market applications for energy storage methods and res ... · contract n°: eie/07/159/si2. 466845...

STORIES Addressing regulations on STORage technologies for increasing the

penetration of Intermittent Energy Sources, Contract N°: EIE/07/159/SI2. 466845

Workpackage 2 :

Market applications in specific island power systems

Deliverable D2.1:

Market applications for energy storage methods and RES

units: Case studies

Authors :

Dr Antonios Tsikalakis

Prof. Nikos Hatziargyriou

Mr Panagitois Koukoutsakis

Dr George Caralis

Prof. Arthouros Zervos

National Technical University of Athens

Dr Manos Zoulias

Dr Manos Stamatakis

Dr. Olga Parissis

Mr. George Tzamalis

Centre for Renewable Energy Sources

Dr Salvador Suarez Garcia

Mr Daniel Henríquez Alamo Canary Islands Institute of Technology

March 2009

2

Distribution List:

Organization YES

Centre for Renewable Energy Sources

Nacional Technical University of Athens

Canary Islands Institute of Technology

Instituto de Engenharia Mecanica – Polo IST

Regulatory Authority for Energy of the Hellenic Republic

Western Isles Council – ISLENET

European Renewable Energy Council

SOFTECH Energia Tecnologia Ambiente s.r.l.

University of Zagreb

Cyprus Energy Regulatory Authority

European Commission

EXECUTIVE AGENCY FOR COMPETITIVENESS AND INNOVATION

Contract N°: EIE/07/159/SI2. 466845

3

1 CONTENTS 1 Contents ............................................................................................................................................3 2 Executive Summary .........................................................................................................................6 3 Introduction ......................................................................................................................................7

3.1 Technologies used ....................................................................................................... 7 3.1.1 Batteries................................................................................................................... 9 3.1.2 Pump hydro storage............................................................................................... 10 3.1.3 Hydrogen ............................................................................................................... 13 3.1.4 Desalination systems ............................................................................................. 15

3.2 Short description of the Island systems selected to be simulated .............................. 20 3.2.1 Milos case study .................................................................................................... 20 3.2.2 Ios Case study........................................................................................................ 23 3.2.3 Cyprus case study .................................................................................................. 36 3.2.4 La Graciosa case study .......................................................................................... 40 3.2.5 Corvo case study.................................................................................................... 44 3.2.6 San Pietro case study ............................................................................................. 45 3.2.7 Mljet case study..................................................................................................... 49

3.3 Matching Technologies and islands .......................................................................... 55 4 Simulation with Battery Storage...................................................................................................56

4.1 Case study 1 results-La Graciosa............................................................................... 56 4.1.1 Scenario 1(Optimization for maximum RES penetration) .................................... 56 4.1.2 Scenario 2-50% RES share.................................................................................... 64 4.1.3 Comparison between RES penetration scenarios .................................................. 70 4.1.4 Summary ............................................................................................................... 71

4.2 Case study 2 San Pietro ............................................................................................. 72 4.2.1 Scenario 1 .............................................................................................................. 72 4.2.2 Scenario 2 .............................................................................................................. 73 4.2.3 Summary ............................................................................................................... 79

4.3 Conclusions of the case studies analysed .................................................................. 81 5 Simulation with pump hydro storage ...........................................................................................82

5.1 Introduction ............................................................................................................... 82 5.2 Methodology ............................................................................................................. 82

5.2.1 Definition of the Wind installed capacity outside the WPS .................................. 82 5.2.2 Operation and architecture of the WHPS .............................................................. 88 5.2.3 Methodology - Simulation..................................................................................... 90 5.2.4 Pre-feasibility study............................................................................................... 95

5.3 Case study 1 results-Ios ........................................................................................... 101 5.3.1 Adding wind without Pump Hydro Storage ........................................................ 101 5.3.2 Scenario 2-Pump Hydro Installation ................................................................... 101 5.3.3 Summary ............................................................................................................. 111

5.4 Case study 2 results-Cyprus .................................................................................... 114



4

5.4.1 Wind potential ..................................................................................................... 114 5.4.2 Scenario 1 – Adding wind power without Storage in Cyprus ............................. 115 5.4.3 Scenario 2 - Pump Hydro Installation ................................................................. 115 5.4.4 Summary.............................................................................................................. 129

5.5 Case study 3 results-Corvo ...................................................................................... 131 5.5.1 Scenario 1- Wind power in Corvo without storage ............................................. 131 5.5.2 Scenario 2 - Pump Hydro Installation ................................................................. 131 5.5.3 Summary.............................................................................................................. 139

5.6 Conclusions of the case studies analysed ................................................................ 141 6 Simulation with Hydrogen ..........................................................................................................143

6.1 General Description................................................................................................. 143 6.2 Milos RES & Hydrogen results-Case Study 1......................................................... 143

6.2.1 Simulation results of the existing power system of Milos island scenario 1 ....... 143 6.2.2 Scenario 2: Introduction of hydrogen (installed capacity of fuel cells accounts for a

10% of peak demand).......................................................................................................... 144 6.2.3 Comparison between the existing and the proposed hydrogen-based power system

of Milos island..................................................................................................................... 146 6.2.4 Summary of Milos Results .................................................................................. 147

6.3 Corvo RES & Hydrogen results-Case Study 2 ........................................................ 149 6.3.1 Scenario 1-Simulation of the existing power system in Corvo ........................... 149 6.3.2 Scenario 2-Results from the proposed Hydrogen-based power system .............. 149 6.3.3 Comparison between the existing and the proposed hydrogen-based power system

of Corvo island .................................................................................................................... 150 6.3.4 Summary of the results for Corvo ....................................................................... 152

6.4 Conclusions of the case studies analysed ................................................................ 153 7 Simulation with Desalination ......................................................................................................154

7.1 Methodology followed for Milos and Mljet ............................................................ 154 7.1.1 Algorithm explanation......................................................................................... 155 7.1.2 Water level calculation ........................................................................................ 156 7.1.3 Level check subroutines ...................................................................................... 156

7.2 Case study 1 results-Milos....................................................................................... 160 7.2.1 Adding Wind power production without desalination-Scenario 1 ...................... 160 7.2.2 Adding wind power production and desalination plant-Scenario 2..................... 161 7.2.3 Adding wind power production and desalination plant co-operation-Scenario 3 165 7.2.4 Adding desalination plant co-operation without wind-Scenario 4 ...................... 168 7.2.5 Summary.............................................................................................................. 170 7.2.6 Conclusions ......................................................................................................... 176

7.3 Case study 2 results-Mljet ....................................................................................... 178 7.3.1 Scenario 1 ............................................................................................................ 178 7.3.2 Scenario 2 ............................................................................................................ 186 7.3.3 Summary.............................................................................................................. 201 7.3.4 Conclusions ......................................................................................................... 204

7.4 Case study 3 results-Cyprus .................................................................................... 206



5

7.4.1 Explanation of scenarios 4 & 5............................................................................ 207 7.4.2 Results ................................................................................................................. 208 7.4.3 Summary -Cyprus................................................................................................ 214 7.4.4 Conclusions of the case studies -Cyprus ............................................................. 218

7.5 Conclusions of the case studies analysed ................................................................ 219 8 General synopsis &comparisons .................................................................................................221 9 Bibliography .................................................................................................................................223



6

2 EXECUTIVE SUMMARY Intermittent nature of Renewable Energy Sources (RES) is one of the significant technical

barriers in increasing RES penetration especially in non-interconneted power systems. Unfortunately demand does not coincide with the RES production creating either significant shortage or significant excess of energy produced by RES. To make matters worse, the thermal units operating to meet the most significant part of the demand cannot always decrease their output beyond a certain technical limit. Even worse, especially for larger island systems, quite often these units cannot be switched off and then switch back on in very short time in favour of increased RES production. Additionally, they cannot always follow the sometimes rapid changes in both demand and RES production. These reasons create limitations to RES penetration especially on island systems.

Matching supply and demand and provide support to the power system either by absorbing energy when it is excess or by injecting energy when there is shortage, is one of the main tasks that energy storage can perform in a power system. Therefore, studying the impact of combining RES and energy storage for islands is essential in order to examine what are the benefits of such an operation for a power system with respect to increasing RES penetration and power system operation. This is the main goal of this deliverable.

More precisely this deliverable provides results from the simulation of 7 different islands from 6 different countries with different sizes and electricity demand patterns, from as small as Corvo in the Azores Archipelago to as large as Cyprus in the Mediterranean for various levels of RES penetration combined with energy storage. More precisely, 3 different storage methods, namely batteries, Pump Hydro Storage and use of Hydrogen, produced by electrolysis driven by RES, for electricity production via a Fuel Cell are considered. Additionally, a demand side management methodology for desalination plants via Reverse Osmosis is also examined since water scarcity is a common characteristic not only for some of the islands studied, but also of other islands targeted by the STORIES action. The provided results have focused on the impact that the combination of RES and storage means can have on the operation of an island power system mainly in terms of system economics and emissions.

Further analysis on the results of the cost benefit analysis for the society as well, will be provided by the Deliverable D2.3. Investigation of the tariff schemes in order to support the simulated scenarios of operation will be provided in Deliverable 3.3. The results of this deliverable and the above mentioned deliverable will be systemized and synthesized in order to provide a roadmap for increasing RES penetration on the island power systems.



7

3 INTRODUCTION This section will provide all the introductory information on the report regarding

• Short characteristics of the technologies used

• Island Networks to be simulated

• Matrix of simulated case studies

Results from the simulations are provided in Chapter 4 for battery systems, Chapter 5 for pump Hydro, Chapter 6 for Hydrogen and Chapter 7 for desalination systems.

3.1 Technologies used Regarding the value of the storage systems it is common belief among the researchers that due

to the fact that these systems can play multiple roles, their evaluation does not often represent adequately this multiple role. A storage system, for example, can contribute to the spinning reserve of the system or reduce the RES curtailed but at the same time can also offer services in improving the voltage level. Thus, the recognize as well as the codes proposed by the regulatory authorities of energy do not recognize the multiple roles a storage system can play [1].

Recent reports of DTI [2,3] present the importance of energy storage in managing the uncertainty due to the participation of RES in the energy system of United Kingdom, one of the biggest islands in Europe, as well as the potential use of the storage systems in enhancing the supply security of the network. The study focuses on the contribution of storage in increasing the penetration of RES generation for various penetration levels and especially against the OCGT (Open Cycle Gas Turbines) technology.

The results focus on the following:

1. Reduction in the fuel cost

2. Reduction in the emissions

3. Increase in the absorbability of wind power In addition to the above studies regarding the combination of storage and RES, the combination

of distributed applications of energy storage in a network with the purchase of photovoltaic systems is described in report [4].

Various studies for energy storage in autonomous power systems have been performed. It has been proposed to disintegrate the unit commitment problem from the batteries management problem [5]. In [6] the addition of batteries in an autonomous system with high RES penetration is studied, while, for the storage capacity, the benefit compared to the average cost of the system is calculated. In such a case the value of the storage in time periods that can lead to avoid the dispatch of generating units and, as a result, to higher benefits is not calculated. Moreover in [7], the impact of energy storage in the secure and economic operation of the island of Kythnos has been studied. For the same island the potential impact of managing desalination units has been also studied [8].

To sum up, the possible applications the storage systems can have in an electricity system, some of which can be served by the same storage systems either simultaneously or in different time periods, can be in the area of RES [9].



8

• Use of the storage system to smooth the output by RES and for the efficient participation of RES in the electricity markets [10, 11] by storing energy during the low price periods and using it during high price periods taking advantage of the price difference.

• Reduction of the RES curtailment energy, energy storage for future use when generation is not available, which is a usual practice in small autonomous and islanded systems.

• Deferral of distribution network investments especially if a combination of storage device with some form of RES generation exists [12], as long as the appropriate location for the installation has been chosen.

Table 3-1 presents the various energy methods used for the various storage devices, while Fig. 3.1 presents the various storage methods with regards to their size, application and technological maturity.

Table 3-1 Energy conversion methods for various storage devices

Energy Conversions method Representative storage Devices Electric charge Capacitors and Super capacitors Super conducting materials Magnetic Energy Storage (SMES)

Pump Storage Compressed Air Energy Storage (CAES)

Mechanical power (Dynamic or rotational)

Flywheels Chemical methods Various types of Batteries, Hydrogen Storage Of the above methods the most suitable, mature and used technologies for direct electricity use

in Autonomous power systems with significant penetration in Renewable Energy Sources (RES) are:

• Batteries • Pump Hydro Storage

However, although not an as mature technology as the other two, Hydrogen can be used as a

storage medium for a) producing electricity via a Fuel Cell, and b) using it as a fuel for transportation on island either on land or marine. Therefore, producing Hydrogen via electrolysis combined with RES is another option that is studied in this deliverable.

Finally desalination is not a direct storage method but can be considered as a Demand Management Method. Desalination units via reverse osmosis can be more easily scheduled than other types of loads since water can be much easier stored than electricity and used in a latter time. Thus, desalination scheduling can be made in such a way that more water is produced when more RES production is available or even to exploit RES production that could have been otherwise curtailed, reducing the impact on the power system of simply adding desalination plants.



9

Fig. 3.1 Comparison of size, technological improvement and applications for various storage

devices.

Some details on these 4 technologies are provided below as well as the islands selected to be simulated.

3.1.1 Batteries

Batteries are the most widespread storage devices all over the world. It is estimated that one third of the world population, non –interconnected with the grid, use batteries to meet some of their neeeds. The limited life-time period for the batteries, especially in deep discharge mode, is balanced by their low capital cost, especially for lead-acid batteries. However, the battery manufacturers are constantly searching ways of lenghtening their life time.

The most widespread type is Lead Acid batteries. Other types of batteries are under the stage of development or demonstration applications for hedging demand and RES production. Advance on these technologies is rapid and soon will start been installed in many location including islands.

Table 3-2 summarizes the types of batteries that can be used and their usual applications. The number of crosses denotes how suitable each type of battery is for power or energy nature applications/

For the purposes of the deliverable however, and in order to use a more mature technology, Lead –acid batteries have been used.



10

Table 3-2 Other than lead-acid batteries and their suitability for power/energy applications

Battery Advantages Disadvantages Power Applications

Energy Application

s

Usual Applications

Flow Battery: PSB, VRBr, ZnBr

High Capacity Low Energy Density

+ ++ Demand balancing and long term storage

Ni – Cd High Energy Density

Difficulty in waste management

++ + Demand balancing for hours-minutes

Li – ion High Energy

Density, high efficiency

High Production cost,

++ _ Portable Electronic devices, limited in the field of Power systems

NaS High Energy and power Density, high efficiency

High Production cost, security measures

++ ++ Demand balancing for hours-minutes

3.1.1.1 LEAD-ACID BATTERIES This technology is the most widespread of all ,and the qualities of this tpye of batteries are

improved as the time goes by. One of the significant changes is the use of electrolyte in gel and not as liquid, making the batteries less vulnerable to stratification issues. Significant applications of batteries worldwide for power systems are provided in Table 3-3.

Table 3-3 Applications of Lead-Acid Batteries for providing Ancillary Services in Electricity networks.

Location-Characteristics Size Application Southern California Edison

Chino, CA USA 10 MW/40 MWh Load levelling

Puerto Rico El. Power Authority San Juan, Puierto Rico 20 MW/20 MWh Frequency Support

GNB Industrial Power Metlakatla, Alaska,USA 1 MW / 1.4 MWh Avoidance of load

disconnection

Most Autonomous power systems for providing electricity to very remote applications running on significant percentage on RES, e.g. telecommunication applications are based on lead-acid batteries. On the island of Kythnos, a battery bank of lead-acid can provide significant aid for running for short time the power system under 100% RES [13].

3.1.2 Pump hydro storage

3.1.2.1 DESCRIPTION OF THE SYSTEM Pumped Storage is considered as the most suitable storage technology for achieving high wind

penetration levels in medium or large autonomous power systems. In autonomous power systems wind farms owners face curtailment by the system operator of surplus wind-generated power



11

during periods of low demand. The ability to balance demand with wind power, defines the wind capacity to be installed A hybrid Wind Hydro Pumped Storage (WHPS) - always comprised by new wind farms, two reservoirs for the recycling of water, hydro turbines, pumps and penstocks - is proposed as a mean to increase the wind installed capacity, substitute expensive fuel oil and reduce the required conventional installed capacity (Fig. 3.2). The later is possible because the variable output of wind power is managed and transformed into a guaranteed power supply. and the wind power to be absorbed by the grid.

Fig. 3.2 General concept of the WPS in autonomous power systems [14]

The wind farm and the hydro pumped storage system are not necessarily installed at the same location. The topography should permit the construction of the two reservoirs, in a small distance and with sufficient hydraulic head as Fig. 3.3 shows.

Wind Park

Energy produced by the wind park

Electrical grid

Wind Energy Stored by the Pump station or

energy imported by the electrical grid

Energy Produced by the turbnines

Water Pumps(n pumps in parallel)

Hydraulic Turbines(nT turbines in parallel)

Flow direction for the turbining operation

Flow direction for the pumping operation

Upper Storage Reservoir

Lower Storage Reservoir

Water pipe(s)of diameter D and

lentgh L

z1

Fig. 3.3 General concept of energy and water flows in the WPS [15]



12

3.1.2.2 OPERATION OF THE SYSTEM The components of the WHPS are directly connected to the grid. Wind power may be given

directly to the demand or may be used for pumping through the grid. The priority is dependent on the legislative framework, the operational policy, and the tariffs. The regional use of energy imposes that if the grid could absorb wind power, wind power should be absorbed rather than used for pumping. On the other hand, if there is a wide range between the electricity production cost of the base load units and the peak supply units, then it may be preferable to use wind power in priority for pumping in order to substitute expensive peak supply units.

Wind power which is not absorbed directly by the grid is stored via pumping in the upper reservoir. Pumping station typically comprises several individual pumps. Then, given the available power for pumping, the number of pumps to be committed is decided. Energy stored at the upper reservoir is recovered via the hydro turbines and always provides peak demand supply. Simultaneous operation of the turbine and pumps is permitted with double penstock arrangements, and provides operational flexibility. Daily operational cycle is proposed to provide the feasibility of the investment.

Several policies has been proposed and analysed for the definition of the hydro turbine supply: - There are some voices, which propose the stable daily energy hydro-turbine production,

which is an ideal case for the hybrid’s operation and feasibility. Additionally, during windy periods and high water surplus in the upper reservoir, additional offer of operation of the hydro turbine may be proposed from the hybrid operator to the grid. This amount of energy could be absorbed from the system operator, under a lower price agreement.

- On the other side, the daily peak demand supply by the hydro-turbines is an ideal operation for the power system. In this case, the hydro-turbine’s energy output is modulated taking into consideration the daily energy demand.

As soon as the hydro-turbines provide fully dispatchable generation, they can substitute not only energy but also installed capacity of conventional thermal units. To make this possible, the provision of “guaranteed power” needs to be ensured during long low wind periods, when the water availability in the upper reservoir is exhausted. For this purpose, a limited amount of pumping is allowed using conventional energy from the grid. Grid pumping is reasonable to be scheduled during low demand hours. The amount of conventional power for pumping should be restricted by the regulatory framework, or the system’s operator should send to the WHPS operator a set-point which defines the maximum amount of conventional power that can be used for pumping.

An other restriction is always imposed and correlates the wind power capacity with the pumping station capacity. This restriction provides that the pumping station is able to use the wind power for pumping. In Greece, the wind installed capacity may exceed the pumps capacity only up to 20%.

The operation of the WHPS should not affect the operation and the wind power absorption from the wind farms outside of the hybrid system. The wind power from the wind farms is absorbed in priority, according to the absorption capability of the power system and the contracts between wind farm investor and system operator.

3.1.2.3 REGULATORY FRAMEWORK – TARIFFS The operational policy and the design of the WHPS is dependent on the regulatory framework

and the tariffs for the different energy or guaranteed power components. There are three components of income:

- Wind energy directly given to the grid. The price for this stochastic wind power production is always defined by the regulatory framework. Sometimes, it is dependent on the price of the consumer and / or the variable cost (fuel cost) of the base load power plants. Wind



13

power is considered to substitute only a part of the operation of conventional power plants and power pricing is not given.

- Hydro turbine’s energy production. The price for this fully dispatchable generation should be higher than the wind price and it should be related to the variable cost (fuel cost) of the peak supply thermal power plants.

- The provided “guaranteed power”, which is typically the rated capacity of the hydro-turbines. The definition for the price of the capacity credit should be related to the avoided cost of building a new conventional power station of similar capacity and includes the yearly capital amortization expenses and the fixed operating cost of an equivalent thermal power station. It is supposed that the hydro turbine has the same contribution to the reliability of the system with an equivalent thermal power station. This is true in case that the volume of the reservoir could provide the operation of the hydro turbine for several hours. Otherwise the reliability of the hydro turbine power supply is doubtful.

Finally, the pricing of the power which is derived by thermal power plants and is used for pumping, should be based on the mean annual variable operating cost of the base load power plants of the system.

3.1.3 Hydrogen

Hydrogen is an energy carrier that can be used as an energy storage method, especially in combination with renewable energy sources (RES) in non-interconnected power systems. Over time, hydrogen will provide an ideal storage medium for renewable energy, especially in non-interconnected islands. Hydrogen can be expected to allow the integration of some renewable energy sources, of an intermittent character, in the current energy system [16]. Thus, we can envisage photovoltaic panels (or wind turbines) linked to a water electrolyser, which uses excess electricity to produce H2 during the day (or in windy conditions). When the natural resource (sun, wind etc.) is not available, a fuel cell uses previously stored hydrogen to produce electrical energy and serve the electrical loads and consumes the hydrogen during the night (or in the absence of wind) to produce electricity. In spite of the undeniable lack in efficiency of this system, it is clear that it would provide an uninterrupted supply of electricity. Moreover, hydrogen can facilitate high RES penetration (especially wind energy penetration) in autonomous power systems of islands.

In more detail, the overall hydrogen energy system, envisaged to be installed in autonomous areas, including islands, comprises three basic sections: i) the hydrogen production section, ii) the hydrogen storage section and iii) the hydrogen re-electrification section. When produced through water electrolysis driven by RES, hydrogen is absolutely clean, since the only by-products of the overall procedure is water and heat. A schematic of the complete hydrogen-based power system is depicted in Fig. 3.4 [17]



14

Fig. 3.4 Electrolysis powered from renewable energy sources

The hydrogen production section can be based on either an Alkaline or a Proton Exchange Membrane (PEM) water electrolysis unit. Alkaline electrolysers use an aqueous KOH solution as an electrolyte. Alkaline electrolysis is best suited for stationary applications that are operating at pressures up to 25 bar. Alkaline electrolysers have been commercially for a long time. Usually, commercial electrolysers consist of a number of electrolytic cells arranged in a cell stack. Alkaline electrolysers typically contain the main components shown in Fig. 3.5. The major challenges for the future are to design and manufacture alkaline electrolysers al lower costs with higher energy efficiency and larger turn-down ratios [18].

Fig. 3.5 Process diagram of alkaline electrolysis for the production of H2

On the other hand in PEM electrolysers, there is an organic polymer membrane-based electrolyte, in which protons that are generated at the anode are transferred to the cathode. PEM electrolysers can potentially be designed for operating pressures up to several hundred bars, and are best suited for both stationary and mobile applications [19]. The major advantages of PEM over alkaline electrolysers are the higher turndown ratio, e.g. the operating ratio of part load to full load, the increased safety due to the absence of KOH electrolyte, a more compact design due to higher densities, and higher operating pressures (no need for further compression). With respect to hydrogen storage section of a complete hydrogen-based power system to be installed in autonomous islands, the most suitable hydrogen storage method is gaseous hydrogen storage under pressure. Compressed hydrogen (CGH2) storage is a commercially available hydrogen storage technology. Since hydrogen has a low energy density, it must be compressed to very high pressures to store a sufficient amount of hydrogen, particularly for mobile applications. Moreover, hydrogen cannot be considered as an ideal gas for pressures above 15 MPa [20].

