limit reached [12,52] in Corsica
and Guadeloupe is 34%, in La Réunion 35% and in Martinique 22%.
When this limit is exceeded, the network operator disconnects the
last commissioned PV or wind installations; thus, the last implemen-
ted renewable plants are the first disconnected ones. This rule limits
the development of photovoltaic and wind plants.
This limit is more easily reached in a small territory as an island
because the “profusion” of the renewable energy plants at an island
scale is very small. The “profusion effect” means that the random
fluctuations of wind or solar are statistically reduced when the sys-
tems are spread over a large territory and their production is injected
into the same grid. In cases of significant expansion, the production is
rarely nil or maximum.
Small island perimeters require a balanced geographical distribu-
tion of renewable projects over the island, indeed, if these projects are
geographically concentrated and if the wind falls or if the sky becomes
cloudy, a significant electrical power is suddenly lost which induces

The first constraint implies that the conventional energy system (as
fuel motor) must work at part-load to be able to react rapidly to a
quick variation of power produced by the wind turbine or the PV
plant. In this case, the efficiency of the conventional motor is reduced
involving an increase of the fuel consumption per kW h produced and
of the kW h production cost. Thus, it is impossible to optimise the
operating mode of the conventional energy means [48].
The second constraint implies to have a voltage and frequency
regulation provided by conventional electrical plants and there-
fore, some conventional production means with inertia must work
permanently.
The two constraints show the difficulty or even the impossibility, at
the current stage, to use “fatal” renewable energy systems alone
without using other electrical productions as conventional fossil
generators or at least guaranted renewable plants as hydraulic ones.
Thus, the integration of fatal renewable systems in an energy mix Fig. 5. Characteristic consumption curves for Corsica (in 2010).
allows to reduce the proportional share in the production cost of the
thermal plants (fuel, maintenance …) but not the fixed costs [12].
The penetration rate of renewable systems, i.e. the power
generated by renewable systems compared with the total electrical
consumption must be limited in view to guarantee an electrical grid
stability and a precise production/consumption balance. The high
sensibility of PV and wind system to electrical grid disturbances, as
voltage dips or frequency variations, often leads to a disconnection
of their production when some incidents happen in the network.
This disconnection aggravates the imbalance between production
and consumption and by domino effect increases the risk of major
incidents. Some feedbacks in Denmark show that for a penetration
rate up to 20 or 30% some stability problems can occur [49]. The
barriers to penetration of RES in the insular territory of La Reunion
is discussed by Praene et al. [27]
Thus, a French order of the 23 April 2008 states that any facility
[…] using fatal energy with randomness characters such as wind
turbines or photovoltaic plants can be disconnected from the public Fig. 6. Illustration of the profusion effect for PV production in France (15/03/2013) [53].
266 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
important problem for the electrical management. Similarly, when the
sky clears the production of PV systems increases suddenly and the
total production can exceed the load consumption.
If the production of one wind farm is intermittent, the wind
production at national French level (with a wide distribution of
wind turbines across a territory) is not or less intermittent because
the geographical profusion and the presence of three uncorrelated
wind regimes (oceanic, continental and Mediterranean) allows an
attenuation of the production variation. Such an effect is illu-
strated in Fig. 6, the hourly production of a PV system alone suffers
sudden variations but on both regional and national level; the PV
production is smoothed by “profusion effect” [53]. This smoothing
is strongly reduced in a small territory and more often, the
production of randomness renewable systems spread over the
island area varies in the same direction, complicating the work of
the electrical grid operator.
The « profusion » effect affects also the load consumption: in a
mainland, the profusion of loads due to different electrical demands in
each region allows to predict the future load with a better accuracy
and then, the programming and the management of production
means is more easy than in an island where this profusion effect is
very limited [54].
The small penetration rate limits the development of “fatal”
energy systems and the contribution of these systems to the global
energy balance of the islands. Consequently, and as said by Duic
and Carvalho [13], a higher penetration of renewable energy
sources in islands is limited by its intermittent nature, which can
be increased if some kinds of energy accumulation are used.