Higher storage pressure results in higher capital and operating costs. Industry standards for compressed hydrogen storage are currently set at 35 MPa, with a future target of 70 MPa. High-strength, carbon-fibre composite pressure vessels rated to 70 MPa can achieve a gravimetric



15

storage density of 6 wt% and a volumetric storage density of 30 kg/m3. However, they require the use of expensive materials or composites to achieve a set of different targets including, minimal hydrogen leakage using gas diffusion barriers (polymer liner), maximum mechanical strength (carbon composite) and high impact resistance (foams) [17].

Regarding the re-electrification section, where previously stored hydrogen is used in fuel cells to meet the electricity demand of the island, PEM fuel cells are considered as the most technically viable solution. Their main advantage compared to other fuel cell categories is that they operate on pure hydrogen and not on hydrocarbons (natural gas), which are internally reformed in other fuel cell types. A typical PEM fuel cell assembly includes the polymeric proton exchange membrane, on the opposite sides of which two porous electrocatalytic layers (electrodes) are suppressed. Two conductive and porous collectors are layered over the electrodes in close contact to the hard-plate interconnects, which form the reactants and products flow channels. The proton exchange membrane consists of perfluorosulfonic acid polymers. These materials are gas-tight electrical insulators, in which the ionic transport is highly dependent on the bound and free water in the polymer structure. Nafion is the most commonly used material [17].

With operation voltages 0.7–0.75 V, the maximum efficiency of PEMFCs can be as high as 64%. In today’s applications, certain losses and ancillary equipment decrease efficiency, resulting in a situation in which PEMFCs are more efficient than internal combustion engines only for operation at partial loads [21;22]. Nevertheless, research has focused on increasing performance of PEM fuel cells by minimizing losses, on reducing cost through the use of non-exotic (and inexpensive) materials and on increasing lifetime of the stack as well.

3.1.4 Desalination systems

Desalination is a natural procedure during which sea or brickish water is cleared from the minerals that cause salinity in order to provide water within acceptable standards for drinking or using it in the industry. The desalination plants use sea-water or water with high level of minerals as stock. Many large industries use desalination for diminishing the water pollution During the last century of the previous millennium, the desalination started to be used in a large industrial scale due to the lack of clear water in many areas of the planet.

Due to the fact that the desalinated water is in fact an industrial product, its price will be significantly than the price of the water that can be found in the nature in the proximity of consumption.

Various desalination methods, summarized below, have been developed to provide water with lower salinity :

• The evaporation

• Electrodeionization

• Reverse Osmosis

• Refrigeration

• Hybrid methods

The Desalination method that will be examined for the purposes of the STORIES project is the Reverse Osmosis (RO) desalination method. RO is well developed and has been in commercial use for three decades for desalting low salinity brackish water. Moreover, this methods provides significant flexibility in adding capacity and can be provided in various sizes of view kW consumption and a few liters of potable water per hour up to some decades of cubic meters and capacity of some hundreds of kWs. The modular nature of this technology and its flexibility makes this technology an ideal candidate for studying in more detail the combination of RES and



16

desalination systems. Moreover, on the island power systems to be studied, desalination plants using this technology have been or are being constructed now.

3.1.4.1 DETAILS ON REVERSE OSMOSIS TECHNOLOGY Formally, reverse osmosis is the process of forcing a solvent from a region of high solute

concentration through a membrane to a region of low solute concentration by applying a pressure in excess of the osmotic pressure.

The membranes used for reverse osmosis have a dense barrier layer in the polymer matrix where most separation occurs. In most cases the membrane is designed to allow only water to pass through this dense layer while preventing the passage of solutes (such as salt ions). This process requires that a high pressure be exerted on the high concentration side of the membrane, usually 2–17 bar (30–250 psi) for fresh and brackish water, and 40–70 bar (600–1000 psi) for seawater, which has around 24 bar (350 psi) natural osmotic pressure which must be overcome.

This process is best known for its use in desalination(removing the salt from sea water to get fresh water), but it has also been used to purify fresh water for medical, industrial and domestic applications since the early 1970s.

When two solutions with different concentrations of a solute are mixed, the total amount of solutes in the two solutions will be equally distributed in the total amount of solvent from the two solutions.

Instead of mixing the two solutions together, they can be put in two compartments where they are separated from each other by a semipermeable membrane. The semipermeable membrane does not allow the solutes to move from one compartment to the other, but allows the solvent to move. Since equilibrium cannot be achieved by the movement of solutes from the compartment with high solute concentration to the one with low solute concentration, it is instead achieved by the movement of the solvent from areas of low solute concentration to areas of high solute concentration. When the solvent moves away from low concentration areas, it causes these areas to become more concentrated. On the other side, when the solvent moves into areas of high concentration, solute concentration will decrease. This process is termed osmosis. The tendency for solvent to flow through the membrane can be expressed as "osmotic pressure", since it is analogous to flow caused by a pressure differential.

In reverse osmosis, in a similar setup as that in osmosis, pressure is applied to the compartment with high concentration. In this case, there are two forces influencing the movement of water: the pressure caused by the difference in solute concentration between the two compartments (the osmotic pressure) and the externally applied pressure.

The various levels of treatment in a typical desalination plant in Croatia[23] are presented in the following diagram of Fig. 3.6.



17

Fig. 3.6 Flow Diagram of a typical RO desalination plant in Croatia

3.1.4.1.1 Economic Parameters The major component of desalinated water production cost in commercial RO plants are the

following: • Capital recovery cost

• Operation &Maintainance (labour, spares, membranes, chemicals, etc)

• Energy cost

The contribution of capital recovery cost varies between 30% and 50% of the cost of water produced, depending on several variables like plant size, site, process type, etc. Energy is usually the major component cost over the useful service life of the RO plants, which usually extends up to 30 years for major plants. The O&M cost ranges between 15% and 30% depending mainly on the plant capacity.

Plant capital and water production costs decrease significantly as plant capacities increase up to about 12000 m3/d for brackish water and 20000m3/d for seawater RO. Beyond these limits, costs decrease only slightly with increasing plant size. Desalination plants require major initial capital outlays that need to be depreciated over plant service life which could be 30 years. Hence, the cost of capital has a significant impact on water production cost since capital contribution is usually 30% to 50% of water production cost.

The source and quality of feed water are an important cost factor. The cost of desalting seawater can be from three to as much as seven times more expensive than brackish water when using RO desalination. The cost and availability of water storage, water transfer and other infrastructure required for building a desalination plant or delivering its water can be a major cost component.



18

Such costs are mainly related to the plant location and size. In locations where no infrastructure exists and the population served is small or dispersed, it may be more cost effective to build smaller plants. The cost of product water will normally be very sensitive to both the distance and elevation of the raw water source relative to the location of the desalination facility. This is due to the energy cost associated with pumping the water from the source to the plant.

In the case of seawater desalting, RO plants require about two times the amount of water they produce. Seawater intake costs include intake pipes or channels, screens, intake basins and seawater pumps with all auxiliary equipment. This cost indicates that the seawater intake is an expensive component in the plant. Therefore, due consideration has to be given o the design and location of the intake as well as to the location of the plant. The main purpose of seawater intake systems is to provide good quality feed water free from seaweed, shells, sand and contamination through hydrocarbons. To avoid seaweed, the intake mouth should preferable be installed more than 15 m below sea level to avoid suction of this matter. this may result in intake pipes of 2km length offshore. Also, to reduce pollution through hydrocarbons the intake should be installed several meters below the lowest sea wave level. Due to increasing sea coast pollution, simple onshore intakes with short channels are becoming less and less possible. Seawater intake is one of the vital components of the plant, so it should have a stand-by capacity. Intake pipes may be installed on the ground of the seabed but also above the sea level with a vacuum-based siphon system. Compared to an open seawater intake, beachwells are preferred if plant location and ground conditions allow their use. The reliability of the continuing trouble-free operation of RO plants is a key consideration related directly to the water production cost. The reliability of any desalination plant is a function of proper operation and good design. Proper operation depends on operators’ experience, skills and training. Good design should compensate for or prevent operation errors through sufficient instrumentation and control. Computer and digital controls allow full plant automation, which is highly desirable to increase reliability of large plants. It is a common practice in large RO plants to expect a load factor of 90% or better. With good design and operation, plant availability may exceed 95% of the time over 1 year. All plants need some shut-down time for routine maintenance, which accounts for the remaining percentage. A common approach to increase plant availability is to have spares or duplicate copies of key items installed or in store ready for use when the operating item fails. This practice will increase the initial capital cost. The best approach to increase plant reliability is through proper process design for the available water at the given site. Pre-treatment of feed water before the brine desalting component is an important consideration in plant design. Pre-treatment usually involves several filters and chemical injections and may require clarifiers, activated carbon or other equipment depending on the chemistry and quality of the feed water. The useful service life of RO desalination is defined as the years over which the plant produces the water quantity and quality it is designed for. Large plants are designed for a service life of 25 to 30 years, while small or mobile plants are usually designed for 10 years. The selection of material and equipment specification can have a great impact on the capital and the operation and maintenance costs required over the life of the plant. The O&M costs such as, labour, membranes, chemicals, spare parts and consumables usually range between 15% to 30% of water production cost depending on plant size, process type and design, etc. Labour costs may be up to 20% for small RO plants. Plant automation will reduce the cost of labour with the additional advantage of increased plant reliability. The annual cost of spare parts is usually about 1% to 2% of the capital cost of the plant, excluding membrane replacement. Membrane life expectancy has continuously been improved since the 1970s. Most membrane manufacturers now provide a 5- year warranty, and the average expected life may exceed 7 years for a well designed and operated plant. Membranes cost about 15% to 20% of the cost of capital for large seawater RO plants, and their replacement about 10% of the water production cost for large RO plants. Chemicals may reach 10% of water production costs. Chemical consumption can greatly be reduced by proper process optimization for any given site.



19

Energy contribution to water production cost can range from about 30% to 50% depending on energy cost, process type and design. The energy input for RO desalination is a function of water salinity and plant design. For producing 1m3 of desalinated water, RO requires approximately 2.8-4.5 kWh of electric energy when energy recovery devices connected to the brine stream are used.

The purpose of the studies performed in Chapter 7 is to study the potential impact of desalination systems to the power systems of the island to be connected when RES are utilized. Moreover, in the same chapter, the impact in the energy cost of water for the various case studies and operation scenarios will be assessed.



20

3.2 Short description of the Island systems selected to be simulated Since we have a separate data collection task please do not put more than 3 pages for each case

study. Avoid overwhelming with photos etc. Please focus on the location of the island in the country, the demand characteristics and the RES characteristics that made the selection of the island interesting for our study.

3.2.1 Milos case study Milos is a Greek island sited on the southwest part of the country and specifically in the group

of islands called Cyclades. Its distance from Athens is 86 nautical miles, i.e. 160 km approximately. This island is 151 square kilometres in area and its coastal line is 125 km. It is considered generally as flat island but semi-arid island. The highest mountains reach the high of 751m and 636m, while the rest have altitude less than 400m. It is a volcanic island with significant geothermal and mineral resources that have played significant role in its development throughout the ages in the Greek civilization. There are almost 5000 people living in the island, but the population rises about 5 times during the summer period because of tourism. However, significant percentage of the population is employed in the mines of the island and therefore Milos is not only a touristic island. Most of the people live in its northern part, which has greater plains where the capital and the harbour of the island is.

Table 3-4 General Data on Milos

Population 4771 Interconnected or non Non Interconnected Minimum active power consumption 1750 kW Maximum active power consumption 9880 kW Mean power factor of the global consumption 42.1% Yearly Consumption 36457MWh

3.2.1.1 THERMAL POWER PLANT DATA Milos is an autonomous power system interconnected with the nearby small island of Kimolos.

On the island, one power station operated on fuel oil exists and belongs to the PPC, the operator of the island. Data on these units are provided in Table 3-5. The cosφ is considered equal to 0.85.

Table 3-5 General Data for the power plant of Milos

Unit ID

Unit Type Nominal capacity of each unit (kVA)

Maximum active power production (kW)

Fuel Type

Minimal active power output in steady-state conditions (kW)

Average specific consumption (g/kWh)

1 SULZER 2.224 1.750 880 0.262 2 SULZER 2.224 1.750 880 0.262 3 MAN 965 700 350 0.248 4 MAN 965 700 350 0.248 5 MAN 965 700

Heavy Oil

350 0.248 6 CKD 2.600 2.000 1000 0.233 7 CKD 2.600 1.900 1000 0.231 8 FINCANTIERI 2.217 1.750

Diesel 850 0.241

Total Capacity 14.760 11.250



21

In order to meet peak demand during summer without affecting significantly the reliability of the power system, some units are usually rented during the summer period.

For 2006, the data of these units are provided as described in Table 3-6. All these units use diesel oil. These units should provide determined amount of power, according to the contract and are placed higher in the priority list than rest diesel-oil units.An idea of the maintenance schedule is provided in Table 3-7 and no more than 2 units within 10 minutes are assumed to start up. The average fuel cost is 262.5€/tn for the heavy Oil and 543.2€/tn for the diesel.Table 3-8 provides data on the emissions for each type of fuel used.

Table 3-6 Data on the rented units of Milos during summer period

Unit ID Min active power output in steady-state conditions (kW)

Maximum active power production (kW)

Average specific consumption (g/kWh)

9 550 1032 0.225 10 550 1032 0.225 11 550 1032 0.225 12 550 1032 0.225 13 850 1600 0.228

Table 3-7 Data on Maintenance of the units in Milos

Sulzer Units MAN Units Unit 6 Unit 7 Unit 8 May, June November, December February-October January May

Table 3-8 Data on emissions per fuel type-Milos

Average Emissions (Kg/Tn) Heavy Oil Diesel

Particles 1.86 1.19

SO2 57.12 0.80

NOX 11.4 4.67

CO2 3200 2445

3.2.1.2 RES DATA

3.2.1.2.1 Wind power Additionally to the existing wind park of 2050kW, there are plans for the addition of a 850kW

combined with a reverse osmosis desalination plant. Moreover, on the island there is significant geothermal potential for which there are exploitation plans for installation of a 120MWe utilizing high enthalpy geothermal energy, if Milos is interconnected to the nearby islands or to the mainland.

On Milos one wind park has been installed since 2001, consisting of two Vestas V-44 wind turbines and total installed capacity of 1200kW. One Vestas V-52 wind turbine of 850kW installed capacity was added at the same site and the total wind power capacity reached 2050kW. A summary of the wind power contribution on the island is shown in Table 3-9. Currently another wind turbine of 850kW was installed which is the one that will be used for our scenarios together with desalination plant. The usual practice of many operators in island systems is to impose a limit of 30-40% penetration of wind power production.



22

Table 3-9 Data Description for wind power on Milos on 2006 (year of study)

Number of wind parks 1

Wind parks nominal power. 2050kW

Annual production 4.980MWh.

Number of wind turbines at each wind park 3

Installed capacity of the wind turbines 2*600kW + 1*850kW

Diameter of rotor etc (m) 44/52

Hub-height (m) 45

Meteorological data height and place if different than the location of the park. Kythnos island. For each type of wind turbine, the machine’s characteristic curve that gives the power output as a function of wind speed is necessary.

3.2.1.2.2 Other RES issues The expected amount of installed capacity for Milos island in the future, concerning PV

installations is 1370 kW according to the applications to the Regulatory Authority of Greece (RAE) currently granted to the complex Milos-Kimolos. 1274KW will be installed on Milos while 91.4kW will be installed on Kimolos.

There was one installation of Geothermal energy with capacity 2MWe. This installation ceased its operation a few years ago due to environmental constraints.

3.2.1.3 ESTIMATION OF WATER NEEDS On Milos, the annual average rainfall is not high, resulting in lack of water in the island. The

problem becomes even bigger through summer that is the dryer period of the year and the population increases. A summary of the water demand data is given in Table 3-10

One solution to the problem was given by transferring water from the mainland of Greece. The transportation of water through water ships began in 2001 and lasted for two months (July and August of 2001) transferring water in the island from the mainland and the Athens Water Company. Since then significant quantities of water have been have been shipped to the island as Table 3-11 describes.

Table 3-10 Data Description on water for Milos

Permanent Population on the island 4771 One year water consumption at monthly or seasonal levels on the island Table 3-11

Water transportation per month/season Table 3-12

Cost of transported water 8 €/m3

Table 3-11 Quantity of transported water to Milos

2002 (*1000 m3) 2003 (*1000 m3) 2004 (*1000 m3) 1st quarter 39 28 18

2nd quarter 44 52 46

3rd quarter 59 67 56

4th quarter 45 36 38

Total 187 183 158



23

The only available data were the transferred water to the island from the mainland which is provided in the following table. It can be considered that the needs are the maximum per season transferred water to the island multiplied by a factor of 2, as Table 3-12 describes. The amount of water that is considered produced at 20 minutes interval for each month is provided in Fig. 3.7

Table 3-12 Estimation of Water consumption in Milos

Total Water consumption (*1000 m3) 1st quarter 78

2nd quarter 104

3rd quarter 134

4th quarter 90

Total 406

6

9

12

15

18

21

24

0 1 2 3 4 5 6 7 8 9 10 11 12

Months

Cub

ic m

eter

s (m

3 )

Fig. 3.7 Water demand at 20 minutes interval

3.2.2 Ios Case study

3.2.2.1 GENERAL DESCRIPTION Ios is an island in the Cyclades group in the Aegean Sea as shown below. Ios is a hilly island

with cliffs down to the sea on most sides, situated halfway between Naxos and Santorini. It is about 18km long and 10km wide, with an area of about 109 km². Population was 1,838 in 2001 (down from 3,500 in the 19th century).

The Port of Ios is at the head of the Ormos harbor in the northwest. From there the bus or 15-minute walk up the steep donkey path takes you to the village, known as Chora. This is a white and very picturesque cycladic village, full of stairs and narrow walks, that makes it inaccessible for cars of any kind. Today, the main path through this village is completely taken over by tourism in terms of restaurants, boutiques, bars and discotheques. Apart from the port and the village of Chora, Ios has only a few small settlements, just a group of spread out houses in the background of major beaches (Theodoti, Kalamos, Manganari). Since the 1990s the island mayor Pousseos has worked on Ios development towards attracting different types of tourists. With money from the European Community some roads have been built, all of them paved, and a very scenic amphitheater has been created by the German architect Peter Haupt (who died in 2003) on the very top of the village hill. Unfortunately, cultural events rarely take place up there.



24

Ios attracts very large numbers of young tourists, many of whom used to sleep on their sleeping bags during the 1970s on the popular beach of Mylopotas after partying through the night. Today Mylopotas beach has been developed to a mass package tourism resort.

Geography

http://en.wikipedia.org/wiki/Image:GR_Ios.PNG

Coordinates: 36°43′N 25°20′E

Island Chain: Cyclades

Area: 109.024 km²

Government Periphery: South Aegean

Prefecture: Cyclades

Capital: Chora

Statistics Population: 1,838 (2001)

Density: 17 /km²

Website www.ios.gr



25

Fig. 3.8 Map of Ios

3.2.2.2 THE ELECTRICAL SYSTEM Ios is part of Paro-Naxia autonomous system, which includes Paros, Antiparos, Folegandros,

Ios, Irakleia, Koufonisi, Naxos, Sikinos and Sxoinousa. It is located in the central part of the Southern Aegean Sea. This is one of the largest insular power system in Aegean Sea, with high wind potential and several existing water reservoirs which are currently used for irrigation and in parallel can be exploited for a WHPS.

The power station of this system with 10 Internal Combustion (IC) power units of cumulative capacity 61.4MW, is located in Paros. The interconnections between the islands, all in Medium Voltage (MV) 20/15kV are shown in Fig. 3.9. Details on the characteristics of each interconnection is shown in Table 3-13 while their electrical characteristics are provided in Table 3-14

Table 3-13 Underwater connections between islands in the system of Paro-Naxia

Connection Length (km)

Number of cables

Type of Cable

Capacity of each cable (MVA)

3 3X150 AL 7,8 Paros – Naxos 7,5

2 3X95 Cu 9,1

Paros – Antiparos 1,9 4 1Χ50 AL 5,3

Naxos – Koufonisi 6,2 1 3Χ35 Cu 4,9

Koufonisi – Sxoinousa 9,2 1 3Χ35 Cu 4,9

Sxoinousa – Irakleia 4,6 1 3Χ35 Cu 4,9

Ios – Sikinos 10,3 2 3Χ35 AL 3,8

Sikinos – Folegandros 18,5 2 3Χ35 AL 3,8

Naxos – Irakleia 8,8 1 3Χ35 Cu 4,9

Paros – Ios 25 2 3Χ95 Cu 9,1



26

Paros has an area of 196.3 km², 12,853 inhabitants and 724m maximum altitude (Mountain Marpissa). Naxos is the largest island in the Cyclades with an area of 429.8 km², 18,188 inhabitants and 1004 maximum altitude. Ios, it the third in order and in size island with 1,838 inhabitants and an area of 109km2. Folegandros has a population of 667 inhabitants and an area pf 32.2km2. Finally, Sikinos has 238 inhabitants, area of 42.5 km2, and maximum altitude Mountain Troulos with 552 m.

Fig. 3.9 The local power system of Paro-Naxia (interconnections between islands) and wind

potential [24] (CRES 2001).

Table 3-14 Underwater connections between islands in the system of Paro-Naxia (source: PPC)

Undersea cables 95Cu 35Cu 35Al R (Ω/km) 0,215 0,617 1,035

X (Ω/km) 0,626 0,115 0,115

Imax (A) 300 190 145

Vn (kV) 15 15 15

Smax (kVA) 7.785 4.931 3.763

Maximum permitted continuous load (Α) 300 185 145

Maximum permitted continuous load (kVΑ) 7.785 4.800 3.763

A summary of the demand for this power system for 2005 is shown in Table 3-15. The annual rate of increase is estimated to 12%.

The Estimated energy demand in Paros and Ios islansd for 2006 are provided in Table 3-16.



27

Table 3-15 Data from the local power station of Paros (PPC, 2005)

Population Peak demand (kW)

Annual electrical energy demand (GWh)

Electricity production cost (€/kWh)

Fuel cost (€/kWh)

Share of FC to the EPC (%)

Load factor (%)

48397 56000 179 0,1306 0,08 63% 37%

Table 3-16 Data from the local power station of Paros (PPC, 2005)

Annual electrical energy demand (GWh) Peak demand (ΜW) Paros 84,4 26,3

Ios 12,6 3,9

3.2.2.2.1 Load demand and wind data In Fig. 3.10-Fig. 3.13, the load demand and the wind data distribution over the year are

presented respectively. The annual energy demand in power system of Paros is estimated for 2010 to 246.3GWh, the peak demand 74.8MW and the load factor 37.6%. The peak demand of each month is provided in Table 3-17

Table 3-17 Peak demand per month in the Paro-Naxia power system (MW) (source: PPC)

MAX 2006 61.2 MW

jan feb mar apr may jun 30 30.6 27.5 37.8 30.1 40

jul aug sep oct nov dec 49.8 61.2 39.5 27 28.3 28.8



28

Power system of Paros - Annual mean load:28.1MW

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1 8760hours

Load

dem

and

(MW

)


0

10

20

30

40

50

60

1 5 9 13 17 21 25 29 33 37 41 45 49 53weeks

aver

age

load

dem

and

per

wee

k (M

W)


0.05.0

10.015.020.025.030.035.040.0

Janu

ray

Febr

uary

Mar

ch

April

May

June

July

Augu

st

Sept

embe

r

Oct

ober

Nov

embe

r

Dec

embe

r

months

aver

age

load

dem

and

per m

onth

(MW

)

Fig. 3.10 Electricity demand in power system of Paros: a) Hourly time series data, b) average load

demand per week, c) average load demand per month.