5. The islands: Territories suitable for the renewable energy
development
It may seem a contradiction: managing an electrical grid in
islands is more complex than in inland for all the reasons previously
exposed and even more when “fatal” renewable energy systems are
integrated in this grid; despite this, in these areas the rate of fatal
renewable energy (i.e. without hydraulic and biomass energy) in the
total production reached 5.5% in 2012 (14% of installed power) i.e.
more than in the French mainland territory (3.5% in energy and 8.5%
for the installed power). The parts of the total renewable energy and
“fatal” renewable energy in the electrical production are shown in
Fig. 7 [14–24]. Despite the difficulties for integrating fatal resources in
Fig. 7. Part of the renewable energy in the installed electrical power and electrical energy. (Cor: Corsica; Gua: Guadeloupe; Mar: Martinique; Reun: Reunion; W&F: Wallis &
Futuna; Tah: Tahiti; May: Mayotte; SB: St Barthelemy; SM: St Martin; NC: New Caledonia; All: All islands; Fra: France).
island grids, the islands reached a higher level of integration of fatal
renewable energy than in mainland.
The part of the installed renewable power in France is higher than
in the islands because the share of hydraulic power in France is high
(second position after nuclear energy). The difference between the
islands is important: in term of installed renewable power, from 42%
for Corsica and Guadeloupe to around 3% for St Martin, St Pierre &
Miquelon and Wallis and Futuna and practically nil for St Barthelemy
and in term of electrical energy, a maximum of 35% is obtained for
Reunion. The fatal energy part reaches 19% of the total installed power
in Guadeloupe and Reunion and up to 8.3% in energy for Reunion.
This strong interest for fatal renewable energy systems in the
islands is mainly due to the high energy production cost underlined in
the second paragraph. The utilisation of costly fossil fuels due to the
fluctuations in crude oil prices, the small size of the production means
and a more expensive maintenance due to the remoteness cause such
high costs and encourage the emergence of innovating solutions as the
development of renewable energy and the setting up of a policy of
electricity demand reduction and energy management.
Fig. 8 shows the electrical kW h cost for various energy means
according to different recent references (From A to E):
– (A) Lacoste [12]: Production costs in insular context; for the
thermal means, there are a fixed cost (independent of the
production) and a proportional cost (combustibles, …)
– (B) Cour des comptes (Court of Auditors) [55]: The report was
written for the French Senate on the real cost of electricity.
– (C): SCM [56]: Evaluated costs of electricity in Reunion Island
according to the production means.
– (D) EDF [57]: Average costs of electricity in France;
– (E) CRE [58]: Renewable energy costs in French mainland
As said by Lacoste [12], the cost of electrical production by
fossil fuel includes a fixed cost independent of the production and
a proportional cost. This cost depends also on the operating mode:
a fuel plant operating in semi-base has a doubled fixed part
compared with the same plant in basic regime. If a guaranteed
renewable energy system saves the complete cost, the unguar-
anteed renewable energy saves only the proportional cost.
Even if the kW h cost for a same electrical energy system varies in a
wide range according to the reference, Fig. 8 allows to see that:
– the “fatal” renewable energy (unguaranteed energy) is cheaper
than heavy fuel and, of course, than domestic fuel (used in
 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
Fig. 8. Electrical kW h costs for several energy means according to various references.
combustion turbines): this concerns wind energy, small hydraulic
energy and sometimes PV energy;
– the guaranteed renewable energy are less expensive than the
fuel in basis utilization as geothermal, biogas and are close to
the fuel cost as biomass.
The insertion of renewable energy systems in Gökceada, the
biggest island of Turkey; was recently studied and showed an
interesting reduction of the kW h cost of the electricity [59].
The islands have everything to gain by developing the use of
renewable energy from an economic point of view but also for
ecological reasons.
Various energy planning using renewable energy were discussed
and developed and particularly in Greece [60,61] and Crete [62].
In view to delete or to decrease the constraint due to the
integration limits, three solutions must be developed:
 To predict the production of fatal renewable energy sources in
view to achieve by anticipation an optimal switchover towards
conventional electricity plants in keeping in mind that the
starting of a conventional system requires a long time; the
prediction of these productions is generally less accurate in an
insular networks [54,63];
 To develop smart grids for the transport and the management
of the production means. The network will be managed with
more flexibility for reaching the production/demand balance in
controlling both the production and the consumption;
 To develop energy storage means in order to absorb the surplus
energy and to release it when the consumption requires. Seve-
ral energy storage means were studied for islands [64–68].
The objectives of this paper are not to develop more the three
interdependent solutions but to show that insular territories are
well adapted to implement and to test these solutions before being
extended to any other parts of the World. Moreover, the electricity
companies that manage electricity in the islands acquired a good
experience and reached a high level of knowledge because they
have been facing this challenge for a long time.
267
Several projects and works were developed for integrating renew-
able energy systems in island and to create Renewable Island. Bağcı
[69] showed that implementing renewable energy technologies in
Peng Chau Island, Hong Kong, is a good solution and introduced the
name of “Zero Energy Island”. Through a plan to promote renewable
energy use, in 2007, the Danish island of Samsø reduced its annual
energy balance to the extent that it now produces as much renewable
energy as it consumes from the mainland. This island achieved 100%
energy independence in less than five years [70]. Another good
example of renewable island is El’Hierro in Canary islands [71].
Centeno Brito et al. [72] imagined an island which electricity, heat
and mobility demand are fulfilled with sustainable and renewable
energies only using hydropumping system and other storages.
The French government launched two calls for tenders in 2010
and 2011 on the implementation of wind farms and PV systems in
insular territories with energy storage and prediction of produc-
tion [52]. Concerning the implementation of smart grids, two
projects were developed Millener in La Réunion, Corsica and
Guadeloupe [73–74] and Address in the islands of Houat and
Hoëdic in Brittany [75]. Two other experimentations on energy
storage were also recently developed:
– Pegase (Reunion) with a prediction at short term (48 h) and very
short term (some hours before) of the production, the implemen-
tation of a smart grid and an energy storage using NaS battery;
– Myrte (Corsica) with a PV plant coupled to an electrolyser and a
fuel cell in view to smooth the PV production and to control the
electrical production [76].
In each island an adapted politic on energy-saving and devel-
opment of efficient and/or renewable systems for heat production is
implemented and participates to the reduction of the electrical
consumption.
The French government, as part of the implementation of the
“Grenelle de l’Environnement” fixed some ambitious objectives for
the not-interconnected areas and particularly the islands:
268 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
– to reach the energy autonomy in 2020 with an objective of 30%
of renewable energies in the final consumption of Mayotte and
50% minimum for the other communities;
– to develop energy storage technologies and smart grids for
increasing the part of intermittent renewable sources;
– to develop for Guadeloupe, Guyana, Martinique and Reunion,
some programs in view to reach the energy autonomy in 2030;
– to develop a program of consumption management in specific
Energy-Climate plans;
– to adopt thermal rules for the production of solar domestic hot
water and for improving the thermal insulation and to develop
the utilisation of PV systems.
6. Conclusion
The integration and development of renewable energy in the
insular territories are in progress and often reached a higher level
than in French mainland.
In a first part, the general problem of the electricity manage-
ment in an electrical grid was reminded with the need at all times
to reach a precise balance between production and consumption.
In a mainland, this objective is achieved by using all the energy
means interconnected in the electrical grid.
For an insular area, generally not connected with a large
network, this balance is more difficult to obtain. Several specific
problems were underlined for the islands in the second paragraph:
– the need for low power electrical systems with rapid starting,
very costly and almost always using fuel;
– the presence of high energy losses in the electrical grid incr-
easing again the electricity cost;
– a high number of electrical failures and shedding;
– and consequently, an electrical kW h cost very high, 3 to 4 times
higher than in mainland.
These specific aspects and constraints have enabled the insular
electrical operators to develop for a long time a higher experience
in the network managing.
The intermittent character of photovoltaic and wind energy
production raises complex technical problems, both for mainland
and island electrical grid and these additional constraints were
reviewed in the third paragraph.
The higher cost of electricity production pushes islands to use
their natural resources in view to reach energy independence, to
reduce their fuel dependence and to preserve their environment.
The fourth paragraph showed that the share of randomness
renewable energy is higher in French insular territories than in
mainland France because in these areas these energies reach the
break-even point more easily, and particularly compared with the
cost of the electricity produced by fuel.