29

Power system of Paros - Annual average wind speed: 9.1m/s

02468

1012

Janu

ray

Febr

uary

Mar

ch

April

May

June July

Augu

st

Sept

embe

r

Oct

ober

Nov

embe

r

Dec

embe

r

months

aver

age

win

d sp

eed

(m/s

)


02468

1012141618202224

1 8760hours

win

d sp

eed

(m/s

)


02

46

810

1214

1 5 9 13 17 21 25 29 33 37 41 45 49 53weeks

aver

age

win

d sp

eed

(m/s

)

Fig. 3.11 Wind data in Paros: a) Hourly time series data, b) average wind speed per week, c)

average wind speed per month.



30

Power system of Paros - Annual wind Capacity Factor:47.1%

0%

20%

40%

60%

80%

100%

120%

1 8760hours

win

d pr

oduc

tion

(MW

pr

oduc

ed /M

W in

stal

led)


0%

20%

40%

60%

80%

100%

1 5 9 13 17 21 25 29 33 37 41 45 49 53weeks

aver

age

win

d pr

oduc

tion

(MW

pro

duce

d /M

W

inst

alle

d)


0%10%20%30%40%50%60%70%80%

Janu

ray

Febr

uary

Mar

ch

April

May

June

July

Augu

st

Sept

embe

r

Oct

ober

Nov

embe

r

Dec

embe

r

months

aver

age

win

d pr

oduc

tion

(MW

pro

duce

d /M

W

inst

alle

d)

Fig. 3.12 Wind production distribution in Paros: a) Hourly time series data, b) average wind speed

per week, c) average wind speed per month.



31


02468

1012141618202224

1 8760hours

win

d sp

eed

(m/s

)


0%

20%

40%

60%

80%

100%

120%

1 8760hours

win

d pr

oduc

tion

(MW

pr

oduc

ed /M

W

inst

alle

d)Power system of Paros - Annual mean load:28.1MW

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

1 8760hours

Load

dem

and

(MW

)

Fig. 3.13 Duration curves of a) annual demand, b) wind speed, c) wind production, in power system

of Paros.

3.2.2.2.2 Thermal power station Data Table 3-19 presents some details on the units of the thermal power station according to their

priority order in unit commitment. In the same table the energy produced by each unit is provided. Additionally to these 10 units, one rented unit also procides energy during the summer months July-August of 359MWh.Table 3-18 provides a summary of the produced energy by the thermal units of the island. 38127tn of heavy Fuel Oil (HFO) and 523tn of Diesel have been consumed for providing this amount of energy for each year.

Table 3-18 Energy produced by the thermal power station (MWh) (source: PPC)

Total 189560MWh

jan feb mar apr may jun 13629 12064 12286 12691 13885 17641

jul aug sep oct nov dec 22473 28758 16927 12904 12530 13853



32

Table 3-19 The installed Internal Combustion Engines on the power system of Paro-Naxia according to their priority-order.

Priority order Unit Provider Rated power

(kW) Maximum capacity (kW)

Fuel (HFO- Heavy fuel oil or diesel)

Technical minimum (kW)

Average Specific consumption (g/kWh)

Energy Produced (MWh)

1 FINCANTIERI 10720 10720 HFO 5360 208.5 25547 2 FIAT-G.M.T 3920 3700 HFO 1960 212.4 2634 3 FIAT-G.M.T 3920 3700 HFO 1960 212.4 10073 4 WARTSILA 10360 10360 HFO 5180 203.6 46275 5 WARTSILA 10360 10360 HFO 5180 203.6 42551 6 H.S.D. 11000 11000 HFO 8000 199.5 49040

7 SULZER 6300 6000 HFO 3150 184.5 12559

8 -SULZER 2900 2600 HFO 1450 228.8 95

9 SULZER 3104 3100 HFO 1552 228.8 408

10 GENERAL ELECTRIC 11700 11700 Diesel 5850 570

100



33

3.2.2.3 RES POTENTIAL - CURRENT DEVELOPMENT OF RES IN PARO-NAXIA POWER SUPPLY SYSTEM

The wind potential on this complex is significant as Fig. 3.14 shows. The technically exploitable potential is shown in Fig. 3.15. However, the current development is not so high as one would expect. The total installed wind power capacity is 2.46MW, while PPC constructs a new wind park of 3MW. There is also a samll PV plant on the roof of an hotel in Paros of 10kW. Fig. 3.16 shows the current development of RES in Paro-Naxia power system.

Fig. 3.14 Wind potential map of Paro-Naxia complex



34

Fig. 3.15 Technically and economically available wind potential map of Paro-Naxia complex

Fig. 3.16 Current development of RES in Paro-Naxia power supply

Current development in Paros: • There is only one old wind turbine in operation with rated capacity 110kW belonging to

the National Telecommunications Company,. OTE S.A [25].

• There is one PV on the roof of an hotel of 10kW [26]

• Under construction a 3MW park on the island of Paros, Kamares area[27].



35

• The PV capacity that has been granted authorization on the island but not yet installed is 3021kW.

Current development in Naxos: • Wind park of 1.2MW (2 600kW wind turbines near the port)

• There are several applications for photovoltaics with total capacity that has been granted authorization 2986kW.

• Six applications for wind farms (one with negative response, one with production license and the rests under the evaluation procedure). The total capacity granted authorization on Naxos is 7.56MW.

Current development in Ios: • There are two wind turbines in operation (0.6MW – Purgos – with operation license,

0.56MW – Pelekania - with operation license)

• There are also applications for photovoltaics (99.36kW in Almiros and 149.04kW in Pelekania, both of them under the evaluation procedure. The total granted authorization capacity on Ios is 504kW.

Current development on the rest of the islands of the complex • Currently a small PV of 10kW exists on Folegandros island.

• PV granted authorization on these islands are summarized in Table 3-20 :

Table 3-20 The Authorized capacity by the Regulatory Authority of Greece on the rest Paro-Naxia islands

Island Authorized PV capacity (kW) Island Authorized PV capacity (kW)

Antiparos 241.6 Iraklia 21.3 Koufonisi 78.2 Sikinos 42.6 Schinousa 49.7 Folegandros 120.8



36

3.2.3 Cyprus case study

Cyprus is the third largest island in the Mediterranean, after Sicily and Sardinia, with an area of 9.251 sq. kms (3.572 sq. miles). It is situated at the north-eastern corner of the Mediterranean, at a distance of 300 km north of Egypt, 90 km west of Syria, and 60 km south of Turkey. Greece lies 360 km to the north-west (Rhodes -Karpathos). Cyprus lies at a latitude of 34°33' - 35°34' North and longitude 32°16'-34°37' East. Fig. 3.17 provides a map and the location of Cyprus in east Mediterranean.

Fig. 3.17 Location of Cyprus in East Mediterranean and map of the island

Cyprus is an isolated island not interconnected in any way with other countries. The country has two mountain ranges: the Pentadaktylos range which runs along almost the entire northern coast, and the Troodos massif in the central and south-western parts of the island. Cyprus' coastal line is indented and rocky in the north with long sandy beaches in the south. The north coastal plain, covered with olive and carob trees, is backed by the steep and narrow Pentadaktylos mountain range of limestone, rising to a height of 1.042 m. In the south, the extensive mountain massif of Troodos, covered with pine, dwarf oak, cypress and cedar, culminates in the peak of Mount Olympus, 1.953 m. above sea level.

The population of Cyprus is estimated (Dec. 2006) at 867.600 of whom 660.600 belong to the Greek Cypriot community, (76,1%), 88.900 (10,2%) to the Turkish Cypriot community and 118.100 (13,7%) are foreigners residing in Cyprus. The capital of the island is Nicosia with a population of 228.400 in the sector controlled by the Cyprus government. It is situated roughly in the centre of the island and is the seat of government as well as the main business centre. It is the only divided capital in the world due to Turkey’s military occupation of part of Cyprus. The second largest town is Limassol, on the south coast, with a population of around 180.100 inhabitants. Since 1974 it has become the island’s chief port, an industrial centre and an important tourist resort. Larnaka, also on the south coast of the island, has a population of 80.400 and is the country’s second commercial port and an important tourist resort. The Larnaka International Airport is located to the south of the city. Finally, Pafos, on the south-west coast, with a population of 54.000, is a fast developing tourist resort, home to the island’s second international airport and an attractive fishing harbour. The towns of Famagusta, Kyrenia and Morfou as well as part of Nicosia, have been under military occupation since the Turkish invasion of 1974. The Greek Cypriot inhabitants of these towns were forced to flee to the government controlled area. In their place the Turkish authorities installed settlers, mostly from Anatolia, Turkey.

The economy of Cyprus is dominated by the services sector, including the public sector, trade, tourism and education, with smaller agriculture and light manufacturing sectors. Cyprus has been sought as a base for several offshore businesses for its highly developed infrastructure. Economic policy of the Cyprus government has focused on meeting the criteria for admission to the European Union. The island has witnessed a massive growth in tourism over the years and as such the property rental market in Cyprus has grown along side. Added to this is the capital



37

growth in property that has been created from the demand of incoming investors and property buyers to the island.

Cyprus has an open, free-market, services-based economy with some light manufacturing. Internationally, Cyprus promotes its geographical location as a "bridge" between three continents, along with its educated English-speaking population, moderate local costs, good airline connections, and telecommunications.

In the past 20 years, the economy has shifted from agriculture to light manufacturing and services. Currently, agriculture makes up only 3.1% of the GDP and employs 8.5% of the labor force. Industry and construction contribute 18.6% and employ 20.5% of the labor force. The services sector, including tourism, contributes 78.3% to the GDP and employs 71.0% of the labor force. In recent years, the services sector, and financial services in particular, have provided the main impetus for growth, while the traditional growth sector, tourism, has leveled off. Manufactured goods account for 58.5% of domestic exports, while potatoes and citrus constitute the principal export crops. The island has few proven natural resources. Trade is vital to the Cypriot economy and most goods are imported. Cyprus has to import fuels, food, most raw materials, heavy machinery, and transportation equipment. More than 67% of its imports come from the European Union, particularly Greece, Italy and the United Kingdom, while 1.2% come from the United States.

In the run-up to EU accession (May 1, 2004), Cyprus dismantled most investment restrictions, attracting increased flows of Foreign Direct Investment (FDI), particularly from the EU. Cyprus has good business and financial services, modern telecommunications, an educated labor force, good airline connections, a sound legal system, and a low crime rate. Cyprus' geographic location, tax incentives and modern infrastructure also make it a natural hub for companies looking to do business with the Middle East, Eastern Europe, the former Soviet Union, the European Union, and North Africa. As a result, Cyprus has developed into an important regional and international business center. Non-EU investors (both natural and legal persons) may now invest freely in Cyprus in most sectors, either directly or indirectly (including all types of portfolio investment in the Cyprus Stock Exchange). The only exceptions concern primarily the acquisition of property and, to a lesser extent, restrictions on investment in the sectors of tertiary education, banking, and mass media.

3.2.3.1 ELECTRICITY DATA There are 3 power stations, on the island (Vasilikos, Dekheleia and Moni) operated by the

Electricity Authority of Cyprus (EAC). Details on these units are provided in Table 3-21. Table 3-21 Thermal units on Cyprus

Unit Type Installed Capacity

(MW)

Specific Consum

ption (g/kWh)

Fuel Cost (€/tn)

CO2 emissions (g/kWh)

NOx (g/kWh)

SOx (g/kWh)

PM-10 (g/kWh)

Vasilikos STEAM 3*130 0.227 195.6 687.6 1.41 1.94 0.14 Vasilikos GT 1*38 0.365 409.98 985 2.11 1.27 0.14 Dekheleia Steam 6*60 0.285 171.78 861 1.80 8.29 0.104 Moni Steam 6*30 0.366 176.93 1,092 2.28 11.36 0.21 Moni Gas 3*38 0.393 399.51 1,032 2.16 1.30 0.18



38

The year under study was 2005 during which the peak load was 850MW and the total demand was 4.4TWh.

3.2.3.2 RES DATA The currently installed capacity on the island is rather low. Despite the fact that Cyprus presents

the largest per capita installed capacity of solar water heaters on Cyprus, the installed PV capacity only recently started to increase and reaches now 1.7MW producing 1.6GWh during 2008. There are also plants that use biomass/biogas with installed capacity of 3.3MW producing 7.8GWh during 2008. The wind resource is not as favorable as for instance on the Gree4k islands but there are few area where wind power casn be installed. During 2005 the authorized capacity by the Regulatory Authority of Cyprus (CERA) is 289.7MW.

The average pattern of wind power resource on the island at 10 m, as used for the simulations, is shown in Fig. 3.18.

0

1

2

3

4

5

6

7

8

9

10

00.

5 11.

5 2

2.5 3

3.5 4

4.5 5

5.5 6

6.5 7

7.5 8

8.5 9

9.5 10

10.5 11

11.5 12

12.5 13

13.5

Wind velocity (m/s)

(%)

Fig. 3.18 Annual wind speed pattern for 10m used for simulations in Cyprus

3.2.3.3 ESTIMATION OF WATER NEEDS Cyprus does not have plenty of potable water from natural resources. The water demand is met

by water dams, springs, Desalination plants, recycling and reuse of water for gardening purposes and imports water from abroad, mainly Greece.

The total annual water demand all over Cyprus for the year 2000 is estimated to be 265.9 million m3. Its distribution over the various sources is shown in Fig. 3.19 and Table 3-22.



39

Anticipated source of supply - Year 2000

Groundwater47.9%

Surface water38.2%

Desalination12.6%

Springs1.3%

Fig. 3.19 Distribution of Total Water Demand amongst various sources of supply for year 2000 in

Cyprus

Evidently, groundwater remains the main, most secure and low-cost source of water for the agricultural sector and in areas outside the government-owned irrigation schemes. Thus, the groundwater resources in Cyprus are overexploited by about 40% of sustainable extraction. What is more, the risk to agriculture has increased because of depleted aquifers. Unless groundwater is left to recover to a reasonable level, the resource will be of limited help to mitigate future water shortages.

The aggravated water scarcity in the 1990’s and the deficient situation of water supply for domestic uses, including the economically important tourist sector, caused state intervention to ensure a stable, continued supply of good quality drinking water through seawater desalination. The measure was successful in providing to the domestic sector, including the economically important tourist industry, a steady supply. As a result, the main source for domestic water supply is desalination, which is equivalent to 12.6% of the total annual water demand all over Cyprus for the year 2000 (Fig. 3.20and Table 3-22). This fact brings to the foreground seawater desalination as a reliable source of good quality drinking water.

Table 3-22 Water demand by sector and anticipated source of supply for the year 2000.

Surface water Groundwater Springs Desalination TOTAL (106) m3

% (106) m3

% (106) m3

% (106) m3 % (106) m3 %

Agric 82 43 100.4 57 - - - - 182.4 68.6

Domes 14.5 21.6 16 23.1 3.5 5.2 33.5 50 67.5 25.4

Indus - - 3.5 100 - - - - 3.5 1.3

Envir. 5 42 7.5 58 - - - - 12.5 4.7

TOTAL 101.5 127.4 3.5 33.5 265.9 100

% 38.2 47.9 1.3 12.6 100.0



40

37.8%

30.8%

6.0%5.5%

12.6%1.3%

1.3% 2.8%1.9%0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

Wat

er D

eman

d (%

of t

he to

tal

annu

al d

eman

d)

Agriculture Domestic Industry Environment

Year 2000

Groundwater Surface water Desalination Springs

Fig. 3.20 Water demand by sector and anticipated sources used for the year 2000 [28]

3.2.4 La Graciosa case study La Graciosa is a small volcanic island located to the north of the island of Lanzarote, separated

from this island by 1 km wide channel of sea water, that the locals call “El Río”. Today, according to the Spanish National Statistics Institute (INE, 2006), La Graciosa is inhabited by 658 people. The average yearly growth is 2,3 %. The population density is 21,93 hab./km².

The island is part of the Natural Park of the Chinijo Archipelago, which includes, together with La Graciosa other two small islands, La Alegranza (1,2 km²) y Montaña Clara, El Roque del Oeste (0,06 km²), y El Roque del Este (0,71 km²), as well as the maritime area encircled by the small island and the northern Lanzarote coast (total of 46.263 hectares).

La Graciosa is an idyllic place, with exceptional natural conditions. The island has been declared by UNESCO, Biosphere Reserve (Oct 7, 1993). The objective is to make the social an economic development of the island compatible with the protection of the natural environment and the preservation of biodiversity and its fragile ecosystems. The Regional Government has approved regulations to protect most of La Graciosa’s territory, through its legal consideration as regional Natural Park (Decree 89/1986, may 9).

The economy is mainly based on fishing and tourism, receiving numerous tourists all year around, and has a small port in the village of La Caleta del Sebo that communicates the island through regular boats to the small fishing port of Orzola in northern Lanzarote. According to the last census, the tourist rental apartment, pensions, and houses which are normally occupied by tourist, offer more than 400 beds, with a tourist occupation rate of 90 % during summer and Easter holiday.

Agriculture is marginal, with 45 very small plots close to the Pedro Barba mountain. Subsistence agricultural activities where a small quantity of onion, corn, garlic and potatoes are grown.

In spite of tourism, fishing remains the main means of subsistence for the families in La Graciosa. There are 58 fishing boats that capture approximately 120 ton of fish a year, which are disembarked in the Caleta de Sebo port).



41

La Graciosa has only two small fishing villages, Caleta del Sebo (the capital) and Casas de

Pedro Barba (north of La Graciosa). The population of the island lives mainly in Caleta del Sebo, while Casas de Pedro Barba is without any population most of the year, and just inhabited during the summer and other longs holiday vacations

The Caleta del Cebo residential area is restricted to 20 hectares, with maximum housing capacity for 1,000 residents (Decree 95/2000, May 22, Partial Revision of the Insular Territorial Organization Plan). A stretch of land of approximately 1.300 m long, by 150 m wide that currently has 325 houses, 128 of them inhabited with permanent residents all year. The houses are all painted in white, with doors and windows painted either green or blue. The other village, Pedro Barba, has a surface of 4 hectares and a total of 120 people resident capacity according to the Revision of the Insular Territorial Organization Plan. Currently it has17 houses, none of which is inhabited with permanent residents all year.

3.2.4.1 ELECTRICITY ON THE ISLAND Electricity is supplied from Lanzarote trough a submarine cable. The cable was installed in

1985. It runs from the town of Ye in Lanzarote, down the cliff at Famara to the power transform substation located at Las Salinas del Río, from where the submarine cable runs through “El Río”, the water channel, to La Graciosa, to supply power to the village of Caleta del Sebo. The village of Pedro Barba is supplied of electricity by means of small gen-sets, and photovoltaic panels

The transmission line has a capacity of 1.030 kW, and the maximum electricity demand is currently of 668 kW, with a capacity factor of 41 %. The current power transform station has enough capacity for current demand. The minimum power demand is 204 kW.

In La Graciosa there is a diesel generator, which is available for any possible power emergency. Although there has not been an important increase in the number of residents in La Graciosa in

past years, the island has been experiencing an increase in the construction of new houses and tourist apartments, which is increasing the electrical demand. In 2008, with 666 kW of peak demand is still far from the maximum grid capacity. The yearly demand is shown in Fig. 3.21 and some more statistically details are provided in Fig. 3.22.



42

J a n F e b M a r A p r M a y J u n J u l A u g S e p O c t N o v D e c3 0 0

3 5 0

4 0 0

4 5 0

5 0 0Po

wer

(kW

)

Fig. 3.21 Yearly demand pattern for La Graciosa

200 300 400 500 600 70

2

4

6

8

10

12

14AC Primary Load PDF

Value (kW)Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann

0

100

200

300

400

500

600

700Graciosa Load Monthly Averages

Month

maxdaily highmeandaily lowmin

Fig. 3.22 Details on annual electricity demand of La Graciosa

3.2.4.2 RES POTENTIAL DATA

3.2.4.2.1 Solar radiation Solar radiation conditions in La Graciosa are excelent, with an average of 4.9 kWh/m2-day, and

hourly maximums in July of 1.1 kW/m2

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann0.0

0.2

0.4

0.6

0.8

1.0

1.2

(/

)

Global Solar Radiation Monthly Averages

Month

maxdaily highmeandaily lowmin

Fig. 3.23 Details on Solar Radiation on La Graciosa

kWh/m2-day January 3.253 February 3.731 March 4.664 April 5.283 May 5.925 June 6.129 July 6.551 August 6.240 Sept 5.764 Oct 4.287 Nov 3.387 Dic 2.983 Average 4.856



43

3.2.4.2.2 Wind Conditions Wind conditions in La Graciosa are determined by the Trade Winds that blow constantly from

the north east. Annual average speed is of 5.7 m/s.

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec0

2

4

6

8

10

12

14Wind Speed

0 2 4 6 8 10 12 14 16 180

2

4

6

8

10F

(%)

Wind Speed PDF

Value (m/s)Wind speed data Best-f it Weibull (k=2.33, c=6.36 m/s)

Average wind speed (m/s)

Jan 6.4

Feb 4.4

Mar 6.7

Apr 6.0

May 5.8

Jun 5.9

Jul 6.3

Ago 5.5

Sep 5.1

Oct 5.0

Nov 5.3

Dic 5.5

Annual av.

5.7



44

3.2.5 Corvo case study

The island of Corvo is one of the 9 islands in the Azores archipelago and together with S. Miguel and Flores form the Western Group. It is the smallest island on the archipelago with an area of 17 km2 and is situated at 31° 5′ West longitude and 39° 40′ North latitude. The island is an inactive volcano named Monte Grosso and its crater, a lake, is the island highest point, circa 720 m. There is only one settlement, Vila Nova do Corvo, where some 400 people reside.

Fig. 3.24 Corvo island crater



45

Fig. 3.25 A map of Corvo

The supply of electricity to an isolated small island such as Corvo is very limited, while there is

a great concern with environmental issues related to fossil fuel supply, such as water and land contamination and pollution by oil products and wastes through leakage during shipping handling and storage. The incidence of small-scale oil spills during loading that occurs adjacent to the storage facilities is very common in Corvo. The fuel cost in Corvo is the highest of the entire archipelago, nearly 5 times more than average in Azores. On Corvo Island, the security of supply is a real and frequent concern, since due to bad weather conditions it is common to have oil shortages in this island. To reduce Corvo's dependency and secure supply, the implementation of an energy system that combines RES and storage is the best solution.

The island’s demand of approximately 1086 MWh and peak of 199kW is covered by two sets of 120kW and two of 160kW, i.e. 560kW. Based on these figures, it is evident that the four generators are never operating simultaneously. Thus, the demand is met by two generators one of each group.

3.2.6 San Pietro case study

San Pietro Island is an island approximately 7 km off the South western Coast of Sardinia, Italy,

facing the Sulcis peninsula. With 51 km² it is the sixth Italian island by surface. The population is mostly concentrated in the town of Carloforte, the only municipality in the island, counting 6629 inhabitants (2547 families) with a peak of 25,000 residents during summer.

The island is of volcanic origin. The 18 km of its coasts are mostly rocky; the western and

northern part includes some natural grottoes and harbours with a few small beaches. The eastern coast, on which the port of Carloforte lies, is instead low and sandy.

Off the north-western coast are two small islands, the Isola dei Ratti and Isola Piana. The latter includes the remains of one of the largest tonnara in Italy, now turned into a tourist resort. The island has no rivers or streams, but features numerous ponds and marshes. The interior is hilly, the highest points being the Bricco Guardia dei Mori (211 m) and Bricco Tortoriso (208 m).



46

Its coastline is a series of ragged and sandy areas. The backland hills are covered with scrub interrupted here and there by cultivated fields and pine-woods. The vegetation is that typical of the Mediterranean coast, with Cistus, Mastic, Strawberry Tree, Juniper, Aleppo Pine and Holm Oak. Cultivation, held especially in the eastern and more protected region, includes grape (Vitis vinifera).

The island has a very long history and has had many names: the Phoenicians called it "Inosim, the Greek "Hioera" or "Nesos"; the Romans "Accipitrum Insula”. Its present name is derived from a legend which tells the story about the apostle Peter who once landed on its shores during his journey to Rome. Its recent history is just as rich and unique. This rocky and impracticable island remained uninhabited for centuries until 1736. Charles Emmanuel III of Savoy then conferred it to the descendants of the Ligurian families that had been forced to settle in Tunisia during the XVI century. That is why San Pietro has Ligurian and Arab tastes and characteristics which can still be seen in the colours of the houses in Carloforte (named after the sovereign) and in its tastes (the cascà, derived from the Tunisian cuscus, and the Ligurian specialities, such as the panisse, the farinate and the buridde) as well as in its sounds (the inhabitants still speak with a strong Ligurian accent).

3.2.6.1 ELECTRICITY DEMAND The annual consumption of San Pietro is 14.544 MWh/y of which 6370 MWh/y for the

residents only The power capacity of the undersea cables equals EPR 40 kV (40,000 Volts) in 4 cables from

the Sardinian coast of Portovesme to San Pietro Island, for a total power of 5.5 MW at 250 A to primarily supply power to the town of Carloforte. In San Pietro there are no autonomous generators, with the exception of a PV-wind power station of a total 1200 kWp installed, currently under retrofit.

Although there has not been an important increase in the number of residents in San Pietro in past years, the island has been experiencing a relevant increase in tourists presence and new houses, which are increasing the electrical demand. The typical seasonal electricity demand is shown in Fig. 3.26 (left) and the average monthly electricity load is shown on the right.

0,00

0,50

1,00

1,50

2,00

2,50

3,00

3,50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

MW

San Pietro -electricity load

Summer

Winter

0 ,0 0

0 ,5 0

1 ,0 0

1 ,5 0

2 ,0 0

2 ,5 0

3 ,0 0

3 ,5 0

jan

feb

mar ap

rmay ju

n jul

aug

sep

oct

nov

dec

Power (MW)

San Pie t ro - e lect ricit y loadaverage m o nthly e le ctricity load

Fig. 3.26 Details on yearly demand from San Pietro

3.2.6.2 RES POTENTIAL DATA

3.2.6.2.1 Solar radiation In San Pietro Island (located 39° 11’ North, 8° 18’ East) the solar radiation conditions are

excellent, with an average of 4.0 kWh/m2-day and maximums in July of 6,75 kWh/m2d, as shown in Fig. 3.27. The annual production per kWp corresponds to 1450 kWh.



47

Fig. 3.27 Solar radiation in San Pietro

San Pietro has a pre-existent PV installation, with a nominal power of 600 kW, based on 13500 PV polycrystalline modules, at originally 10% of efficiency. This installation, shown in Fig. 3.28 combined with a wind generation plant, is now under recycling.

Concerning solar thermal, the estimate installed capacity in the island is 600 m2 of mostly flat plate water solar collectors, with an annual production of 628.200 kWh/y

Fig. 3.28 Pre-existing RES installation on the island of San Pietro

3.2.6.2.2 Wind power The winds are intense and frequent over the whole year, with predominance of those in the

northern quadrant. Its annual rate is 35%, with peaks of maximum intensity in winter. The southern quadrant winds are frequent on average by 15% over the year with special

intensity in spring and fall. Eastern wind is often accompanied by significant rain storms and other events, especially during the fall and winter. In summer is almost completely absent.

The wind condition of the San Pietro Island, portrayed in Fig. 3.29, shows that the variation of wind speed is in the range of 9-10 m/s.



48

Fig. 3.29 Wind power conditions in San Pietro (left) and comparison with wind in Sardinia

More specifically, the relation between wind speed and frequency, portrayed in Fig. 3.30, shows that the 83% of wind frequency belongs to the speed range 6-20 m/s

For two decades, San Pietro island has been hosting a wind park. This first installation of 960 kW nominal power, made of 3 wind turbines of 320 kW installed capacity, producing 2500 MWh/y, is now under recycling. The turbines were often stopped for operation troubles and after few years of operation were broken because of the high wind speeds. Safety components were installed in order to avoid failures, but now the old turbine technology needs to be substituted. A new planning has been considered and already approved for the installation of modern wind turbines of higher capacity.

0

2

4

6

8

10

12

14

16

18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Frequence %

wind speed m/ s

San Pietrowind speed and frequence

wind speed m/s

Fig. 3.30 Wind velocity histogram in San.Pietro



49

3.2.6.3 OTHER INFORMATION

3.2.6.3.1 Mobility The impact on the mobility in San Pietro Island is very strong in summer and less in winter and

intermediate season, as in most Mediterranean island. This highlights two fundamental aspects: summer traffic causes air pollution and noise, but as a side effect, tourists bring work and prosperity to the island. Looking for solutions that can mitigate the hardships caused by congestion in the summer, it must be taken into account the needs of residents, many of which are involved in local tourism, without penalizing their activities.

Therefore, developing a network of electrically-powered vehicles, which do not discriminate mobility, but reduce air pollution and noise, can represent an opportunity for operators in the tourism sector and for the island environmental quality.

A new mobility plan includes the following actions: • purchase of tourist facilities and rentals of electric bikes and scooters;

• purchase electric mini-buses;

• creation of a network of electric recharging via the installation of photovoltaic roofs

3.2.6.3.2 Water Supply The water supply on the island of San Pietro, which serves about 4,000 customers, is limited to

the urban center of Carloforte. Only recently has been extended to the nearest suburbs with coordinated actions involving the province, the water service provider and the community of Carloforte. The water network is powered by two tanks of a total capacity of 3650 m3, fed by a submarine pipeline coming from Sant'Antioco island.

A recent financial support from the Autonomous Region of Sardinia enabled the rehabilitation of urban water supply of the old town through the substitution of steel pipes and polyethylene existing pipeline with new spheroidal cast iron pipelines, in conjunction with flow and pressure measuring instruments. In 2007, an amount of 883.000 m3 of water was collected and distributed, at 28 lit/s.

3.2.7 Mljet case study The Island of Mljet (Fig. 3.31)is situated in the southern Dalmatian archipelago, 30 km west

from Dubrovnik and south of the Peljesac Peninsula, separated from it by the Mljet channel. Mljet is an elongated island, with an average width of 3 km, 37 km long, total area of the island is 100.4 km2 and the highest peak is Veli Grad (514 m a.s.l.). The climate is Mediterranean; an average air temperature in January is 8.7 °C and in July 24 °C.

Fig. 3.31 The southern Dalmatian archipelago.

Economy is based on farming, viticulture, production of wine, olive growing, cultivation of medicinal herbs, fishing and tourism. The regional road (52 km) runs throughout the island. Mljet has ferry lines with Peljesac and Dubrovnik.



50

Tourism is the most valuable economy branch on the island but it also makes big stress on the resources (water, environment, electricity) especially during the summer months when population on the island is two to three times bigger than in the winter.

The following diagram shown in Fig. 3.32, shows the electricity demand at hourly level for one year.

0

200

400

600

800

1000

1200

1400

1600

1800

124

148

172

196

112

0114

4116

8119

2121

6124

0126

4128

8131

2133

6136

0138

4140

8143

2145

6148

0150

4152

8155

2157

6160

0162

4164

8167

2169

6172

0174

4176

8179

2181

6184

0186

41

Hrs

Dem

and

(kW

)

Fig. 3.32 The annual demand of the Mljet at hourly level.

Mljet island is an interconnected island to the Mainland Croatia via two sub-sea cables. The on-line diagram of the island is shown in Fig. 3.37 provides the power of each 10/0.4kV substation on the island, separated in Eastern and Western part. There is not a power station on the island. The major nodes on the islands and the capacity of the transformers is provided in Table 3-23.

Table 3-23 Substation of Mljet island

Western Eastern Substation Capacity Substation Capacity

Goli Vojska 20 Babino Polje1 100 H.Odisej 250 Zukovac 250 Pomena 50 Sobra 100 Govedari 125 Zaglavac 50 H.Melita 50 Prozurski 30 Pristaniste 30 Prozura 50 Soline 50 Okuklje 50 Polace 100 Maranovici 50 PS Polace 0 Korita 30 Ropa 50 Saplunara 30 Tatinica 30 Saplunara 2 Voljna 100 Kozarica 30 Blato 30 Babino Polje 2 100 Hotel Odisej is the major consumer on the island, especially during the summer months Table

3-24 provides the monthly consumption of the island, the consumption in its eastern part, the western part without the Hotels and the monthly demand of the hotel. The hotel consumes about 16% of the island demand while its demand during May-October is 20-23%. Therefore, installation of a desalination plant in a major consumer can be an interesting option.



51

In order to evaluate the losses avoided two undersea cables interconnecting Croatia and this Island exist. The first one is FXBTV 3x1x150 Cu (it is similar to cable XHP 81) and it connects Dingac (Peljesac peninsula) with Goli Vojska (the Island of Mljet) –western part of the island. The second cable type PP41 3x25 Cu connects Prapratno (Peljesac peninsula) with Zaglavc (the Island of Mljet). Each cable supplies the one part of the island and the last place supplied via PP41 is Babino Polje. The characteristics of the cables are provided in Table 3-25.

Table 3-24 Monthly distribution of the demand to each part of the island and hotel Odissej

Month Total Island (kWh)

Eastern Part (kWh)

Western Part (kWh)

Hotel Odissej

January 281588 149164.3 117024.8 12920 February 240517 121898.3 95868.1 18928 March 260185 131747.8 103371.9 20756 April 286669 144488 113344.8 23900 May 318482 137275.2 107789.8 61023.5 June 410146 166377 130572.1 94260 July 584697 241002 189229.9 128587 August 716584 299423 235103 151807 September 452052 190003 149201 94099.5 October 344002 144983.3 113937.1 70909.5 November 238349 124334.6 97514 13730 December 268644 141837.2 111308.2 12989 Total(MWh) 4401.92 1992.53 1564.26 703.91

Table 3-25 Undersea cables for interconnecting Croatia and Mljet

resistance of conductor at 200C, max. 0,124Ω/km 0,727 Ω/km

resistance of conductor at 900C max. 0,1593Ω/km 0,9271 Ω/km

capacitance, max 0,190µF/km 0,201 µF/km

inductance, max. 0,41mH/km 0,444 mH/km

The major distances on the island are summarized in Table 3-26. Table 3-26 Major cable distances between substations on the island of Mljet

Western Eastern Line Distance(km) Line Distance(km)

Goli Vojska-Dingac 15.43 Zaglavac-Prapatno 9.5 Goli Vojska-Govedari 2.6 Zaglavac-Sobra 1.72 Polace- Govedari 1.33 Zaglavac-Korita 11.1 Govedari-Blato 10 Babino Bolje-Sobra 3.14 RS Polace -Babino Polje 6.3 Tatinica-Lines Link 4.53 Blato-babino Polje2 7



52

In order to evaluate the impact of the scenarios studied on the emissions avoidance the critical 24-hour emissions curve from the power system of Croatia has been constructed using the methodology of [29]. The idea behind this methodology is to estimate how often each type of units from the Croatian power system is expected to be marginal, i.e. to ones to be influenced by the introduction of RES in some part of the power system.

For CO2 typical emissions curves for summer and winter are shown in Fig. 3.33. Similar curves are produced for the other Green House Gas (GHG). These curves use the hourly load curve for the Croatian power system [30], information on the estimated operation hours and emission levels as obtained from [31].

450510570630690750810870

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour

kg/M

Wh

Winter Summer

Fig. 3.33 Typical CO2 emissions curve for the upstream Croatian Network

3.2.7.1 RENEWABLE ENERGY SOURCES DATA Mljet island is quite abundant on RES. The wind power conditions are quite favourable and

located in the southern part of the island, the solar potential is higher than the rest of the country. More precisely, the average wind speed at 10m is 5.88m/s on the eastern part of the island. The

distribution of wind velocity is shown in Fig. 3.34.

0

2

4

6

8

10

12

0 1.5 3 4.5 6 7.5 910

.5 12 13.5 15 16

.5 18 19.5 21 22

.5 24 25.5 27 28

.5 30

Wind Velocity (m/s)

Freq

uenc

y (%

)

Fig. 3.34 Wind Velocity distribution on Mljet island



53

Solar radiation is also abundant on the island. The average PV production for 1kWp and 39o slope is shown in the following Fig. 3.35. The distribution of PV production is provided in Fig. 3.36.

0

20

40

60

80

100

120

140

160

180

1 2 3 4 5 6 7 8 9 10 11 12

Month

kWh/

KW

p

Fig. 3.35 PV production share per month

Fig. 3.36 PV production distribution



54

Govedari

Zaglavac

RS Korita

H.Odisej

CROATIA MAINLAND

Soline

H.Melita

Pomena

Goli Vojska

Korita

Saplunara

Saplunara 2 Voljna

RS Prozura

RS Sobra

Sobra

Babino Polje

Babino Polje 2

Blato

PristanistePolace

Ropa

RS Polace

Tatinica

Kozarica

Zukovac

Sobra link

Lines link

Lines link

Maranovici

Prozura Link

Prozurski Porat

Prozura

Okuklje

Fig. 3.37 One-line diagram of Mljet Island in Croatia



55

3.3 Matching Technologies and islands Having the description of storage technologies and details on the islands to be simulated, a matrix

of assigning storage technologies to islands has been created. Especially for pump hydro, which is quite site specific technology, the criteria of Table 3-27 were used for the islands selection.

Table 3-27 Criteria for the selection of islands for pumped hydro

Criteria Ideal case Interconnection Autonomous island with local power station (not

interconnected with the mainland) Current electricity production cost High current cost of the energy production

RES potential High wind potential (always wind with pumped hydro) or the existence of other renewable source economically feasible to explore.

Terrain morphology Not flat (maybe the existence of one reservoir in a suitable site).

For desalination technology, the selected islands are the ones that have lack of water and either have installed desalination units or such a unit is to be installed shortly. For battery power systems, the islands with rather low demand have been evaluated. For Milos and Corvo local interest was the one which lead us in simulating operation with hydrogen. The final matching of technologies and islands is provided in Table 3-28.

Table 3-28 Matching responsible partners, islands and technologies for the simulationsCountry Battery system Pump-hydro H2 Desalination Spain La Graciosa Greece Ios Milos Milos Croatia Mljet Portugal Corvo Corvo Italy San Pietro Cyprus Cyprus Cyprus Responsible Partner ITC NTUA CRES NTUA



56

4 SIMULATION WITH BATTERY STORAGE

4.1 Case study 1 results-La Graciosa A microgrid for supplying electric power to the island system is proposed. The system would

include a photovoltaic system, a wind farm, a diesel engine and batteries for energy storage. The system has been simulated using HOMER “Hybrid Optimization Model for Electric Renewables” developed by NREL (National Renewable Energies Laboratory, USA). It assumes that it is an isolated grid, where the submarine cable connecting La Graciosa to the neighbouring island of Lanzarote does not exist.

HOMER, the software used for techno-economic analysis and optimization of electrical microgrid for La Graciosa, allows to:

• Determine whether the renewable energy resources are adequate

• The optimal size of the system components of a hybrid system: number of photovoltaic modules, power of wind generators, size of backup diesel gen-set, number and capacity of battery storage, power of rectifiers and inverters connecting the DC and AC bus.

• Investment cost of the hybrid system and annual O&M costs

• Economic sensitivity analysis to changes in cost, RES availability and consumption loads

The objective of the simulation was to optimally design an electric system that can give energy autonomy to the island by maximizing the penetration of RES and minimizing the needs for fossil fuels to satisfy the electricity demands from households, productive activities and public services in La Graciosa.

4.1.1 Scenario 1(Optimization for maximum RES penetration)

The microgrid will combine photovoltaic, wind and diesel systems to supply, without interconnection to other islands electrical grids, the electrical needs of the island of La Graciosa as, the electrical needs of the island of La Graciosa as Fig. 4.1 shows. The control concept is shown in Fig. 4.2.

Wind turbine

Diesel genset) Inverter

Rectifier

Batteries

Loads La Graciosa9.5 MWh/day 668 kW (Max.) Photovoltaic

CA

bus

DC

bus

Losses

Losses

Losses

Converter

Fig. 4.1 Proposed flow chart between the components to be installed on La Graciosa



57

Batteries

Control and power conditioning unit

Loads Fig. 4.2 Control concept for the components to be installed on La Graciosa

4.1.1.1 COMPONENTS DESCRIPTION

4.1.1.1.1 PV system for La Graciosa The photovoltaic modules proposed for the system are Isofoton IS-200 monocrystal 200 Wp

(+/- 5 %)

• Efficiency 12 %

• Nominal voltage 12 V

• Open circuit voltage 57.6 V, and at maximum power 46.1 V.

• Open circuit intensity 4.7 A, and maximum power current of 4.35 A

From the simulation carried out with HOMER, we obtained that the optimal system should have a total installed photovoltaic power of 300 kWp, and would be mad of a total of 500 Modules. Other parameter values obtained from the simulation for the photovoltaic system are provided in Table 4-1

Table 4-1 Results for the PV system-La Graciosa

Variable Value Average output: 1,378 kWh/d

Minimum output: 0.0 kW

Maximum output: 320 kW

Maximum Solar penetration (if could have been absorbed) 14.44%

Capacity factor: 19.15%

Hours of operation: 4,729 hr/yr


20

40

60

80

100

120

Pow

er (k

W)

0 50 100 150 200 250 300 350

0

10

20

30

40

50

60PV Array Power Output PDF

Value (kW)



58

Fig. 4.3 PV power output time-series (left) Power output PDF (right)

4.1.1.1.2 Wind farm for La Graciosa For the HOMER simulation the Fuhrländer 250 kW was chosen with power curve and icon

shown in Fig. 4.4. The results from the simulation gives that optimum size of the wind farm for La Graciosa would be 6 wind turbines and the production characteristics are provided in Table 4-2 and Fig. 4.5. Table 4-3 presents the overall production data of wind for the studied scenario.

0 5 10 15 20 250

100

200

300Power Curve

Wind Speed (m/s)

Fig. 4.4 Wind turbine data velocity to power curve (left)- Fuhlander icon (right)

Table 4-2 Results for the wind power system-La Graciosa

Variable Value Model Fuhrländer 250 Unit max. power 250 kW Number 6 Total power 1,500 kW Production 3,426,194 kWh/y


200

400

600

800

1,000

1,200

1,400

1,600

()

0 200 400 600 800 1,000 1,200 1,400 1,600 1,8

0

5

10

15

20

25Fuhrländer 250 Power Output PDF

Value (kW)

Fig. 4.5 Wind turbine production time-series(left)- Production wind power PDF (right)


Variable Value Total capacity: 1,800 kW Average output: 391 kW Minimum output: 0.000 kW Maximum output: 1,754 kW Potential wind power penetration 98.3% Actual wind power penetration 68%



59

Capacity factor: 21.7% Hours of operation: 8,182 hr/yr

4.1.1.1.3 Diesel generator The hybrid electrical generation system that will power the microgrid for La Graciosa will also

include a diesel genset. Although sun and conditions in La Graciosa are excellent, the photovoltaic system and the wind

farm would not be able by themselves to guarantee 100 % electricity supply to the island. On the hourly energy balance there are moments of excess electricity production from the RES systems, and there are other moments when there is a power deficit which has to be supplied by conventional energy power system, namely a diesel genset.

The graphs obtained from the HOMER simulation show the hourly power generation from the diesel genset in Fig. 4.6, while a summary of the results are provided in Table 4-4.


100

200

300

400

P(kW

)

0 100 200 300 4000

10

20

30

40

50

60

70Generator 1 Electrical Output PDF

Value (kW)

Fig. 4.6 Diesel production time-series (left)- Diesel production PDF (right)

Table 4-4 Results for the diesel unit-La Graciosa

Variable Value Hours of operation 3,153 hr/yr Number of starts 415 starts/yr Average electrical output: 378 kW Min. electrical output: 100 kW Max. electrical output: 400 kW Annual fuel usage: 398,605 L/yr Specific fuel usage: 0.335 L/kWhAv. electrical efficiency: 30.4%

4.1.1.1.4 Battery Energy Storage The simulation only considered energy storage in batteries. The simulation was based on the

Trojan L16P batteries, and the results gave the values of Table 4-5 for the battery system: Table 4-5 Results for the battery storage device-La Graciosa

Variable Value Nº of batteries 2,000 Trojan L16P Battery throughput 448,448 kWh/yr Battery life 4.79 yr Battery autonomy 7.60 hours



60

Fig. 4.7 describes the histogram for the State of Charge (SOC) of the battery in frequency and statistics. Fig. 4.8 presents another important parameter of the storage device operation the hourly exchange level.

Fig. 4.7 Hourly battery state of charge (left)- Monthly statistics (right)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec-100

-50

0

50

100

150

P(kW

)

-600 -400 -200 0 200 4000

5

10

15

20

25

30

35Battery Charge Power PDF

Value (kW)

Fig. 4.8 Exchange of Battery with the rest of the network Hourly (Left)-PDF(Right)

4.1.1.2 ANNUAL ELECTRIC ENERGY PRODUCTION The optimal system configuration of the microgrid for La Graciosa seeks to achieve a maximum

RES penetration, and could have been achieved if there had been higher correlation between load and RES production but understanding that aiming at a 100 % supply of electrical power would mean a disproportionate investment cost, mainly related to the energy storage battery system. That is the reason why in the optimal solution of the HOMER simulation, RES penetration is estimated at 78 %, and the rest, 22 %, would have to be supplied by the diesel generator. The production of each of the components are provided in Table 4-6 and for each month is graphically represented in Fig. 4.10. Electricity production, when compared to actual energy consumption, gives excess of electricity production on the yearly bases 1,395,873 kWh (difference of production of 5,045,442 kWh – consumption of 3,485,007 kWh) as Table 4-7 shows. This amounts to 40% of the demand of the island and reduces significant wind power penetration to 78%.

Table 4-6 Summary of the components penetration-La Graciosa

Component Production (kWh/yr) Fraction PV array 503,146 10%

Wind turbines 3,426,194 68%

Diesel Gen. 1,116,102 22%

Total 5,045,442 100%

Table 4-7 Excess energy and RES fraction on La Graciosa



61

Variable Value Primary load 3,485,007 kWh/yr

Renewable fraction: 0.779

Excess electricity: 1,395,872 kWh/yr

Fig. 4.9 Monthly average production by the units installed on La Graciosa

In the following more details on this issue, as well as ways of exploiting it are provided.

4.1.1.2.1 Excess electricity Given the variability in the energy demand curve (on daily and yearly bases), the high level of

RES penetration in the microgrid proposed for La Graciosa, and the limited battery energy storage capacity (due to battery investment cost restrictions), there will be moments when excess electricity from the photovoltaic system and the wind farm will be produced. This excess electricity, on yearly bases, has been estimated at 1,395,872 kWh/yr and is distributed as Fig. 4.10 shows.