The continuity of the renewable energy development will not
be accomplished without the development of efficient prediction
of the energy production, without the development of energy
storage means and without an optimal management of the energy
flux via the utilization of smart grids.
The islands are becoming renewable energy development
laboratories and probably the islands maps the road for the future
of the energy supply in the World.
References
[1] Duić N, Alves LM, Chen F, Da Graça Carvalho M. Potential of Kyoto protocol
clean development mechanism in transfer of clean energy technologies to
small island developing states: case study of Cape Verde. Renewable Sustain-
able Energy Rev 2003;7:83–98.
[2] Duić N, Lerer M, Carvalho MG. Increasing the supply of renewable energy
sources in island energy systems. Int J Sol Energy 2003;23-4:177–86.
[3] Action 21. In: Declaration of Rio on environment and development, conference
of the United Nations on Environment and Development (CNUED). United
Nation Publications, New York, NY, 1993; 92-1-200147-5, p. 140–41.
[4] Insula – Unesco – European Commission—Consell Insular de Menorca. In: First
European conference on sustainable island development, The Minorca Com-
mitments, European Island Agenda, Ciutadella Declaration, Minorca, ECSID
Edition; 1997.
[5] Marin, C. Towards 100% RES supply. An objective for the islands . Available from:
〈http://www.eurocaribbean.org/chania/folders/08_Marin.pdf〉; 2004 (accessed on
29.01.09).
[6] Duić N, Da Graça Carvalho M. Increasing renewable energy sources in island
energy supply: case study Porto Santo. Renewable Sustainable Energy Rev
2004;8:383–99.
[7] Saguan M. L’analyse économique des architectures de marché électrique.
Application au market design du “temps réel”. PhD dissertation, April. University
Paris Sud XI; 2007 In French.
[8] EDF. Technical reference: Key parameters for the insular electrical systems safety. SEI
REF 05, Technical report, 01/12/2008. Available from: 〈http://corse.edf.com/fichiers/
fckeditor/Commun/SEI/corp/SEI_REF_05_V1.pdf〉; 2008 (accessed on 12.04.14). (In
French).
[9] EDF. Intermittent random renewable energy in the insular energy system: why a
maximum integration rate of 30%, EDF report, 19/05/2009, CCP SEI. Available from:
〈http://sei.edf.com/fichiers/fckeditor/Commun/SEI/corp/Pourquoi-30-dans-SEI.pdf〉;
2009 (accessed on 09.04.14). (In French).
[10] Marin D. Integration of wind turbines in insular electrical network. PhD
dissertation, April. Ecole Centrale de Lille; 2009 In French.
[11] Barlier, Y. The situation of electricity generation in Corsica. In: Proceedings of
the dissemination of the advanced control technologies and SCADA systems
for the isolated power networks with increased use of renewable energies;
3 March 2000, Ajaccio, France. p. 27–35.
[12] Lacoste, J. Les systèmes électriques insulaires. IEEE P&ES, 25/04/2013, RTE, La
Défense, Paris. Available from: 〈http://ewh.ieee.org〉; 2013 (accessed on
12.02.14) (In French).
[13] Duic N, Carvalho MG. Increasing renewable energy sources in island energy supply:
case study Porto Santo. Renewable Sustainable Energy Rev 2004;8:383–99.
[14] EDF SEI Corse. Generation adequacy report on the electricity supply—demand
balance.. Available from: 〈http://sei.edf.com〉; July 2013 (accessed on 09.04.14).
(In French).
[15] EDF SEI Guadeloupe. Generation adequacy report on the electricity supply—
demand balance. Available from: 〈http://sei.edf.com〉; July 2013 (accessed on
09.04.14). (In French).
[16] EDF SEI Martinique. Generation adequacy report on the electricity supply—
demand balance. Available from: 〈http://sei.edf.com〉; July 2013 (accessed on
09.04.14). (In French).
[17] EDF SEI La Réunion. Generation adequacy report on the electricity supply—
demand balance. Available from: 〈http://sei.edf.com〉; July 2013 (accessed on
09.04.14). (In French).
[18] IEOM (Institut d’Emission d’Outre Mer), Wallis et Futuna. Report ISSN 1635-2262.
Available from: 〈http://www.ieom.fr〉; 2013 (accessed on 01.02.14). (In French).
[19] EDF SEI Saint-Barthélemy. Generation adequacy report on the electricity
supply—demand balance. Available from: 〈http://sei.edf.com〉; July 2013
(accessed on 09.04.14). (In French).
[20] EDF SEI Saint-Martin. Generation adequacy report on the electricity supply—
demand balance. Available from: 〈http://sei.edf.com〉; July 2013 (accessed on
09.04.14). (In French).
[21] EDF SEI Saint-Pierre-et-Miquelon. Generation adequacy report on the elec-
tricity supply—demand balance. Available from: 〈http://sei.edf.com〉; July 2013
(accessed on 09.04.14). (In French).
[22] Enercal. Activity report 2012–2013. Available from: 〈www.enercal.nc〉; 2013
(accessed on 09.04.14). (In French).
[23] RTE (Réseau de Transport d’Electricité). Electricity balance2012. Available
from: 〈http://www.rte-france.com〉; 2013 (accessed on 02.03.14). (In French).
[24] Lucking, R. Production manager of «Electricité et Eau de Wallis et Futuna»,
Personal communication (01/06/2014).
[25] CEA. Energy handbook. Edition. ISSN-1280-9039. Available from: 〈www.cea.fr/
content/download/121828/.../Memento_sur_l_energie_2013.pdf〉; 2013 (accessed
on 01.05.14). (In French).
[26] Corsican Territorial Authority. Energy policy: climate, air, transport and
sustainable mobility. Report no. 2011/O2/230; 2011. (In French).
[27] Praene JP, David M, Sinama F, Morau D, Marc O. Renewable energy: progres-
sing towards a net zero energy island, the case of Reunion Island. Renewable
Sustainable Energy Rev 2012;16:426–42.
[28] Liotard C. L’électricité dans les îles: vers une production durable? In:
Guerassimoff G, Maïzi N, editors. Îles et énergie: un paysage de contrastes.
Presses de l’école des Mines; Collection Libres Opinions; 2008.
[29] Weisser D. On the economics of electricity consumption in small island
developing states: a role for renewable energy technologies? Energy Policy
2004;32:127–40.
[30] Weisser D. Power sector reform in small island developing states: what role
for renewable energy technologies? Renewable Sustainable Energy Rev
2004;8:101–27.
[31] Chen F, Duic N, Alves LM, Da Graça Carvalho M. Renewislands—renewable energy
solutions for islands. Renewable Sustainable Energy Rev 2007;11:1888–902.
 G. Notton / Renewable and Sustainable Energy Reviews 47 (2015) 260–269
[32] Duić N, Krajačić G, Da Graça Carvalho M. RenewIslands methodology for
sustainable energy and resource planning for islands. Renewable Sustainable
Energy Rev 2008;12:1032–62.
[33] European Union. Directive 2003/54/EC of the European Parliament and of the
Council of 26 June 2003 concerning common rules for the internal market in
electricity and repealing Directive 96/92/EC. Official J. European Union, L176-
37-L176-55, 15/01/2003; 2003.
[34] Legislative French Assembly. French Act 2000-108 (10/02/2000) on the
modernisation and development of the public electricity service; reinforced
on 01 January 2012. (In French).
[35] Legislative French Assembly. French Act SPEGEEG no. 2004-803 of August 9,
2004 relating to the public service of the supply of electricity and gas and to
other electric and gas companies; reinforced the 01 January 2013. (In French).
[36] Legislative French Assembly. Energy Program Act POPE no. 2005-781 of 13
July 2005 establishing France’s energy policy priorities, reinforced the 01/06/
2011; 2005. (In French).
[37] CRE (Energy Regulatory Commission). Available from: 〈www.smartgrids-cre.
fr〉; 2014 (accessed on 11.04.14). (In French).
[38] EDF-SEI Corse. Available from: 〈http://www.corse-energia.fr〉; 2014 (accessed
on 09.04.14). (In French).
[39] Syrota J. Evaluation des missions de service public de l'électricité. Paris:
Commission de Régulation de l'Electricité, La Documentation Française; 2000
(In French).