Fig. 4.10 Hourly excess electric power (Left)-PDF(Right)

This excess power would have to go to dissipating loads that would consume this electricity and produce heat, or a curtailment policy would have to be implemented to cut production from the wind farm and photovoltaic system when supply exceeds demand. Both are inefficient energy solution indeed. A possible solution could be to make use of the excess energy on reverse osmosis desalination plants or hydrogen production for transport fuel.

4.1.1.2.1.1 Water production from excess electricity La Graciosa does not have any source of its own of fresh water, and receives it from Ye in

neighbouring Lanzarote through a water pipe that communicates both islands. The water pipe descends from the Famara Cliff, in Lanzarote, and crosses the channel (El Río). Nevertheless the



62

current volume is insufficient to cover La Graciosa’s needs in high tourist season. The water production, transport and distribution infrastructure belongs to the water company of Lanzarote, INALSA.

Estimating a specific consumption of 2.4 kWh of electricity per cubic meter of water (RO desalination plants with pressure recovery systems), the use of the excess energy from La Graciosa microgrid for sea water desalination we would obtain the water production as described in Table 4-8.

Table 4-8 Summary of the desalination plant results-La Graciosa

Variable Value Excess electricity: 1,311,431kWh/yrSpecific consumption: 2.4 kWh/m3 Water production 546,430m3/year

4.1.1.2.1.2 Hydrogen production from excess electricity La Graciosa has a small car fleet and consumes also transport fuel for its fishing boats and

transport ships moving passengers and cargo that connect the island to neighbouring Lanzarote. The fossil fuel consumed by these vehicles could be substitute in the future by hydrogen to be used in fuel cells and internal combustion engines.

If excess RES electricity was to be used to run electrolysers for hydrogen production, and estimating a specific consumption of 4,5 kWh per normal cubic meter of hydrogen (11,2 Nm3H2 in a kg of H2), the hydrogen production that could be obtained is provided in Table 4-9

Table 4-9 Summary of the Hydrogen production when excess electricity is utilized

Variable Value Excess electricity 1,395,872 kWh/yr Specific consumption 4,5 kWh/ Nm3H2 Hydrogen production 310,194 Nm3H2/yearHydrogen production 27,695 kg H2/year

4.1.1.3 ECONOMIC ANALYSIS The HOMER simulation gives the optimal system architecture for the La Graciosa microgrid for

maximum RES penetration according to the following strategy as described in Table 4-10.

Table 4-10 Summary of the components-La Graciosa

PV Array: 200 kW Wind turbine: 6 Fuhrländer 250/1500kWDiesel Generator: 400 kW Battery: 2,000 Trojan L16P Inverter: 400 kW Rectifier: 400 kW Dispatch strategy: Cycle Charging

The cost breakdown for the different system components is shown in Table 4-11. In terms of net present cost, considering an interest rate of 6 % for the cash flow analysis, the results are:

The emissions avoidance due to the proposed solution is provided in Table 4-12.



63

Table 4-11 Cost breakdown per component-La Graciosa

Components Initial Capital(€) Annualized Capital (€/yr)

Annualized Replacement (€/yr)

Annual O&M(€/yr)

Annual Fuel(€/yr)

Total Annualized (€/yr)

PV Array 1,000,000 87,185 -7,249 4,000 0 83,935

Fuhrländer 250 1,500,000 130,777 0 60,000 0 190,777

Diesel Genset 600,000 52,311 -7,563 126,120 398,605 569,472

Battery 600,000 52,311 80,878 40,000 0 173,189

Converter 40,000 3,487 730 400 0 4,618

Other 0 0 0 104,966 0 104,966

Totals 3,740,000 326,07 66,796 335,486 398,605 1,126,957

Table 4-12 Emissions avoidance due to the proposed solution

Parameter Emissions (kg/yr) Carbon dioxide 1,049,657

Carbon monoxide 2,591

Unburned hydrocarbons 287

Particulate matter 195.3

Sulphur dioxide 2,108

Nitrogen oxides 23,119



64

4.1.2 Scenario 2-50% RES share Due to high excess electricity, the option of trying to maintain 50% RES penetration is studied.

The capacity of the components changes.

4.1.2.1 COMPONENTS DESCRIPTION

4.1.2.1.1 PV system for La Graciosa The installed photovoltaic capacity for achieving 50% penetration will be 150 kWp, comprising

500 Modules. A summary of the simulation results are provided in Table 4-13. Hourly production and PDf are provided in Fig. 4.11

Table 4-13 Results for the PV system- La Graciosa

Variable Value Average output: 643 kWh/d

Minimum output: 0 kW

Maximum output: 149.5 kW

Maximum Solar penetration (if could have been absorbed) 6 %



Fig. 4.11 PV power output time-series (left) Power output PDF (right)

4.1.2.1.2 Wind farm for La Graciosa Table 4-14 presents the overall production data of wind for the studied scenario, while Fig. 4.12

presents the hourly production of the wind turbine. Table 4-14 Results for the wind power system- La Graciosa

Variable Value Model Fuhrländer 250 Unit max. power 250 kW Number 3 Total power 750 kW Production 1,713,097 kWh/y



65

Fig. 4.12 Wind turbine production time-series(left)- Production wind power PDF (right)


Variable Value Total capacity 900 kW

Average output 195.6 kW

Minimum output 0.000 kW

Maximum output 877 kW

Potential wind power penetration 44%

Capacity factor 21.7%

Hours of operation 8,182 hr/yr

4.1.2.1.3 Diesel generator The configuration for the diesel Generation is increased by 100kW compared with the previous

scenario of 100% RES. The updated data on the diesel capacity to be installed is described in Table 4-16.Fig. 4.13 presents the hourly power generation from the diesel genset.

Ja n Fe b Mar A pr May Jun Jul A ug Se p Oc t No v De c0

10 0

20 0

30 0

40 0

Pow

er (k

W)

Fig. 4.13 Diesel production time-series (left)- Diesel production PDF (right)

Table 4-16 Results for the diesel unit-La Gracioa

Variable Value Hours of operation 5,383 hr/yrNumber of starts 451 starts/yrAverage electrical output: 349 kWMin. electrical output: 100 kWMax. electrical output: 400 kWAnnual fuel usage: 642,072 L/yrSpecific fuel usage: 0.342 L/kWh



66

Av. electrical efficiency: 29.7 %

4.1.2.1.4 Battery Energy Storage Table 4-17 presents the characteristics of the battery system for 50% RES penetration.

Table 4-17 Results for the battery storage device

Variable Value Nº of batteries 1,000 Trojan L16P Battery throughput 460,284 kWh/yrBattery life 2.34 yrBattery autonomy 3.80 hours




4.1.2.2 ANNUAL ELECTRIC ENERGY PRODUCTION The production of each of the components is provided in Table 4-18 and for each month is

graphically represented in Fig. 4.16. Electricity production, when compared to actual energy consumption, gives excess of electricity production on the yearly bases 216,451 kWh (difference of production of 3,888,501kWh – consumption of 3,485,007 kWh) as Table 4-19 shows.



67

Table 4-18 Summary of the components penetration-Scenario 2

Component Production (kWh/yr) Fraction PV array 234,802 6% Wind turbines 1,713,097 44% Diesel Gen. 1,879,266 50% Total 3,827,165 100%

Table 4-19 Excess energy and RES fraction on La Graciosa-Scenario 2

Variable Value Primary load 3,484,556kWh/yr Renewable fraction: 0.50 Excess electricity: 209,764 kWh/yr

Fig. 4.16 Monthly average production by the units installed on La Graciosa-Scenario 2

In the following more details on this issue, as well as ways of exploiting it are provided.

4.1.2.2.1 Excess electricity Given the variability in the energy demand curve (on daily and yearly bases), the high level of

RES penetration in the microgrid proposed for La Graciosa for La Graciosa under the 50 % RES scenario, and the limited battery energy storage capacity (due to battery investment cost restrictions), there will be moments when excess electricity from the photovoltaic system and the wind farm will be produced. This excess electricity, on yearly bases, has been estimated at 209,764 kWh/yr and is distributed as Fig. 4.17shows.




68

This excess power would have to go to dissipating loads that would consume this electricity and produce heat, or a curtailment policy would have to be implemented to cut production from the wind farm and photovoltaic system when supply exceeds demand. Both are inefficient energy solution indeed. A possible solution could be to make use of the excess energy on reverse osmosis desalination plants or hydrogen production for transport fuel.

4.1.2.3 ECONOMIC ANALYSIS The optimal system architecture for the La Graciosa microgrid for 50% RES penetration

according to the following strategy as described in Table 4-20.

Table 4-20 Summary of the components to be installed

PV Array 140 kWWind turbine 3 Fuhrländer 250Diesel Generator 400 kWBattery 1,000 Trojan L16PInverter 400 kWRectifier 400 kWDispatch strategy Cycle Charging

The cost breakdown for the different system components is shown in Table 4-21 In terms of net present cost, considering an interest rate of 6 % for the cash flow analysis, the results are:

The emissions avoidance due to the proposed solution is provided in Table 4-22.



69

Table 4-21 Cost breakdown per component

Components Initial Capital(€) Annualized Capital (€/yr)

Annualized Replacement (€/yr)

Annual O&M(€/yr)

Annual Fuel(€/yr)

Total Annualized (€/yr)

PV Array 700,000 61,029 -5,074 2,800 0 58,755 Fuhrländer 250 750,000 65,388 0 30,000 0 95,388 Diesel Genset 600,000 52,311 -3,683 215,320 642,072 906,020 Battery 300,000 26,155 74,095 20,000 0 120,251 Converter 40,000 3,487 730 400 0 4,618 Other 0 0 0 54,213 0 54,213

Totals 2,390,000 208,371 66,068 332,588 642,072 1,249,100

Table 4-22 Emissions avoidance due to the proposed solution

Parameter Emissions (kg/yr) Carbon dioxide 1,690,788Carbon monoxide 4,173Unburned hydrocarbons 462Particulate matter 315Sulphur dioxide 3,395Nitrogen oxides 37,240



70

4.1.3 Comparison between RES penetration scenarios

The three scenarios correspond to the two situations modelled and simulated using HOMER, against the current situation with no renewable energies and all electric power supplied through the connecting submarine cable to Lanzarote:

• Optimization for maximum RES penetration (78% RES)

• 50 % RES penetration

• Business-as-usual situation (0% RES)

Table 4-23 shows the differences between certain parameters of the three scenarios for La Graciosa electrical system:



71

Table 4-23 Summary of the results for the La Graciosa Case study

SUBMARINE CABLE

AUTONOMOUS ELECTRICAL SYSTEM

0 % RES 50 % RES 78 % RES (optimum)

PV Installed power 0 140 kWp 200 kWpAverage output: 0 643 kWh/d 919 kWh/dMaximum output: 0 149.5 kW 214 kWSolar penetration: 0 6 % 10 %Wind Number (Fuhrländer 250) 0 3 6 Total power 0 750 kW 1,500 kWProduction 0 1,713,097 kWh/y 3,426,194 kWh/yWind penetration 0 44 % 68 %Total capacity: 0 900 kW 1,800 kWAverage output: 0 195.6 kW 391 kWBatteries Number (Trojan L16P) 0 1,000 2,000Battery throughput 0 460,284 kWh/yr 448,448 kWh/yrBattery life 0 2.34 yr 4.79 yrBattery autonomy 0 3.80 hours 7.60 hoursDiesel Hours of operation: 8760hr/yr 5,383 hr/yr 3,153 hr/yrNumber of starts: 1000 starts/yr 451 starts/yr 415 starts/yrAverage electrical output: 398 kW 349 kW 378 kWMin. electrical output: 210 kW 100.0 kW 100.0 kWMax. electrical output: 668 kW 400 kW 400 kWProduction 3,485,039 kWh/yr 1,879,266 kWh/yr 1,190,835

kWh/yr Annual fuel usage: 1,361,824 L/yr 642,072 L/yr 398,605L/yrCarbon dioxide 3,586,130 kg/yr 1,690,788kg/yr 1,049,657 kg/yrTotal electricity production 3,485,039 kWh/yr 3,827,165 kWh/yr 4,952,459

kWh/yrExcess Electricity 16 kWh/yr 209,764 kWh/yr 1,311,431 kWh/yr

4.1.4 Summary

La Graciosa is a small island located in the north of Lanzarote. Currently is being provided of electric power and water from neighboring Lanzarote through a submarine cable with capacity for 1.3 MW, and a submarine water pipe.

The island has a small resident population of 658 people which lives all year long in two small communities, Caleta del Sebo (the capital), and Casas de Pedro Barba (north of La Graciosa.



72

During the summer the island population increases due to tourism, which together with fishering represents the main economic activities.

La Graciosa, as happens with the rest of the Canary Islands, is exposed to the Trade winds that blow predominantly from the northeast. Strong constant winds throughout the year, although in the summer months the mean average speed is higher than in the winter. Solar radiation is also high most of the year. Wind and solar radiation gives an important potential for renewable energy electricity production.

The island yearly electricity consumption is 3,485,039 kWh which, considering the specific consumption of the Canary Islands thermal plants of 0.25 kg of fuel oil per kWh, represents a yearly consumption of 871,260 kg of fuel oil. Associated to this consumption of fossil fuels there are 3,140 tons of CO2. Peak electric power demand is 668 kW

An analysis for la Graciosa has been made under two scenarios for RES penetration that assumes that no interconnection through submarine cable exists, and that all electric energy is produced in the island from renewable energies and a diesel system backup installed in the island. The first scenario considers the optimum maximum penetration is terms of investment cost, and a 78 % penetration was estimated. The second scenario looks at a 50 % RES penetration. Finally both scenarios are compared to the business as usual scenario, where electricity is supplied through the submarine cable from Lanzarote.

The simulations were carried out using HOMER (“Hybrid Optimization Model for Electric Renewables” developed by NREL - USA). For the maximum penetration scenario a 78% RES for la Graciosa was estimated. The investment cost was estimated at 3,740,000 €. It included a hybrid system made up of 200 kWp photovoltaic, 1,500 kW wind and 400 kW diesel genset. The need to meet as much as possible from RES even at moments of low wind speeds and low solar radiation, creates excess electricity at valley hours of the demand curve. To try to adjust the electricity demand curve and the supply of the RES electricity generation system, battery storage is considered. Nevertheless battery storage is expensive, so enough capacity for all the excess electricity is not a suitable option.

To reduce excess electricity, one possible option is the implementation of a curtailment policy (reduce power output from RES generation system at valley hours) or burning excess electricity in dissipation loads. Other possibilities include the use of Reverse Osmosis desalination plants that could be connected at moments when electricity supply from RES is higher than demand from the conventional electric loads. The production from the RO plants could substitute fresh water being supplied through a submarine water pipe from Lanzarote (which is currently being also produced by RO plants). There is another interesting application for the excess electricity: hydrogen production. Electrolisers could operate also as programmable loads that can consume excess electricity when supply is higher than demand. The H2 could be used as transport fuel for the small road vehicle fleet of the island, and even for the fishing boats.

Under the 50 % RES penetration scenario (also assumes that no submarine interconnecting cable exists), the investment cost is 2,390,000 €. Under this scenario also small excess electricity is produced, if no curtailment policy is implemented, or no water desalination plants and H2 production system are installed.

4.2 Case study 2 San Pietro

4.2.1 Scenario 1 A preliminary feasibility study for a microgrid with high penetration of RES for supplying

electric power to the island of San Petro has been performed. The generation systems will be based on a wind and solar photovoltaic energy, with diesel back-up system installed in the island, and battery for energy storage. The system has been simulated using HOMER “Hybrid Optimization Model for Electric Renewables” developed by NREL (National Renewable Energies Laboratory,



73

USA). It assumes that it is an isolated grid, where the submarine cable connecting San Pietro to other grids does not exist.

The objective of the simulation was to optimally design an electric system that can give energy

autonomy to the island by maximizing the penetration of RES and minimizing the needs for fossil fuels to satisfy the electricity demands from households, productive activities and public services of San Pietro. The HOMER software used for the technical analysis and optimization of the electrical microgrid, allows to determine the size of the system components suitable to cover electrical in the most cost-effective way, considering renewable energy resources in the island. Gives the necessary installed power of the wind-PV hybrid system, the power of the diesel back-up, the needed battery storage capacity, and the power of rectifiers and inverters connecting the DC and AC bus.

In the economic analysis, HOMER estimates the investment cost of the optimal hybrid system and annual O&M costs.

4.2.2 Scenario 2

The simulation made for San Pietro assumes that no interconnection exist through submarine cable to other electrical systems. The optimization simulation, for a 82 % of RES penetration in the island, gave the following sizes for the system components:

Table 4-24 Summary of the proposed system for San Pietro

PV Array: 900 kW

Wind turbine: 4 ENERCON E-70 2.3 MW

Diesel genset: 2,000 kW

Battery: 8,000Trojan L16P

Inverter: 1,000 kW

Rectifier: 1,000 kW

Table 4-25 Summary of the results for the San Pietro Case study

Component Production FractionPV array 1,484,555 kWh/yr 5%

Wind turbines 21,651,706 kWh/yr 77%

Generator 1 4,939,221 kWh/yr 18%

Total 28,075,482 kWh/yr 100%

4.2.2.1 ANNUAL ELECTRIC ENERGY CONSUMPTION

Table 4-26 Summary of the results for San Pietro study

LOAD Consumption Fraction AC primary load 15,620,034 kWh/yr 100% Total 15,620,034 kWh/yr 100%

Table 4-27 Summary of the results for San Pietro Case study



74

Variable Value Variable Renewable fraction: 0.824 Renewable fraction:

4.2.2.2 COMPONENTS DESRIPTION AND RESULTS FOR EACH OF THEM

4.2.2.2.1 PV System for San Pietro Results from the HOMER simulation show that the optimal system should have a total installed

photovoltaic power of 900 kWp. Other parameter values obtained from the simulation for the photovoltaic system are provided in Table 4-28. The graphic representation of the production is shown in Fig. 4.18.

Fig. 4.18 PV production annual and Histogram for San Pietro

Table 4-28 Summary of the results for PV and San Pietro

Variable Value Average output: 4,067 kWh/d


Maximum output: 1,010 kW

Solar penetration: 9.50 %

Capacity factor: 18.83 %


4.2.2.2.2 AC Wind Turbine: Enercon E-70 2.3 MW The ENERCON E-70 was chosen for the HOMER simulation. The power curve and the wind

turbine picture are shown in Fig. 4.19. The results from the simulation gives that optimum size of the wind farm to be installed in San Pietro is 4 wind turbines and the production characteristics are provided in Table 4-2. Table 4-3 presents the overall production data of wind for the studied scenario, while Fig. 4.20 presents the characteristic of the wind power production.



75

Fig. 4.19 Characteristics of the Enercon E-70

Table 4-29 Summary of the results for the wind turbine production on San Piedro

Variable Value Model Enercon E-70 Unit max. power 2.3 MW Number 4 Total power 9,240 kW kW Production 21,651,706 kWh/yr

Table 4-30 Details for the wind turbine production

Variable Value Total capacity: 9,240 kW

Average output: 2,472 kW


Maximum output: 9,240 kW

Wind penetration: 138.6%


Hours of operation: 8,736

Fig. 4.20 Wind power production characteristics time-series (left)- Production wind power PDF (right)

4.2.2.2.3 Diesel Gen-SET



76

The hybrid electrical generation system that will power the San Pietro will also include a diesel genset for back-up, which when wind and solar systems supported by batteries are not able to supply the electric power demand.

On the hourly energy balance there are moments of excess electricity production from the RES systems, and there are other moments when there is a power deficit which has to be supplied by conventional energy power system, namely a diesel genset. The graphs obtained from the HOMER simulation show the hourly power generation from the diesel genset in, while a summary of the results are provided in Table 4-31.

Table 4-31 Details for the Diesel unit

Variable Value Hours of operation: 3,037 hr/yr

Number of starts: 417 starts/yr

Average electrical output: 1,626 kW

Minimum electrical output: 500 kW

Maximum electrical output: 2,000 kW

Annual fuel usage: 1,720,725 L/yr

Specific fuel usage: 0.348 L/kWh

Average electrical efficiency: 29.2%

4.2.2.2.4 BATTERY CHARACTERISTICS The simulation only considered energy storage in batteries. The simulation was based on the

Trojan L16P batteries, and the results gave the values of Table 4-32 for the battery system Table 4-32 Details for the Battery Characteristic

Variable Value Nº of batteries 8,000 Trojan L16P

Battery throughput 1,058,565 kWh/yr

Battery life 8.12 yr

Battery autonomy 6.78 hours





77


4.2.2.3 ANNUAL ELECTRIC ENERGY PRODUCTION The optimization simulation carried out with HOMER gives the optimal system configuration for

the San Pietro micro-microgrid, for maximum RES penetration, considering that aiming at 100 % RES system would imply a disproportionate high investment cost, mainly related to the energy storage battery system. That is the reason why in the optimal solution of the HOMER simulation, RES penetration is estimated at 82 %, and the rest, 18 %, would have to be supplied by the diesel generator. The production of each of the components is provided in Table 4-33 and for each month is graphically represented in Fig. 4.23. Electricity production, when compared to actual energy consumption, gives excess of electricity production on the yearly bases of 12,002,370 kWh (difference of production of 28,075,482 kWh – consumption of 15,620,034 kWh) as Table 4-34 shows.

Fig. 4.23 Representation of the production from the various components

Table 4-33 Summary of the components penetration

Component Production (kWh/yr) Fraction PV array 335,430 9%

Wind turbines 3,426,194 69%

Diesel Gen. 1,190,835 22%

Total 4,952,459 100%

Table 4-34 Excess energy and RES fraction on San Pietro

Variable Value Total production 28,075,482 kWh

Primary load 15,620,034 kWh/yr

Renewable fraction: 0.82

Excess electricity: 12,002,370 kWh/yr



78

4.2.2.3.1 Excess electricity

Given the variability in the energy demand curve (on daily and yearly bases), the high level of RES penetration in the power system proposed for San Pietro, and the limited battery energy storage capacity (due to battery investment cost restrictions), there will be moments when excess electricity from the photovoltaic system and the wind farm will be produced. This excess electricity, on yearly bases, has been estimated at 12,002,370 kWh and is distributed as shown in Fig. 4.24.

12 GWh of excess electric energy is a very big amount of energy, that if not used, would have to go to dissipating loads, or a curtailment policy would have to be implemented to cut production from the wind farm and photovoltaic system when supply exceeds demand. Possible applications for this excess electricity include reverse osmosis water desalination or electrolysis hydrogen production for transport fuel.


4.2.2.3.1.1 Water production from excess electricity Estimating a specific consumption of 2.4 kWh of electricity per cubic meter of water (RO

desalination plants with pressure recovery systems), the use of the excess energy from San Pietro for sea water desalination we would obtain the water production as described in Table 4-35.

Table 4-35 Summary of the desalination plant results

Variable Value Excess electricity: 12,002,370 kWh/yr

Specific consumption: 2.4 kWh/m3

Water production 5,000,000 m3/year

4.2.2.3.1.2 Hydrogen production from excess electricity If excess RES electricity was to be used to run electrolysers for hydrogen production, and

estimating a specific consumption of 4,5 kWh per normal cubic meter of hydrogen (11,2 Nm3H2 in a kg of H2), the hydrogen production that could be obtained is provided in Table 4-36.