[40] Levratto, N. Fonctionnement et impact du mécanisme de compensation des
charges de service public de l’électricité sur l’offre d’énergie dans les zones
non-interconnectée in L. Vidal, Droit de l'énergie et des réseaux, Annales de
l’Institut André TUNC, volume IX de la Bibliothèque de l’Institut André TUNC,
Paris, LGDJ; April 2005.
[41] EDT (Electricité de Tahiti). EDT in brief. Available from: 〈www.edt.pf〉; 2013
(accessed on 09.04.14). (In French).
[42] DGAC (Direction Générale de l’Air et du Climat). Programmation pluriannuelle
des investissements (PPI) de production d’électricité Période 2009–2020.
Rapport au Parlement. Available from: 〈http://www.developpement-durable.
gouv.fr/IMG/pdf/ppi_elec_2009.pdf〉; 2009 (accessed on 13.04.14). (In French).
[43] Piwko, R., Bai, X., Clark, K., Jordan, G., Miller, N., Zimberlin, J. The effect of
integrating wind power on transmission system planning, reliability, and
operation. Prepared for The New York State Energy Research and Devlopment
Authority Albany, NY; Report March 4, 2005.
[44] Czisch, G, Ernst, B. Hugh wind power penetration by the systematic use of
smoothing effects withing huge catchment areas shown in a European
example. In: Proceedings of the wind power; 2001, AWEA, Washington, DC.
[45] Luther, G. The regional function energy. In: Proceedings 12th European
community photovoltaic solar energy conference; 1992. p. 1313–16.
[46] Hay JE. An assessment of the mesoscale variability of solar radiation at the
earth’s surface. Sol Energy 1984;32:425–34.
[47] Van Hertem, D. The use of power flow controlling devices in the liberalized
market. Catholic University of Leuven report; January 2009. ISBN 978-94-
6018-024-8.
[48] Pestourie, J. Les EnR intermittentes: cas des réseaux insulaires. IEEE P&ES, 25/
04/2013, RTE, La Défense, Paris. Available from: 〈http://ewh.ieee.org〉; 2013
(accessed on 12.02.14). (In French).
[49] Robyns B, Davigny A, Saudemont C, Ansel A, Courtecuisse V, Francois B, et al.
Impact de l'éolien sur le réseau de transport et la qualité de l'énergie. J3eA
2006;5 In French.
[50] Legislative French Assembly. French Act no. 2008-386 of April 23, 2008
relating to the general technical prescriptions for the conception and the
operation for the connection of production installations to the public electrical
networks; 2008. (In French).
[51] Papathanassiou SA, Boulaxis NG. Power limitations and energy yield evalua-
tion for wind farms operating in island systems. Renewable Energy
2006;31:457–79.
[52] Harriche, F., Souletis, R. Technico-economical study of electricity storage; 17/
03/2013. Available from: 〈http://www.smartgrids-cre.fr〉; 2013 (accessed on
23.01.14). (In French).
[53] RTE (Réseau de Transport d’Electricité). Panorama des Energies Renouvelables
2013, 〈www.rte-france.com〉; 2013.
[54] Wang Y. Evaluation de la Performance des Réglages de Fréquence des
Eoliennes à l’Echelle du Système Electrique: Application à un Cas Insulaire.
PhD dissertation. Ecole Centrale de Lille; 20/11/2012.
269
[55] Cour des comptes. La contribution au service public de l'électricité (CSPE):
Suites données aux observations de la Cour dans le rapport public 2011,
Rapport communication à la commission d'enquête du sénat sur le coût réel
de l'électricité afin d'en déterminer l'imputation aux différents agents écon-
omiques; June 2012.
[56] SCM (Société de Calcul Mathématique SA). Production locale d’électricité vs
Réseau maillé: Eléments de comparaison; Rapport présenté au Réseau de
Transport d’Electricité; September 2012.
[57] EDF. Electricity in France: the various energy sources. Available from: 〈https://
www.lenergieenquestions.fr/les-couts-de-production-des-differentes-sour
ces-denergie-en-france-infographie〉; 2013 (accessed on 20.05.14). (In French).
[58] CRE (Energy Regulatory Commission). Rapport d’analyse sur les couts et la
rentabilité des énergies renouvelables en France métropolitaine, 17/04/2014.