Table 4-36 Summary of the Hydrogen production when excess electricity is utilized

Variable Value Excess electricity 12,002,370 kWh/yr

Specific consumption 4,5 kWh/ Nm3H2

Hydrogen production 2,667,193 Nm3H2/year

Hydrogen production 238,142 kg H2/year



79

4.2.2.4 ECONOMIC ANALYSIS The HOMER simulation gives the optimal system architecture for the San Pietro microgrid for

maximum RES penetration according to the following strategy as described in Table 4-37. The total cost for the components is shown in Table 4-38, considering an interest rate of 6 % for

the cash flow analysis. The emissions avoidance due to the proposed solution is provided in Table 4-39

Table 4-37 Summary of the installed components

PV Array: 900 kW Wind turbine: 4 ENERCON E-70 2.3 MW Diesel Generator: 2,000 kW Battery: 8,000 Trojan L16P Inverter: 1,000 kW Rectifier: 1,000 kW Dispatch strategy: Cycle Charging

Table 4-38 Breakdown of the installation cost per component

Initial Capital(€) Annual O&M(€/yr)

Annual Fuel(€/yr)

Totals 23,800,000 1,644,669 1,720,725

Table 4-39 Emissions avoidance due to the proposed solution-San Piedro

Pollutant Emissions (kg/yr) Carbon dioxide 4,531,237

Carbon monoxide 11,185

Unburned hydrocarbons 1,239

Particulate matter 843

Sulphur dioxide 9,100

Nitrogen oxides 99,802

4.2.3 Summary

San Pietro is a volcanic Island located approximately at 7 km off the South western Coast of Sardinia. Has a resident population of 6,660 inhabitants which are mostly concentrated in the fishing town of Carloforte, that more than triples in the summer months, growing to 15-20,000 people.

The island electrical system is interconnected to Sardinia, through a submarine cable, with a capacity of 5.5 MW (at 15 kV). The island electricity demand experiences high seasonal variation due to tourist activity.

Wind and solar radiation gives an important potential for renewable energy electricity production. Average wind speed is 5,3 m/s, with relatively constant wind velocities during the year (4 m/s in July to 6.2 in February). Solar yearly average radiation is 4,5 kWh/m²-d, with yearly variations from 1,9 kWh/m²-d in December, to 7,2 kWh/m²-d in June.



80

The island yearly electricity power demand suffers a high seasonal variation from the winter to the summer months, due to the tourist activity. Yearly consumption is 15,620,034 kWh which produces around 14,000 tons of CO2. Peak electric power demand is 3 MW.

An analysis for a microgrid with maximum RES penetration has been made for San Pietro, assuming that no interconnection through submarine cable exists, and that all electric energy is produced from renewable energies and a diesel system backup installed in the island. The optimum looks for maximum penetration is terms of investment cost, and an 82 % penetration was estimated. A 100 % RES penetration objective is not consider, given the high investment cost of the solution, specially associated to necessary battery energy storage capacity.

The total investment cost was estimated at 23,800,000 €. It included a hybrid system made up of 900 kWp photovoltaic, 9,240 kW wind and 2,000 kW diesel genset. The need to meet as much as possible from RES even at moments of low wind speeds and low solar radiation, creates excess electricity at valley hours of the demand curve. To try to adjust the electricity demand curve and the supply of the RES electricity generation system, battery storage is considered. Nevertheless battery storage is expensive, so enough capacity for all the excess electricity is not a suitable option.

Two alternatives for the excess electricity are proposed: Reverse Osmosis desalination plants that could be connected at moments when electricity supply from RES is higher than demand from the conventional electric loads; Hydrogen production, with electrolysers operating also as programmable loads that can consume excess electricity when supply is higher than demand. The H2 could be used as transport fuel.



81

4.3 Conclusions of the case studies analysed Simulations for preliminary feasibility study for microgrids with high penetration of RES for

three European islands has been carried out: San Pietro, La Graciosa, considering only energy storage in batteries, and that no submarine cables exist connecting these islands to any outside electrical system. Both simulated islands, have abundant RES potential and typical seasonal demand variation from the winter to the summer months, due to the tourist activity

These simulations with the use of the HOMER simulation programme have identified the combination of wind and PVs, with diesel back-up system installed in the island and battery for energy storage.

The optimum microgrid design seeks maximum penetration of RES at minimum investment cost, and under these conditions the optimal is not 100 % coverage, because investment on storage capacity would be too high. The analysis was made under different RES penetration scenarios. The analysis assumes that no interconnection through submarine cable exists, and that all electric energy is produced from renewable energies and a diesel system backup installed in the island.

For both islands, La Graciosa and San Pietro, scenarios that lead to 80% RES penetration were simulated. The cost for La Graciosa is significantly lower due to its smaller size, and the higher correlation between load and RES especially during the periods of high demand compared to San Pietro. For levels of penetration above this percent, the excess electricity is geometrically increased as well as the installation cost.

Two alternatives for the excess electricity are proposed: Reverse Osmosis desalination plants that could be connected at moments when electricity supply from RES is higher than demand from the conventional electric loads; Hydrogen production, with electrolysers operating also as can consume excess electricity when supply is higher than demand. The H2 could be used as transport fuel for the small road vehicle fleet of the island, and even for the fishing boats transport fuel. RO desalination plants and electrolysers could operate as programmable loads that can consume excess electricity when supply is higher than demand

Unless such methods of exploiting excess energy are exploited, RES penetration cannot be further economically increased for the cases studied. A scenario with lower RES penetration target, e.g. 50% for La Graciosa will provide lower benefits for the grid in the emissions levels and fuel consumption, but with significantly lower excess energy. Such a configuration can be used a first step in introducing RES on this Spanish island and during the construction phase to consider ways of exploiting additional excess electricity and apply demand side measures for smoothening the peak of the island. Then the additional works than can lead to much higher penetration, always considering storage, can be constructed increasing the expected benefits.



82

5 SIMULATION WITH PUMP HYDRO STORAGE 5.1 Introduction

The analysis is consisted on the following main steps: • Definition of the wind installed capacity outside the WPS

• Pre-feasibility study of WPS

• Evaluation of existing reservoirs

• Feasibility study of WPS (Dimensioning / Cost estimation / Benefits)

• Conclusions - Discussion

5.2 Methodology 5.2.1 Definition of the Wind installed capacity outside the WPS

Several islands are characterized by large wind-energy potential (i.e. the Greek islands [24]) and therefore there is a much interest by investors for wind applications. Wind farms operating in such autonomous systems are subject to output power limitations, related to technical constraints of the conventional generating units, namely the minimum loading levels of the thermal units (technical minimum) and a dynamic penetration limit, applied for stability purposes [32, 38].

However, the lack of interconnection introduce a restriction on the wind capacity which is permitted to be installed. In Greece, Ministerial Decision 8295/95 determined a maximum limit of installed capacity for non-interconnected islands. The limit was set at 30% of the previous year’s maximum hourly average capacity demand. The same Ministerial Decision gave the right to the System Operator to curtail wind farm production during hours of low demand, (ensuring however a minimal number of hours of operation for the wind farms – 6000 “Conventional Hours of Operation”). After the abolishment of Ministerial Decision 8295/95, RAE established a new methodology for determining the maximum allowed installed capacity in the non-interconnected islands, and the issuing of permissions for the production of RES electricity in the islands. According to the new rules, the limit of new capacity that can be installed and the extent to which production can be constrained will be determined for each island separately and will be revised every two years. Following the revision, deadlines for the submission of applications for generation of electricity from RES will be announced. The capacity and energy penetration limits will be calculated so as to ensure a minimal capacity factor of the order of 27.5%, taking into consideration the new loads, the existing conventional stations and RES projects, and the wind speed data of each island. It is worth mentioning that this procedure was put in place for the first time in Spring 2003.

The maximum installed capacity is defined so the wind power absorbed provides at least a ‘reali

capacity factor’ii which could provide the feasibility of the plant given the current total investment

cost of 1200-1500€/(kW capacity) [43], and the various incentives provided in the various countries.

For example in Greece, the real capacity factor after the wind curtailment in islands should be at least 27.5%iii, given the fixed price for wind power of 0.08€/kWh, paid by the Electricity System Operator (ESO) to the investor and the provided 30% subsidy. Crete was one of the first isolated systems with significant wind penetration which faced the curtailment of surplus wind. In 2000, the

i (real capacity factor)= (capacity factor)*(percentage of annual wind energy absorbed) ii Capacity factor is the average power production as the percentage of the nominal capacity iii i.e. a real Capacity Factor of 27.5% is likely to assure the feasibility of the investment



83

wind installed wind capacity reached 67 MW in Crete, with the wind ivcontribution to the electricity supply about 10% and the annual wind energy curtailed about 1.5% [33]. With the corresponding reduced income, the wind farm owners had significant financial losses [34,35]. By the end of 2006, the wind installed capacity in Crete reached the 130 MW and the annual curtailment was about 10%.

The objective of the current approach is to define, which are the limits of the wind penetration in the three case-studies islands (power system of Paro-Naxia, Cyprus and Corvo), without storage.

The analytical simulation of various case studies in Greek autonomous islands [42] shows that a higher penetration of wind energy could be achieved in the autonomous islands. The level of the allowed wind penetration is different for each island and it depends on the following technical and economical factors:

• the wind potential

• the duration curve of the demand

• the correlation between wind data series and demand

• the order of commitment of conventional units,

• the type of conventional units and their technical minimums

• the allowed instantaneous wind penetration (δ)

• the current electricity production cost of the system, which is always dependent on the oil price.

• the tariffs for the wind energy produced

• the subsidy or grants provided for such investments

• the current cost of the wind investments

The target should always be the maximization of the benefits to the electrical system. For example a better price for wind power should be given by the ESO to the investors, in case that a higher wind penetration could be useful in a system with higher current cost. By this way the feasibility of the investment is assured regardless of the lower real capacity factor and the higher curtailment of wind power surplus. Additionally, higher benefit occurs for the ESO due to the substitution of expensive conventional production.

In the following diagrams, general charts which have been reproduced by the systematic application in various case studies [36], are presented. The analysis is based on the convolution of annual hourly time-series of demand and wind. The required data for the simulation are the characteristics of the electricity demand, the conventional units and their features (maximum ability, technical minimums, order of commitment, fuel consumption, installation cost, operation and maintenance cost, fuel cost) and the wind potential of the island. These diagrams could provide a first approach for the wind penetration which can be achieved in an autonomous power supply system. The target is to determine the maximum wind installed capacity which can be installed in an autonomous island, given the current electrical infrastructure and without any storage facilities.

The main inputs for the use of these charts are: • the wind potential (average wind speed)

• the allowed instantaneous wind penetration (i.e δ=30% is a typical value, although a slight higher value can be allowed in larger systems, with a relatively geographical dispersion over a wider area of wind farms.



84

• The minimum real wind capacity factor which provides the feasibility of the wind investments. For example in Greece a 27.5% is at least the real capacity factor which is desirable for the investors in Greece given the tariffs for electricity produced by wind farms and the subsidy provided for such investments. In Cyprus, given the lower average wind potential, better tariffs are given, and then a lower capacity factor of the order of 20% could be considered enough for the feasibility of wind investments. Finally, in Corvo thanks to the high wind potential a capacity factor of the order of 30% could be feasible for wind investments, even without any incentives.

The outputs of these charts are: • the wind capacity to be installed in a system, ensuring the safe operation of the system,

• the percentage of wind energy absorbed / curtailed.

• the wind energy contribution

In this stage of the analysis, the several administrative difficulties to achieve the license for the installation of wind farms or land planning issues which may restrain the implementation of wind farms are not examined. The WPS is introduced as a mean to further increase the development of wind energy. Then, a significant degree of wind penetration is considered to be implemented in all the analysed islands, before the WHPS integration.



85

δ=30%

6.3m/s

7.2m/s

δ=40%

8.1m/s

δ=50%

9.0m/s

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

0% 50% 100% 150% 200%wind installed capacity (dimensionless)

Cap

acity

Fac

tor

δ=30%

δ=40%

δ=50%

Fig. 5.1 Diagram for estimation of the real capacity factor of wind farms in autonomous islands as a

function of the wind installed capacity (dimensionless by the mean annual load) and parameterized by the mean annual wind speed (6.3, 7.2, 8.1, 9.0m/s) and the allowed instantaneous wind penetration

(δ=30%, 40%, 50%).



86

Fig. 5.2 Diagram for estimation of the wind energy supply in autonomous islands as a function of the

wind installed capacity (dimensionless by the mean annual load) and parameterized by the mean annual wind speed (6.3, 7.2, 8.1, 9.0m/s) and the allowed instantaneous wind penetration (δ=30%,

40%, 50%).

δ=30%

6.3m/s

δ=50%

7.2m/s

δ=40%

8.1m/s

9.0m/s

0%

5%

10%

15%

20%

25%

30%

35%

40%


% w

ind

ener

gy s

uppl

y

δ=30%

δ=40%

δ=50%



87

δ=30%

6.3m/s

6.3m/s

9.0m/s

δ=50%

7.2m/s

8.1m/s

7.2m/s

δ=40%

8.1m/s

7.2m/s

8.1m/s

9.0m/s

6.3m/s

9.0m/s

20%

30%

40%

50%

60%

70%

80%

90%

100%


Win

d en

ergy

abs

orbe

d

δ=30%

δ=40%

δ=50%

Fig. 5.3 Diagram for estimation of the wind energy absorbed in autonomous islands as a function of

the wind installed capacity (dimensionless by the mean annual load) and parameterized by the mean annual wind speed (6.3, 7.2, 8.1, 9.0m/s) and the allowed instantaneous wind penetration (δ=30%,

40%, 50%).



88

5.2.2 Operation and architecture of the WHPS Different options related with the design and the operation of the WHPS have been analysed,

compared and evaluated [37]. As a result the features that are used for the study of the islands include:

• connection of the Wind farms with the pumping station through the central grid,

• peak demand supply of the hydro turbine,

• consideration of the hydro-turbine as a spinning reserve to increase the direct wind power absorption,

• double penstock and

• complementary pumping using the available conventional power (the amount of conventional units’ spinning reserve is taking into consideration).

5.2.2.1 CONNECTION OF THE WHPS WITH THE ELECTRICAL SYSTEM The connection of the wind farms with the Pumped Storage unit (PSU) is proposed through the

central grid, under the condition that the pumping loads are considered as deferrable loads. This means that in case of wind loss or other stability problem, pumps are disconnected.

Additionally, wind power from the WHPS can be directly absorbed by the electrical system, according to the technical constraints imposed by the Electrical System Operator (ESO). The amount of wind power, which can be absorbed directly by the grid, is dependent on the wind installed capacity outside of the WHPS, and on the allowed instantaneous wind power penetration “δ”. As a result, the wind installed capacity outside of the WHPS should be defined, before going on with the analysis of the WHPS. The wind power absorbed in priority from these wind farms is defined by the technical constraints described by the Regulatory Authority for Energy [38] and the WHPS integration will not effect their operation.

The pumping station should be also directly connected to the main grid, in order to use surplus wind power in priority and sometimes conventional power for complementary pumping.

In Fig. 5.4, the proposed structure and the interconnections of the whole electrical system after the WHPS integration is presented.

Grid

WPS (Hybrid system)

Conventional Units

Wind Farms

Pumped Storage/Hydro turbine

Load Wind Farms outof the WPS

Fig. 5.4 Structure of the electrical system after the WPS integration

5.2.2.2 OPERATIONAL TARGET OF THE HYDRO TURBINE The objective of the hydro turbine is to provide peak demand supply. Since there are seasonal and

daily ups and downs of the demand, and such an operational cycle of the hydro-turbine is needed to provide the financial feasibility, the daily (and not the weekly or the seasonal) peak demand supply is used. So, the hydro-turbine is setting into operation when the demand exceeds a predefined level which is not stable during the year, but follows the daily peak of the demand:

( ) Peak1L A ⋅−= α (Eq.1)



89

where Peak is the last 24-hours peak and α is the upper part of the demand to be covered by the

hydro-turbine. The ESO, who monitors the electricity demand, is proposed to be responsible for the assignment of the turbine’s set-point to the WHPS’s operator (WHPSO), when the demand D exceeds the LA. By this way the set-point of the turbine is defined as

ALD −=TSP (Eq.2) The reliability of the WHPS is measured by the hydro-turbine’s ability to correspond at the set-

points. The required rated power of the turbine is defined by the α and the annual peak demand.

5.2.2.3 CONVENTIONAL POWER FOR PUMPING Conventional power is proposed for complementary pumping, in order to provide the unfailing

weekday operation of the hydro-turbine and avoid the over-dimensioning of the reservoir. The use of conventional power for pumping is going to increase the total demand and the required conventional production, so the available conventional power for pumping should be defined by the ESO. For this purpose the assignment of a set-point by the ESO is suggested. The use of the available conventional electricity for pumping is then decided by the WHPSO considering the level of the water in the higher reservoir and the availability of the surplus wind power.

The preferable definition of the set-point for conventional pumping takes into consideration the spinning reserve of the committed conventional units [37]. This definition permits the use of conventional power for pumping only in case of conventional power surplus, so further conventional units are not committed to cover pumping loads. The advantage of this proposal is that the conventional units operation is improvedv and the available conventional power for pumping is widely distributed during the day.

5.2.2.4 WIND PENETRATION The ability of the conventional power stations to balance out both the variability of the demand

and the wind power, defines the wind power to be directly absorbed by the grid. Wind power is absorbed in priority by the wind farms outside of the WHPS system and then, if there is a margin for further wind power absorption, by the wind farms in the WHPS system.

Before the simulation of the electrical system, the wind penetration which is allowed should be defined. A stable maximum instantaneous wind penetration is used (i.e. δ = 30%)vi, while the wind penetration is further increased considering the hydro-turbine as spinning reserve. This operation presupposes the two-sided communication between ESO and WHPSO. Specifically, the ESO should know the amount of the hydro-turbine’s spinning reserve, which is dependent both on the turbine’s capacity and on the water availability in the higher reservoir, in order to allow equal increase of wind penetration. Finally, the turbine should be committed; otherwise the required time between the possible wind loss and the hydro-turbine response may cause a stability problem in the autonomous electrical system.

5.2.2.5 DOUBLE PENSTOCK Double penstock is used providing operational flexibility and the direct quick response of the

turbine when it is needed. The same time that the stochastic and variable wind power production is curtailed due to the technical constraints imposed for the safe operation of the electrical system and pumps should be set into operation, the system may need the uninterrupted and scheduled production of the hydro-turbine. This can be achieved only with the two penstocks.

v Conventional units operate more efficiently in full load. vi As a percentage of the load demand



90

5.2.3 Methodology - Simulation 5.2.3.1 ENERGY-POWER BALANCE

For the simulation the design parameters of the plant should be known: hydraulic head H, length LP and diameter DP of the penstock, capacity of the higher VH and lower reservoir VL, rated power of the wind farms PW,R, PW,h,R (with the index h the wind farms inside the WHPS are distinguished from the wind farms outside the WHPS), rated power of the turbine PT,R, number of pumps NP and rated power of each pump PP,R and the part of the peak to be supplied by the turbine α. An overview of the basic calculations is presented in Table 5-1.

Table 5-1 Overview of the basic calculations

Input Calculation / assumptions Output 1 U A common power curve for the wind farms is used. PW,h, PW

2 D (Eq. 1 and 2) Constraint: The SPT can not be lower than the technical minimums of the turbine i.e. TMT=30%PT,R. So if TA TMLD <− then 0=TSP .

SPT

SPT, LH, LH, VH, VL,

⎭⎬⎫

⎩⎨⎧

=>−

=000

T

TTT,RT if SP,

if SP,SPPSR

(Eq.3) Constraints considered: -water availability in the higher reservoir / level of the water in the lower reservoir

SRT

3 D, SPT, A strict order of commitment for the conventional units is considered.

TC D-SPMA ≥ (Eq.4) NC, TMC, MAC

4vii D, TMC, δ

CTAbsorbed D, TMδSRP ⋅+= min (Eq.5)

⎭⎬⎫

⎩⎨⎧

≥<

=→WAbsorbedW

WAbsorbedAbsorbedAW Pif P,P

Pif P,PP

(Eq.6a) AWWCW PPP →→ −= (Eq.6b)

⎪⎭

⎪⎬

⎫

⎪⎩

⎪⎨

⎧

<

>⎭⎬⎫

⎩⎨⎧

≥−<−−

=→

→→

→

WAbsorbed

WAbsorbedW,hAWAbsorbedW,h

W,hAWAbsorbedAWAbsorbed

AW,h

Pif P,

P if P,PPif P,P

PP, if PPPP

0 (Eq. 7a)

AW,hW,hCW,h PPP →→ −= (Eq.7b)

PW,h→Α, PW→Α, PW,h→C, PW→C

5 D, PW,h→Α, PW→Α, SPT AW,hAWTC -P-PD-SPSP →→= (Eq.8) SPCU

6 PW,h→C, SPC, MAC,

CCP SPMASP −= (Eq.9)

PCW,hP:Avail SPPP += → (Eq.10) ΡP:Avail

7 LH, LL, VH, VL, PP,R, NP,

Constraints considered: -number of pumps and rated power of the pumping station -level of the water in the higher reservoir / water availability in the lower reservoir

PP:Final

8 PP:Final ⎭

⎬⎫

⎩⎨⎧

<==

≥−==

→

→→→

)Pif (P,P,PP)Pif (P,PPP,PP

RW,hP:FinalP:CP:FianlP:W

RW,hP:FinalRW,hP:FianlP:CRW,hP:W

0

(Eq.11)

PP:W PP:C

9 ΡP:Final Operational curves of the pumps and the penstock. VL→H

10 SPT, LH, LL, VH, VL,

Constraints considered: -Technical minimum of the hydro-turbine -water availability in the higher reservoir / level of the water in the lower reservoir

PT,Produced,

11 PT,Produced Operational curves of the turbine and penstock. VH→L

vii Ιn case that D>++ →→ AWAhW,T PPSP , which could happen when α+δ>1, an iterative procedure is used before the required SPT and the NCU are defined.



91

12 LH, LL, VL→H, VH→L,

LHHLHH,END VVLL →→ −+= (Eq.12)

LHHLLL,END VVLL →→ +−= (Eq.13) LH,END, LL,END,

The main steps of the calculations are: • For every time-step (i.e. 1 hour) the water level in the higher LH and lower reservoir LL

viii, the electricity demand D and the wind velocity U are known. The wind power production of the wind farms PW,h and PW is calculated given the wind installed capacity PW,h,R and PW,R.

• The hydro-turbine’s set-point SPT (Eq.1 and 2) and the number of conventional units committed NC are calculated by the ESO (Eq.4). The cumulative maximum ability of the committed conventional units MAC and the hydro-turbine should be able to meet the demand even if all the wind power is lost, providing the safe operation of the electrical system.

• Given the turbine’s set-point, the amount of the hydro-turbine’s spinning reserve SRT (Eq.3) is taken into consideration to increase the wind penetration. The SRT is calculated given the hydro turbine’s rated capacity PT,R, the turbine’s production SPT and the available water in the higher reservoir LH. The ability of the electrical system to absorb wind power is increased by the SRT (Eq.5).

• The wind power absorbed (PW,h→Α, PW→Α) and curtailed (PW,h→C, PW→C), are calculated (Eq. 6a, 6b, 7a, 7b).

• The conventional units set-point SPC (Eq. 8) and the available conventional power for pumping SPP (Eq.9) are calculated. So, the total available power for pumping ΡP:Avail is derived (Eq.10).

• The final power used for pumping ΡP:Final is constrained by the rated power of pumps (PP,R, NP) the capacity of the reservoirs (VH, VL) and the available water. The part of the ΡP:Final which is derived from the wind farms or the grid (ΡP:W, ΡP:C) is calculated (Eq.11). Then, the water flow from the lower to the higher reservoir VL→H is calculated given the operational curves of the pumps and the penstock.

• The desirable hydro-turbine production is defined by the set-points sent by the ESO. The final turbine’s power production PT,Produced depends on the availability of the water in the upper reservoir. Then the water flow from the higher to the lower reservoir VH→L is calculated given the operational curves of the turbine and penstock.

• Finally at the end of each step the level of the higher and lower reservoirs (LH,END, LL,END), are calculated (eq. 12, 13).

When all the above calculations for all the time-steps (8760hours) are completed, then the following annual energy amounts are derived:

• desirable hydro-turbine’s production ESP_T,

• final hydro-turbine’s production ET,

• production of wind farms (EW, EW,h, EW,Total),

• wind energy absorbed by the grid directly (EW→A, EW,h→A),

• wind energy curtailed (EW→C, EW,h→C),

• conventional energy available for pumping ESP_P,

viii For the first step of the simulation, an initial water level in the two reservoirs is needed. The independence of the results is proved, using different initial water levels.



92

• total energy available for pumping (conventional and curtailed wind energy) EP:available

• energy used for pumping: cumulative, conventional and wind power (EP:final, EP:C and EP:W,h)

5.2.3.2 EVALUATION CRITERIA The contribution of the WHPS, together with economical and reliability indexes, are used to

describe the performance of the electrical system after the WHPS integration. The conventional units’ EPC EPCC, the Electrical system’s EPC EPCS and the turbine’s EPC EPCT are used to describe the economic impact of the WHPS to the electrical system. The most critical is the EPCS, when it is compared with the current cost, the resulting benefit -if any- from the WHPS integration is defined. The EPCT is important for the private investor, indicating a first estimation of the required price for the turbine’s electricity production which provides the feasibility of the investment. Finally, the modification of the EPCC due to the WHPS integration is critical for the ESOix in order to accept this price.

The electricity production cost of the turbine EPCT is defined under the assumption that the whole investment is considered as a mean to provide guaranteed electricity supply during peak demand, so the wind energy sold in a fixed price is considered as inflow:

T

WCWPSWPST E

-BECEOMCRTICEPC

++⋅=

(eq.14)

where TICWPS is the total investment cost, OMCWPS the Operation and Maintenance cost of the WHPS, CEC the cost of conventional energy used for pumping and BEW the benefit from the wind energy directly absorbed by the grid. If the market price is pm, then

P:CmC EpCE ⋅= (eq.15)

and if the fixed price for wind power is pw, then

AW,hWW EpBE →⋅= (eq.16)

The degree of accomplishment of hydro-turbine’s set-point DA_SPT is the index to measure the reliability of the WHPS. This index is highly dependent on the design of the WHPS; namely the wind installed capacity, the capacity of the reservoir, and the operational target of the turbine. It is defined as the ratio of the actual energy production ET to the desirable production defined by the hydro-turbine’s set-points (ESP_T):

SP_T

TT E

EDA_SP = (eq.17)

The electricity production cost of the conventional units EPCC is defined as:

C

CCC E

OMCRTICEPC

+⋅=

(eq.18)

where TICC is the total investment cost of the essential conventional units, OMCC the Operation and Maintenance cost, and EC the conventional energy production. The OMCC has a fixed cost part, a variable cost part and the fuel cost.

FuelCoststVariableCoFixedCostOMC ++=C (eq.19) The electricity production cost of the electrical system EPCS is calculated, assuming that the

redundant units are uninstalled:

TotalS E

OMCRTICEPC +⋅= (eq.20)

ix The autonomous islands are excluded from the market liberalization and the system operator remains the owner of the local power stations.



93

where TIC includes the cumulative investment cost of all the power plants (essential conventional units, WHPS and wind farms outside the WHPS); the OMC includes the fixed cost, the variable cost and the fuel cost for the operation and maintenance of the system; and ETotal is the total electricity demand.

The annuity factor R is defined as:

Nii

−+−=

)1(1R (eq.21)

where, i is the discount rate and Ν the lifetime of the investment.

5.2.3.3 SIMULATION OF THE HYDRO SUBSYSTEMS (PUMPS, PENSTOCKS, HYDRO TURBINE) The problem of the simulation of the pumping station can be described as “Which is the required

power to commit i pumps and which is the flow of the water”, and in the case of the hydro-turbine’s simulation as “Which is the required volume of the water to produce the required power”

The pump and turbine characteristics (Head-flow curve and efficiency-flow curve) are introduced dimensionless (Fig. 5.5 and Fig. 5.6). The presented curves were assumed as representative for a wide range of centrifugal pumps and Pelton turbines with different nominal characteristics [39].

0%

20%

40%

60%

80%

100%

120%

0% 50% 100% 150% 200%Qdimensionless

η dim

ensi

onle

ss

0%

20%

40%

60%

80%

100%

120%

140%

160%

0% 50% 100% 150% 200%Qdimensionless

Hdi

men

sion

less

Fig. 5.5. Dimensionless Efficiency-flow and Head-Flow curves for a representative centrifugal pump

0%

20%

40%

60%

80%

100%

120%

0% 20% 40% 60% 80% 100% 120%

Qdimensionless

ηdim

ensi

onle

ss

Fig. 5.6. Dimensionless Efficiency-flow curves for a representative Pelton turbine

The operation point (Fig. 5.7) of the pumping station is calculated as the intersection of the pumps characteristic (H, Qi), where i is the number of pumps in operation and the pipeline characteristic HP(Q).

( ) 2QHQH P ⋅+= ζ (eq.22)

where ζ is the loss coefficient of the penstock, H the hydraulic head and Q the water flow.



94

The operation point of the hydro-turbine is calculated as the intersection of the hydro-turbine characteristic (H, Q), and the pipeline characteristic HP(Q).

( ) 2QHQH T ⋅−= ζ (eq.23) The loss coefficient is calculated:

221

AgDL

P ⋅⋅⋅= λζ (eq.24)

Where, λx is the dimensionless loss coefficient, LP the length of the penstock, DP the diameter of the penstock, and A the area of the penstock. For the calculation of the λ the formula of the Colebrook and White is used:

⎥⎦

⎤⎢⎣

⎡+

⋅⋅−=

71,3Re51,2log21 sε

λλ (eq.25)

Where εs is the relative roughness of the penstock’s wall Dsεε =

and ε the absolute roughness A short iterative procedure is used to solve the above complex formula. The Reynolds is defined

as: vDc ⋅

=Re, where c is the water velocity in the penstock A

Qc = and ν is the cinematic viscosity of

the water )(1031.1 126 −− ⋅⋅= smν for water temperature 10˚C.

0%20%40%60%80%

100%120%140%160%180%200%

0%

100%

200%

300%

400%

500%

600%

700%

800%

900%

1000

%

Qdimensionless

Hdi

men

sion

less HΣ

(H,Q 1) (H,Q 2)

(H,Q 10)

Fig. 5.7. Definition of the operation point of the pumping station

5.2.3.4 DEFINITION OF THE PENSTOCK’S DIAMETER The diameter of the two penstocks for the turbine DP,T, and the pumping station DP,P is calculated,

assuming a maximum permitted water velocity Vmax, which occur for the maximum water flow Qmax. The following equations are applied for the penstocks of the pumping station and the hydro turbine.

21max

max4/P π)(V

QD

⋅⋅

= (eq.26)

( ) 50max 21250 .Hg.V ⋅⋅⋅= (eq.27)

In the case of the pumping station, the maximum water flow Qmax,P is defined by the number of pumps and the water flow of each pump in the nominal point of operation QP,R:

HNηgρ

NQNQ P,RPPP,RP,P

⋅⋅⋅⋅=⋅=max

(eq.28)

x The loss coefficient and the dimensionless loss coefficient are calculated for the different water flows.



95

Where ηP=0.85 is the efficiency of the pump in the nominal point. In the case of the hydro-turbine, the maximum water flow Qmax,T is defined by the hydro-turbine:

HηNgρ

.Q.QT

T,RT,R,T ⋅

⋅⋅⋅=⋅= 3131max

(eq.29)

Where ηT=0.90 is the efficiency of the hydro-turbine in the nominal point. 5.2.4 Pre-feasibility study

5.2.4.1 DIMENSIONING PROCEDURE In fact, given the wind potential, the wind installed capacity is the parameter which defines the

available for exploitation energy. The part of this energy which can be directly absorbed by the grid, depends on the technical constraints of the electrical system. The part of the surplus wind energy which can be finally exploited, depends on the architecture the design and the size of the WHPS. For a given wind installed capacity, there is a direct relation between the size of the reservoirs, and the maximum operational target that can be covered by the hydro-turbine with reliability reaching the 100%. The capacity of the hydro-turbine is then defined by this target.

The aim of the proposed approach is to take comparable results for the case-studies islands, so the basic parameters in issue are introduced dimensionless. Initially, the wind installed capacity was defined by the mean annual load demandxi, which is fair way to levelize wind installed capacity in different in size systems [40]. An advanced approach, says that the wind capacity should be proportional to the mean annual load demand, and it should reversely depended on the wind capacity factor and the efficiency of the hydro pumped storage. As a result, ten different values for wind installed capacity are examined from 5% to 95%, by step 10%, of the following amount:

PSUth,W

L

nCFP

⋅

(eq. 30)

Where, LP is the mean annual load demand, th,WCF is the theoretical wind capacity factor and PSUn the efficiency of the hydro pumped storage (typically 60%, for this calculation only). Similarly, the capacity of the reservoir should be defined proportionally to the average hourly

water pumped volume. Then, sixteen different values are examined from 140 to 335 by step 13, of the following amount (annual average hourly water pumped volume):

HnCFP1023600 Pth,WR,h,W ⋅⋅⋅⋅ (eq. 31)

Where, th,WR,h,W CFP ⋅ is the average wind production, Pn is an average pumping station efficiency (80%), and H the available hydraulic head.

For each wind installed capacity and reservoirs capacity, the maximum operational target that can be achieved by 100% is calculated using an iterative procedure. That is to say, that the maximum turbine size which is justified by the wind installed capacity and the reservoirs capacity should be defined. It is obvious that a bigger target could be set, but it would be achieved in less than 100% of the year. The target of the dimensioning is to guarantee the reliability of the turbine operation (achieve DA_SPT=100%), and conventional units are supposed to be removed. So, the conventional installed capacity is redefined by the peak demand, the supply of the turbine and a safe margin of conventional installed capacity back up (20%).

For each different value of wind installed capacity, there is an optimum value for the reservoir which provides the lower electricity production cost for the turbine operation (EPCT). By the investor’s point of view the target is the maximum contribution with the least cost. So, from 160

xi Annual electricity demand divided by 8760 hours



96

cases resulted -all of them provide the reliability of the turbine operation-, the 10 cases with the maximum target achievement and the least cost comprise the optimum solution (lower envelope curve of the EPCT’s curves) for the current island and scenario. An example is given for the island of Crete in figure 5, where the twenty different curves represent different wind installed capacity inside the WHPS (153-1314MW).

0.08

0.10

0.12

0.14

0.16

0.18

0.20

0% 20% 40% 60% 80%WPS share of the peak

Turb

ine'

s EP

C (€

/kW

h)

153

214275336397458519580641703764

825886947100810691130119112521314

0

20000000

40000000

60000000

80000000

100000000

120000000

140000000

160000000

0% 20% 40% 60% 80%

WPS share of the peak

Rese

rvoi

r's v

olum

e (m

^3)

153214275336397458519580641703764

825886947100810691130119112521314

Fig. 5.8. Example of the dimensioning procedure: a) Reservoir’s capacity, b) Turbine’s EPC (Crete)



97

1st loop for the wind installed capacity

2nd loop for the capacity of the reservoir (the examined values are dependent on the wind capacity)

Simulation for all the hours of the year (hour=1:8760) – analytically described in the table 2.

Calculation of the annual energy balance and indexes of evaluation

Reliability controlDegree of turbine’s set-point achievement =100%

Calculation of the required turbine capacity

Cost-effectiveness subroutine for the definition of the optimum solutions, which provide the maximum peak supply with the lower

turbine’s electricity production cost

3rd loop for the calculation of the maximum WHPS share of peak demand, which can be guaranteed, and the turbines required capacity.

Record of the examined case (PW,h,R, VR, PT,R) and the evaluation in the

various evaluation indexes

End of the loops

Recording of the optimum solutions (PW,h,R, VR, PT,R)

Dimensioning procedure of WHPS

Fig. 5.9. Simplified algorithm for the dimensioning procedure of the WHPS

5.2.4.2 REQUIRED DATA AND ASSUMPTIONS

5.2.4.2.1 Assumptions For the initial calculations of the pre-feasibility study, the following typical assumptions for

wind-hydro pumped storage hybrid systems are considered:: • Analytical data of the demand (hourly data for one year), of the conventional power units

and wind velocity (hourly data or mesoscale are used for each case study island.



98

• The representative wind velocity has been assumed 8.1m/s for the Paro-Naxiaxii (38% initial capacity factor), 6.3m/s in Cyprusxiii (22.3% initial capacity factor) and 9.4m/s in Corvoxiv (47% initial capacity factor).

• For all the examined case studies, the hydraulic head between the two reservoirs has been considered as H=400m, and the distance between them LP=3000m

• Financial evaluation without any subsidy

• The reference year of the analysis is the 2010. The demand and the required conventional power stations in 2010 have been estimated according to the rate of increase recorded and to the available data, during the last years

• As a reference price for oil the last years annual mean values varies from 30$/b up to 100$/b (Table 5-2). In Fig. 5.10, it is obvious that from 2007 until the middle of 2008, an excessive increase of the price occurred, while then an excessive decrease occurs. Then a reference value of the order of 60$/b could be considered as a conservative assumption for the lifetime of the project.

Table 5-2 Historical data of annual minimum, maximum and average price for Brent

year min max average 2000 22,23 33,56 29,33

2001 17,68 30,91 24,62

2002 18,17 31,74 25,12

2003 23,59 34,88 28,87

2004 28,64 51,95 38,28

2005 43,28 68,06 53,65

2006 55,60 76,73 65,11

2007 49,00 96,57 73,32

2008 34,08 143,33 97,71

xii Taking into consideration available measurements xiii Taking into consideration available measurements and the assumptions undertaken for the definition of

the tariffs-incentives for the development of RES in Cyprus. xiv Taking into consideration available measurements in two sites of the island.



99

Fig. 5.10.Brent Oil price and future estimations

• The Wind installed capacity out of the WHPS is considered as a percentage of the mean annual load. In Paro-Naxia it is considered as the 60% of the annual mean load, in Cyprus, as the 63% of the annual mean load and in Corvo, as the 65% of the annual mean loadxv. Further details on this assumption are presented in the first paragraph of each case-study.

• The diameter of the two penstocks is calculated given the rated capacity of the pumping station and the hydro-turbine

• Ten centrifugal pumps connected in parallel are used to provide operational flexibility [41]. Given the available power for pumping the numbered of committed pumps is calculated.

• The efficiency of the pumping station and hydro generation are analytically calculated for the different points of operation. The results show that the whole efficiencyxvi of the plants in the examined cases varies between 55% and 69%.

• For the estimation of the WHPS investment cost, empirical formulas have been used [42]. For the wind farms an investment cost of 1200€/kW has been considered [43]. An overview of the WHPS’s cost estimation is presented in Table 5-3

• The calculation of the conventional EPC is based on data of the installation cost, the fixed O&M cost, the variable O&M cost and the fuel consumption in the various operational points.

• The discount rate is considered i=5% and the lifetime N of the investment 20 years for the wind applications and the WHPS and 30 years for the conventional units.

• Market price for the conventional electricity used for pumping pm is 0.12€/kWh (much more expensive than the regular tariff of the consumer 0.09€/kWh), taking into consideration the higher marginal electricity production cost.

• Fixed price for the wind production pw is 0.08€/kWh.

Table 5-3 Overview of the formulas and assumptions for the WPS cost estimation

Equipment – Cost symbol Data/Formula for Cost Estimation (€) Wind Farms (CW) 1200 /kW

Pumps (CP) 820

300

,

,P

P,rated,PPP H

PCNC ⎟⎟

⎠

⎞⎜⎜⎝

⎛⋅⋅=

, 18140 =,PC Hydro-turbine (CT) 820

300

,

,T

T,rated,TT H

PCC ⎟⎟

⎠

⎞⎜⎜⎝

⎛⋅=

, 46870 =,TC Reservoir (CR) 70420 ,

R VC ⋅=

xv Always, as far as the wind potential is decreased, the wind curtailment is decreased but also lower wind

curtailment could be afforded by the wind farms owners. On the other hand, the higher the wind potential and the incentives-tariffs, the higher the wind curtailment that can be afforded by the investor. In Cyprus, for example the tariff scheme has been designed to provide the feasibility of the investments despite the low wind potential. xvi It is defined as the ratio of the annual energy produced by the hydro-turbine compared to the annual energy consumed for pumping.



100

Penstock (CPenstock)

∑

⎪⎪⎪

⎭

⎪⎪⎪

⎬

⎫

⎪⎪⎪

⎩

⎪⎪⎪

⎨

⎧

⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢

⎣

⎡

⋅⎟⎟⎠

⎞⎜⎜⎝

⎛⋅⋅

+⋅⋅+⋅⋅⋅⋅

⋅I

CostExcavation

EI

CostInsulation

II

ostMaterial C

MIIM

CLπD.

CL)(ππCL)eπD(W

.

444 3444 21

4 34 21444 3444 21

451

251 2

Grid connection (CGC) 4%*(CP+CT+CR+CPenstock) Control system (CCS) 1,6%*(CP+CT+CR+CPenstock) Transportation of equipment (CT) 2,4%*(CP+CT+CR+CPenstock) Personal (CP) 30%*(CP+CT+CR+CPenstock) Others (CO) 2%*(CP+CT+CR+CPenstock) Operation and Maintenance (OMCWPS) 2%*(CP+CT+CR+CPenstock+CW)

5.2.4.2.2 Electricity demand For the simulation, the analytical power demand data (hourly time series for one year) are

required. The load factor, shows how fully the annual duration curve of the load demand data is, and it is defined as:

)MW(PeakDemand)h(8760)MWh(gyDemandAnnualEner

⋅=LF(%)

The load factor in 2007 is 37.6% in Paro-Naxia, 58.2% in Cyprus and 60.3% in Corvo. It is very useful, for the design of the systems and the understanding of the results, to figure the

duration curve of the dimensionless load demand. Two approaches have been implemented: • comparison of the duration curve of the dimensionless -by the peak demand- load (Fig.

5.11)

• comparison of the duration curve of the dimensionless -by the annual mean demand- load (Fig. 5.12)

cypruscorvo

paros0%

20%

40%

60%

80%

100%

120%

0 2000 4000 6000 8000

cypruscorvoparos

Fig. 5.11. Load duration curve (dimensionless by the peak demand)



101

cyprus

corvo

paros

0%

50%

100%

150%

200%

250%

300%

0 2000 4000 6000 8000

cypruscorvoparos

Fig. 5.12 Load duration curve (dimensionless by the mean load demand)

5.2.4.2.3 Conventional power units All the examined islands are based on diesel power stations. The technical minimums of such

units are always around 40-50%. The required conventional capacity is higher than the annual peak demand by at least 10%. Higher cool reserve is needed for smaller islands due to the small number of conventional units. This has a great importance to the reliability of the electrical system. The probability analysis shows that as soon as the number of units is increased, the reliability is increased, assuming a same degree of cool reserve [44].

5.2.4.2.4 Wind potential The wind potential maps and the maps of economical exploited wind potential gives a first idea

of the wind potential and the sitting of the wind farms in the islands. These maps have been created from [24], taking into consideration restrictions of slope, land use, restricted areas, urban areas, etc.

Analytical timeseries of wind data are required for the simulation of the WHPS. The use of wind atlases is not a solution since they only provide an estimate of the spatial distribution of the mean wind speed without any information on its temporal variation.

5.3 Case study 1 results-Ios 5.3.1 Adding wind without Pump Hydro Storage

Using the methodology explicitly described in section 5.2 and more specifically diagram of Fig. 5.1, for representative mean annual wind speed 8.1m/s, for allowed instantaneous wind penetration 30% and desirable real capacity factor at least 27.5% (initial capacity factor for 8.1m/s is 22.3%), the maximum allowed wind installed capacity without storage is estimated to the 60% of the annual mean load. The electricity demand is 170GWh or 20.4MW annual mean load. Then, the maximum allowed wind installed capacity without storage is estimated to 12.2MW. The energy supply of this wind installed capacity is 16.4% and the wind energy curtailment is 27.6%.

5.3.2 Scenario 2-Pump Hydro Installation 5.3.2.1 RESULTS OF THE PRE-FEASIBILITY STUDY

In Fig. 5.3 the results of the pre-feasibility study of WPS in the power system of Paro-Naxia are presented. Pre-feasibility study aims to examine the effect of the WPS integration into the autonomous power supply system. Then, in these charts the electricity production cost of the power system without pumped storage considering a Brent price of the order of 100$/b is presented in straight red line. This price is considered as a long term reference price.



102

The Market price for the conventional electricity used for pumping pm is 0.12€/kWh (much more expensive than the regular tariff of the consumer 0.09€/kWh), taking into consideration the higher marginal electricity production cost. Fixed price for the wind production pw is 0.085€/kWh in Greece.

The Market price for the conventional electricity used for pumping pm is 0.12€/kWh (much more expensive than the regular tariff of the consumer 0.09€/kWh), taking into consideration the higher marginal electricity production cost. Fixed price for the wind production pw is 0.085€/kWh in Greece.

The main conclusions of this approach in the power system of Paro-Naxia are: • For the large scale RES integration i.e. peak demand supply up to 50% or hybrid’s

electricity supply up to 55% (Fig. 5.13.a), and cumulative RES supply in the island up to 70% (Fig. 5.13.b), the required reservoirs volume is up to 13 million m3 (Fig. 5.13.i) and the required wind installed capacity is up to 100MW (Fig. 5.13.j).

• The electrical system’s EPC could be reduced (Fig. 5.13.d) as a result of the WPS integration (from the level considered with Brent price 100$/b).

• The conventional units EPC is increased (Fig. 5.13.e) as soon as the same units operate less after the WPS integration and their low load factor is further decreased (Fig. 5.13.g).

• As regards, the turbine’s and hybrid’s EPC, the former varies around 0.18-0.22€/kWh and the latter around 0.17-0.21€/kWh (Fig. 5.13.c). This cost is comparable to the peak supply cost, even a lower Brent price is considered. The turbine’s electricity production cost is independent to the Brent price. Then, the profitability of this investment is strongly depends on the tariffs for the turbine’s electricity production.

• The investment cost of the WPS is estimated to 1800-2700€/kW installed (cumulative wind and hydro turbine capacity) or 6000-9000€/kW of turbine capacity (Fig. 5.13.m, Fig. 5.13.n). Lower cost is estimated for larger WPS systems.

• The efficiency of the WPS plant is estimated to 55-67% (Fig. 5.13.h).

• The RES supply in the Paro-Naxia can be significantly increased thanks to the WPS integration. Around 20% RES supply could be achieved if wind installed capacity is introduced without any storage devices. For further RES supply, the WPS integration could contribute up to an additional 50% (or up to 70% cumulative RES supply), for the examined cases of this study (Fig. 5.13.b).

• For this achievement the wind installed capacity in the WPS should be increased. Unfortunately, only the 45-60% of the wind energy is exploited (Fig. 5.13.l). There are several cases where the wind energy is not exploited due to the fulfillment of the upper reservoir.

On the other hand the conventional power which is used for pumping is always lower than 20% (Fig. 5.13.k).