Available from: 〈www.cre.fr〉; 2014 (accessed on 20.05.14). (In French).
[59] Demiroren A, Yilmaz U. Analysis of change in electric energy cost with using
renewable energy sources in Gökceada, Turkey: an island example. Renewable
Sustainable Energy Rev 2010;14:323–33.
[60] Giatrakos GP, Tsoutsos TD, Mouchtaropoulos PG, Naxakis GD, Stavrakakis G.
Sustainable energy planning based on a stand-alone hybrid renewable energy/
hydrogen power system: application in Karpathos island, Greece. Renewable
Energy 2009;34:2562–70.
[61] Georgiou PN, Mavrotas G, Diakoulaki D. The effect of islands’ interconnection
to the mainland system on the development of renewable energy sources in
the Greek power sector. Renewable Sustainable Energy Rev 2011;15:2607–20.
[62] Tsioliaridou E, Bakos GC, Stadler M. A new energy planning methodology for the
penetration of renewable energy technologies in electricity sector—application for
the island of Crete. Energy Policy 2006;34:3757–64.
[63] Voyant C, Paoli C, Muselli M, Nivet ML. Multi-horizon solar radiation
forecasting for Mediterranean locations using time series models. Renewable
Sustainable Energy Rev 2013;28:44–52.
[64] Kaldellis JK, Zafirakis D, Kavadias K. Techno-economic comparison of energy
storage systems for island autonomous electrical networks. Renewable Sus-
tainable Energy Rev 2009;13:378–92.
[65] Kaldellis JK, Kapsali M, Kavadias KA. Energy balance analysis of wind-based
pumped hydro storage systems in remote island electrical networks. Appl
Energy 2010;87:2427–37.
[66] Carapellucci R, Giordano L. Modeling and optimization of an energy genera-
tion island based on renewable technologies and hydrogen storage systems.
Int J Hydrogen Energy 2012;37:2081–93.
[67] Rious V, Perez Y. Review of supporting scheme for island power system
storage. Renewable Sustainable Energy Rev 2014;29:754–65.
[68] Karellas S, Tzouganatos N. Comparison of the performance of compressed-air
and hydrogen energy storage systems: Karpathos island case study. Renew-
able Sustainable Energy Rev 2014;29:865–82.
[69] Bağcı B. Towards a zero energy island. Renewable Energy 2009;34(3):784–9.
[70] Jørgensen, P.J., Hermansen, S., Johnsen, A., Nielsen, J.P., Jantzen, J., Lundénn M. Samsø
– a renewable energy island – 10 years of development and evaluation. Report
2007. ISBN 978-87-92274-01-4. Available from: 〈http://energiakademiet.dk/wp-con
tent/uploads/samso-renewable-energy-island.pdf〉; 2007 (accessed on 20.05.14).
[71] Iglesias G, Carballo R. Wave resource in El Hierro—an island towards energy
self-sufficiency. Renewable Energy 2011;36:689–98.
[72] Centeno Brito M, Lobato K, Nunes P, Serra F. Sustainable energy systems in an
imaginary island. Renewable Sustainable Energy Rev 2014;37:229–42.
[73] Chardin-Segui, T, Mathiou, S; Jeanjean, Y. The Millener project: five hundred
RES & one thousand fifty DSM installations for households in French islands,
aggregated for grid support. In: Proceedings of the PowerTech (POWERTECH);
2013, IEEE Grenoble, 16–20 June 2013.
[74] Marcel, J.C. The Millener project, 1500 installations for households in French
insular electric systems, aggregated for grid support. In: Proceedings of the
IRED é012, Berlin; 5/12/2012.
[75] Belhomme, R; Cerero, R; Valtorta, G, Eyrolles, P. The ADDRESS project:
developing active demand in smart power systems integrating renewables.
In: Proceedings of the power and energy society general meeting; 2011 IEEE,
San Diego, CA, 24–29 July 2011.
[76] Darras C, Muselli M, Poggi P, Voyant C, Hoguet JC, Montignac F. PV output
power fluctuations smoothing: the MYRTE platform experience. Int J Hydro-
gen Energy 2012;37:14015–25.