103

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

0% 20% 40% 60% 80% 100%WHPS share of the peak

Turb

ine'

s E

PC

(€/k

Wh)

0,26

0,27

0,28

0,29

0,30

0,31


Elec

trica

l sys

tem

's E

PC

(€/k

Wh)

0%

5%

10%

15%

20%

25%

30%


Load

Fac

tor o

f con

vent

iona

l un

its

0,00

0,05

0,10

0,15

0,20

0,25

0% 20% 40% 60% 80% 100%

WHPS share of the peak

Hybr

id's

EP

C (€

/kW

h)

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35


requ

ired

pri

ce fo

r IR

R=8%

0,15

0,20

0,25

0,30

0,35

0,40

0,45

0,50


Conv

entio

nal u

nits

' EP

C (€

/kW

h)

Power system of Paros

0%

10%

20%

30%

40%

50%


Hybr

id's

ene

rgy

supp

ly

(%)

0%10%20%30%40%50%60%70%80%90%

100%


% r

es s

uppl

y

0100000020000003000000400000050000006000000700000080000009000000

10000000


Rese

rvoi

r's v

olum

e (m

^3)

0%

10%

20%

30%

40%

50%

60%

0% 20% 40% 60% 80% 100%


% c

onve

ntio

nal p

ower

for

pum

ping

a) b)

c) d)

f) e)

g) h)

j) i)



104

0%10%20%30%40%50%60%70%80%


effic

ienc

y of

the

hydr

o pu

mpe

d st

orag

e sy

stem

0%

10%

20%

30%

40%

50%

60%

70%

0% 20% 40% 60% 80% 100%


% w

ind

ener

gy e

xplo

ited

0100020003000400050006000700080009000

10000

0% 20% 40% 60% 80% 100%

WHPS share of the peakco

st o

f the

pla

nt p

er M

W o

f gu

aran

teed

pow

er (€

/MW

)

0

500

1000

1500

2000

2500

0% 20% 40% 60% 80% 100%


cost

of t

he p

lant

per

MW

of

inst

alle

d ca

paci

ty (€

/MW

)

0,0

20,0

40,0

60,0

80,0

100,0

0% 20% 40% 60% 80% 100%


Win

d ca

paci

ty (M

W)

Fig. 5.13 Results of the pre-feasibility study in Paro-Naxia complex: a) hybrid’s energy supply, b) total RES electricity supply in the island, c) Turbine’s EPC, d) Electrical system’s EPC, e)

Conventional units’ EPC, f) Required price for the turbines electricity production to provide IRR=16%, g) Load factor of conventional units, h) Hybrid’s EPC, i) Conventional power for pumping,

j) Reservoir’s volume, k) efficiency of the pumped storage system, l) wind capacity (MW), m) wind energy exploited (%), n) cost of the plant per MW of guaranteed power, o) cost of the plant per MW of

installed capacity

5.3.2.2 EVALUATION OF EXISTING RESERVOIRS In the complex of Paro-Naxia there are several existing reservoirs. Two of them (Fig. 5.14)Fig.

5.18) are considered as more suitable for the parallel exploitation towards a WPS integration: • Reservoir of Eggaron in Naxos

• Dam of Milopotamos in Ios (Fig. 5.14)

The reservoir in Naxos (the right side one), is in an altitude of 118m, while in a distance of 3160m, there is possible site of a second upper reservoir in an altitude of 910m. Then the available hydraulic head will be 790m.

The reservoir in Ios, is in an altitude of 70m, while in a distance of 2300m, there is possible site of a second upper reservoir in an altitude of 690m. Then the available hydraulic head will be 620m.

Although a higher hydraulic head is provided in Naxos, the development of the first WPS of Paro-Naxia is proposed to be realized in Ios. This is also justified due to the high rates of development in Ios, the distance between Ios and Paros (high losses) and the distance between the upper and lower reservoir in the case of Naxos

k) l)

n)m)

o)



105

Fig. 5.14 Photo of the Dam of Milopotamos in Ios (capacity: 215,000m3, surface: 36,000m2)

Fig. 5.15 Photomap of the existing reservoir (lower reservoir) in Naxos

The reservoir in Ios, is in an altitude of 70m, while in a distance of 2300m, there is possible site of a second upper reservoir in an altitude of 690m. Then the available hydraulic head will be 620m



106

Fig. 5.16 Photomap of the existing reservoir (lower reservoir) in Ios



107

Fig. 5.17 Map with the proposed site for the installation of the WHPS in Naxos

Fig. 5.18 Map with the proposed site for the installation of the WHPS in Ios



108

5.3.2.3 FEASIBILITY STUDY The system of Paro-Naxia has a special interest, because it is consisted of five main islands

(Paros, Naxos, Ios, Sikinos, Folegandros) and some smaller and the only local power station is located in Paros, which is the first in order island. As a result this system is considered as a weak and centralized system with high energy transportation losses and problems of stability.

Additional constraints are imposed on the design and the dimensioning of WPS in Ios due to the capacity of the existing underwater cables between the islands.

The dimensioning in this case, starts from the existing reservoir. Taking into consideration that the existing reservoir is also used for irrigation, then only a part of the reservoir’s capacity can be used for energy use. The topography of the site of the upper reservoir imposes an additional constrain. Then, the useful capacity of the reservoirs has been considered 120.000m3, to start with the simulation and the dimensioning procedure. The hydraulic head is 620m and the distance between the two reservoirs is 2300m.

The rest components of the WPS has been then estimated: • Hydro turbine: 8MW,

• Wind Farms’ capacity: 8MW

• Upper reservoir: 120.000m3 - Lower reservoir (existing): 260.000m3

• Pumping station: 6,5MW (10 pumps)

The general design of the WPS is compromised: • double penstock

• use of the turbine as a spinning reserve to increase the direct wind absorption

• hydro-turbine’s supply of the daily demand peaks (stable daily energy production, distribution according to the needs)

The main effect on the contribution to 2010 of this plant is estimated to 7.9% of the electricity demand. A small part is derived by direct wind absorption (1.9%) and an other 6% is derived by the hydro turbine supply. The “Guaranteed power” is 8MW or the 7% of the annual peak demand.

The results from the simulations are summarized in Table 5-4, while Table 5-5 presents the energy feautures.Graphically the results from the simulation are presented in Fig. 5.19.



109

Table 5-4 Results of the simulation for the power system of Paros – development of WHPS in Ios

Island Ios Year of simulation 2010 Average wind speed at the hub height (m/s) 8,2 Scenario of operation 7 Hydro turbine's share of the peak 10,7% hydraulic head (m) 620 Length of the penstock (m) 2300 Main parameters of dimensioning Rated power of the turbine (MW) 8,0 Rated power of wind farms (in the hybrid) - MW 8,0 Rated power of wind farms (out of the hybrid) - MW 18,3 Rated power of pumps - MW 6,5 Capacity of the reservoir - m3 120000 Energy calculations Peak - MW 74,8 Minimum demand - MW 11,8 Annual energy demand - GWh 246,3 Mean annual load - MW 28,1 Wind energy production (out of the hybrid) - GWh 60,7 Wind energy absorbed (out of the hybrid) - GWh 43,1 Wind energy curtailed (out of the hybrid) - GWh 17,6 Wind energy production (in the hybrid) - GWh 26,5 Wind energy absorbed (in the hybrid) - GWh 6,9

Wind energy curtailed (in the hybrid) - GWh 19,7 Cumulative wind energy produced 87,3 Cumulative Wind energy absorbed - GWh 49,9 Cumulative wind energy curtailed - GWh 37,3



110

Table 5-5 Energy features of the reverse hydro system

energy used for pumping given the number and rated of pumps 24,3 Wind energy rejected (out of the hybrid) 29,0% energy used for pumping given the number and rated of pumps, and the capacity of the upper reservoir 21,1 Wind energy rejected (in the hybrid) 74,2% energy used for pumping given the number and rated of pumps, and the capacity of the upper and the lower reservoir 21,1 Part of the rejected wind energy (in the hydrid), used for pumping 75,1%

cumulative final energy used for pumping 21,1 Part of the rejected energy (out of the hybrid), used for pumping 0,0% wind energy surplus (in the hybrid), used for pumping 14,8 Part of the energy used for pumping derived by conventional units 29,9%

wind energy surplus (out of the hybrid) used for pumping 0,0 Ratio of conventional power for pumping and of turbine's production 43,1%

conventional energy used for pumping 6,3 Theoretical Capacity factor of wind farms (out of the hybrid) 37,9% Turbine's production - GWh 14,6 Real Capacity factor of wind farms (out of the hybrid) 26,9% Renewable energy poroduction - GWh 8,3 Theoretical Capacity factor of wind farms (in the hybrid) 37,9% Conventional energy production - GWh 188,0 Real Capacity factor of wind farms (in the hybrid) 9,8% Degree of turbine's setpoint achievement 100,0% CF of wind farms in the hydrib (absorbed and used for pumping) 30,9%

Turbine's Capacity Factor 20,8% Efficiency of the reverse hydro scheme (turbine's production / energy used for pumping) 69,3%

Wind farm's capacity factor 30,9% Degree of conventional for pumping set-point achievement 7,52% wind energy used (%) 81,5% Water moved through the pumps to the upper reservoir - 1000m3 10031 Capacity factor of the upper reservoir 47,8% Water moved through the turbine to the lower reservoir - 1000m3 9982 Capacity factor of the lower reservoir 52,2% Energy consumed by the pumps - GWh 21,1

Maximum conventional production / peak demand 93,2% Energy produced by the hydro turbine - GWh 14,6 Wind energy contribution (out of the hybrid) 17,5% Dynamic energy of water moved through the pumps - GWh 16,9

Wind energy contribution (in the hybrid) 2,8% Dynamic energy of water moved through the turbine - GWh 16,9 Turbine's contribution 5,9% Pumping operation efficiency 80,4%

WHPS contribution 8,7% Turbine's efficiency 86,6% Cumulative RES contribution 23,7% Efficiency of the reverse hydro scheme 69,3%



111

The tariffs for the hydro turbine production, the pricing of guaranteed power and the price for the conventional power for pumping are objectives of negotiation between investor, system operator and regulatory authority (RAE).

The proposed WPS is in accordance with the current grid infrastructure and the underwater cables. The maximum aggregated production (wind and turbine) is 13.1MW (and it occurs in peak demand load). Only 77 hours annually the aggregated production exceeds 10MW, which is lower than the demand, and the ability of the two underwater cables between Ios-Paros (~10MW) and Ios-Sikinos (~5MW).

WPS integration in Ios contributes to the decentralization of the electrical system, with positive effect on the stability, reduction of transportation losses and better quality of power.

0.000

0.050

0.100

0.150

1

106 m

3

Water level of the upper reservoir (m3) Water level of the down reservoir (m3)

0.0

2.0

4.0

6.0

8.0

10.0

1

MW

Cumulative Wind absorbed rejected used for pumping wind rejected which is not used for pumping

-10.0-5.00.05.0

10.015.020.025.030.035.0

1

January

MW

rejected used for pumping Thermal stations Wind absorbed Turbine production Grid electricity for pumping Demand

0.0

10.0

20.0

30.0

40.0

MW

TM of conventional units conventional production above TM Rest conventional units ability

Fig. 5.19 Simulation of the WHPS in power system of Paros (January) – development of WHPS in Ios

5.3.3 Summary

In this paragraph the summary of results for the power system of Paro-Naxia are presented. The results are comparably presented for the case with or without WPS. A summary is provided in Table 5-6.



112

Table 5-6 Summary of results for IOS

Current situation (2005)

Current infrastructure (2010)

Adding RES without storage devices (2010)

Proposed operation (2010)

Demand (MWh) 179000 246260 246260 246260+6300(conventional pumping)=251540

Peak demand (MW) 56 74.8 74.8 74.8 Annual mean load (MW) 20.4 28.1 28.1 28.1

18.3 (outside the WPS) Wind installed capacity ~1.3MW 1.8MW 18.3MW

8 (inside the WPS) Conventional capacity(MW) 61.4 82 82 72.2

40890 (wind outside WPS) 4670 (wind in WPS) RES Production (MWh) ~4000 5540 40890 14600 (hydro turbine)

60160

16.3% (wind outside WPS) 1.9% (wind in WPS) RES supply (%) 2.2% 2.2% 16.6% 6% (turbine)

23.9%

24150 (wind outside WPS) RES curtailment (MWh) - - 24150 (36%)

23800 (wind in WPS) Thermal units (MWh) 175000 240720 (97.8%) 205370 (83.4%) 191380 (76.1%) Diesel (MWh or tn) 524tn 720tn 615tn 573tn HFO (MWh or tn) 38127tn 52445tn 44744tn 41696tn CO2 Avoided (tn) 3388 4692 34634 50955 SO2 Avoided (tn) 14.32 19.8 146.4 215.4 NOx Avoided (tn) 6.16 8.5 63 92.6 PM10 Avoided (tn) 1.36 1.88 13.9 20.5 Estimated O&M cost (€) 15,76 21,65 19,09 18,21

Estimated O&M Cost of Energy (COE)- 0,088 0,088 0,078 0,072



113

€ct/kWh Estimated change in Installation cost pay -back(€) 2,15 2,90

Total Cost of Energy (COE)- €ct/kWh 0,127 0,126 0,124 0,121 Scenario 100$/B Estimated O&M Cost of Energy (COE)- €ct/kWh 0,153 0,153 0,133 0,123

Total Cost of Energy (COE)- €ct/kWh 0,192 0,191 0,179 0,171 With the introduction of the WPS in the power system of Paro-Naxia, the wind penetration is increased from 16.6%, which can be achieved with wind

integration without storage, to 23.9%. Simultaneously the system’s EPC remains almost the same. As a consequence the operation of conventional units and their required installed capacity are reduced. The large power system of Paro-Naxia needs a rather progressive renewable energy integration to preserve the reliability of the system. Today the system is rather centralized, with all the local conventional power generated in Paros and transported to the other islands through several underwater cables. With the current infrastructure the proposed WPS, is considered as a good solution to make the first step towards a larger renewable penetration. In parallel, the Hellenic Transmission System Operator plans the interconnection of the Cyclades complex (Paro-Naxia is included) with the mainland, which will permit a larger wind power integration. In this case, clean wind power will be transferred to the mainland and local conventional units will be set to cool reserve. In such a perspective, WPS could be used to ensure the stability and the power quality of a rather weak power system characterized by large wind integration without local conventional production.



114

5.4 Case study 2 results-Cyprus

5.4.1 Wind potential The wind in Cyprus is affected by the following factors:

• From anticyclones moved from west to east, from the Siberian anticyclone during the winter and from the low pressure created in the area of India and expanded until the area of Cyprus during the summer.

• Sea breezes generated in coastal areas as a result of the different heat capacities of sea and land, which give rise to different rates of heating and cooling.

• Mountain-valley winds created when cool mountain air warms up in the morning and begins to rise – while cool air from the valley moves to replace it. During the night the flow reverses.

In Cyprus there is available wind potential, which can be exploited. In the following figure the mean wind velocity in a height of 10 meters above the ground are presented. There are some areas with mean wind velocity 5-6m/sec, while there are few areas with 7m/sec.

Wind farms in Cyprus will be mainly installed on top of mountain or hill chains to take advantage of the speed-up effect. In complex terrain the wind profile is difficult to be predicted and use of logarithmic or exponential law is not recommended [45].

In the above map, the available potential is presented. Additional restrictions should be taken into consideration, in order to discover the exploitable wind energy potential:

• Subregions dedicated to special activities must be excluded

• Subregions of less than 5m/s are of no interest at least for the current level of technology and the legislative framework.

• Subregions of very high altitudes or slope.

m/sec

4,0 4,5 5,0

5,5 6,0 6,5

7,0

© Dr. Ioannis P. Glekas

Fig. 5.20 Wind potential (annual mean wind velocity at 10m)



115

5.4.2 Scenario 1 – Adding wind power without Storage in Cyprus Using Fig. 5.1 for representative mean annual wind speed 6.3m/sxvii, for allowed instantaneous

wind penetration 40%xviii and desirable real capacity factor at least 20% (initial capacity factor for 6.3m/s is 22.3%), the maximum allowed wind installed capacity without storage is estimated to the 63% of the annual mean load. The electricity demand is 4333GWh or annual mean load of 494.6MW. Then the maximum allowed wind capacity to be installed without storage is estimated to 310MW. The energy supply of this wind installed capacity is 12.5% and the wind energy curtailment is 10.3%.

If the target is to ensure 0% curtailment, then the maximum allowed wind capacity to be installed without storage is restrained only to 127MW (25% of the annual mean load) and the wind energy supply to 5.5%.

5.4.3 Scenario 2 - Pump Hydro Installation 5.4.3.1 RESULTS OF THE PRE-FEASIBILITY STUDY

In the following figure the results of the pre-feasibility study of WPS in the power system of Cyprus are presented. Pre-feasibility study aims to examine the effect of the WPS integration into the autonomous power supply system. Then, in these charts the electricity production cost of the power system without pumped storage considering a Brent price of the order of 100$/b is presented in straight red line. This price is considered as a long term reference price.

The Market price for the conventional electricity used for pumping pm has been received 0.10€/kWh (around the existing consumer’s price) and the fixed price for the wind production pw was considered 0.125€/kWh.

The main conclusions of this approach in the power system of Cyprus are: • There is a general remark that the results are less attractive. The reason is the low wind

potential in Cyprus. It should be clarified that WPS is not a fuel, but only a power storage system. Then in areas without reach and cheap renewable source, the main problem is to identify the original source for exploitation. If this is not possible, then, we should expect higher costs and prices. But this is essential, towards the larger renewable contribution.

• Only a moderated RES integration (i.e. peak demand supply up to 16% or hybrid’s electricity supply up to 14% - Fig. 5.21a) can be achieved. Then the cumulative RES supply with wind (outside the WPS) and WPS may reach only up to 20% (Fig. 5.21 b) of the annual electrical demand. In this direction, the required reservoirs volume is up to 31 million m3 (Fig. 5.21.i) and the required wind installed capacity is up to 470MW (Fig. 5.21.j). It is obvious from the wind potential map (Fig. 5.20), that even the estimation of 6.3m/s average wind speed, sounds optimistic for such a massive wind integration. Cyprus has a very limited spatial and qualitative wind potential.

• The electrical system’s EPC seems to get reduced (Fig. 5.21.d) up to a 10% peak supply as a result of the WPS integration (from the level considered with Brent price 100$/b). In fact, this is a thoughtless conclusion as soon as the high consumption of conventional power for pumping (fig. 4.23.k) in a relatively low price strongly affects the results.

• The conventional units EPC is increased (Fig. 5.21 e) as soon as the same units operate less after the WPS integration and their low load factor is further decreased (Fig. 5.21.g).

xvii Mean annual wind speed of the order of 6.3m/s, is considered as a representative wind speed for the

cumulative wind installed capacity. Cyprus has a relatively limited wind potential with rather moderate wind speeds.

xviii Cyprus is a relatively large autonomous system and promotion of the geographical dispersion of wind farms, may allow an instantaneous wind penetration of the order of 40%.



116

• As regards, the turbine’s and hybrid’s EPC, the former varies around 0.25-0.32€/kWh and the latter around 0.29-0.42€/kWh (Fig. 5.21.c). This values represents a large cost, more expensive than the current peak supply cost, even a higher Brent price is considered. The only advantage is that the turbine’s electricity production cost is independent to the Brent price, providing less risk as regards the future costs and the fuel feed. Then, the profitability of this investment is strongly depends on the feed-in tariffs for the turbine’s electricity production. The cost could be reduced if the existing dams of Cyprus are used for energy purposes.

• The investment cost of the WPS is estimated to 1360-1520€/kW installed (cumulative wind and hydro turbine capacity) or 3800-7000€/kW of turbine capacity (Fig. 5.21.m, Fig. 5.21.n). Lower cost is estimated for larger WPS systems. It is relatively lower than in the previous examined case of Paro-Naxia due to the economy of scale.

• The efficiency of the WPS plant is estimated to 37-52% (Fig. 5.21.h).

• The RES supply in Cyprus can be increased thanks to the WPS integration. Only a share of 11% RES supply could be achieved if wind installed capacity is introduced without any storage devices. Further RES supply up to 9% could be achieved via the WPS integration, for the examined cases of this study (Fig. 5.21.b).

• For this achievement the wind installed capacity in the WPS should be significantly increased. In the case of Cyprus a relatively higher share of wind energy exploited (78-88%) is provided thanks to the uniform annual curve of power demand and the low wind potential (Fig. 5.21.l).

• On the other hand, the conventional power which is used for pumping is high, introducing the need for a rational design of the whole system (Fig. 5.21.k).



117

Turbine's EPC

Hybrid's EPC

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

0% 10% 20% 30% 40% 50% 60%


Turb

ine'

s EP

C (€

/kW

h) -

Hyb

rid's

EPC

(€/k

Wh)

without WPS

0,15

0,15

0,15

0,16

0,16

0,16

0,16

0,16

0,16

0% 10% 20% 30% 40% 50% 60%WPS share of the peak

Ele

ctri

cal s

yste

m's

EPC

(€/k

Wh)

32%

33%

33%

34%

34%

35%

35%

36%

36%

37%

0% 10% 20% 30% 40% 50% 60%


Load

Fac

tor

of c

onve

ntio

nal u

nits

Turbine's EPC

required price

system's EPC without WPS

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

0,40

0,45


requ

ired

pri

ce fo

r IRR

=16%

- Tu

rbin

e's

EPC

(€/k

Wh)

cyprus

without WPS

0,16

0,16

0,16

0,16

0,17

0,17

0,17

0,17

0% 10% 20% 30% 40% 50% 60%


Conv

entio

nal u

nits

' EP

C

(€/k

Wh)

0%

10%

20%

30%

40%

50%

60%


effic

ienc

y of

rev

erse

hyd

ro

sche

me

cyprus

0%

2%

4%

6%

8%

10%

12%

14%

16%

0% 10% 20% 30% 40% 50% 60%


Hybr

id's

ene

rgy

supp

ly (%

)

0%

5%

10%

15%

20%

25%


% r

es s

uppl

y

0

5.000.000

10.000.000

15.000.000

20.000.000

25.000.000

30.000.000

35.000.000


Res

ervo

ir's

vol

ume

(m^3

)

0,050,0

100,0150,0200,0250,0300,0350,0400,0450,0500,0


win

d in

stal

led

capa

city

in th

e W

PS

(MW

)

a) b)

c) d)

f) e)

i) j)

g) h)



118

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

0% 10% 20% 30% 40% 50% 60%


% c

onve

ntio

nal p

ower

for

pum

ping

76%

78%

80%

82%

84%

86%

88%

90%

0% 10% 20% 30% 40% 50% 60%


% w

ind

ener

gy e

xplo

ited

0

1000

2000

3000

4000

5000

6000

7000

8000

0% 10% 20% 30% 40% 50% 60%


cost

of t

he p

lant

per

MW

of

guar

ante

ed p

ower

(€/M

W)

1340

1360

1380

1400

1420

1440

1460

1480

1500

1520

0% 10% 20% 30% 40% 50% 60%


cost

of t

he p

lant

per

MW

of

inst

alle

d ca

paci

ty (€

/MW

)

Fig. 5.21 Results of the pre-feasibility study in Cyprus: a) hybrid’s energy supply, b) RES supply, c)

Turbine’s and hybrid’s, d) Electrical system’s EPC, e) Conventional units’ EPC, f) Required price for the turbines electricity production to provide IRR=16%, g) Load factor of conventional units, h) efficiency of the pumped storage system, i) Reservoir’s volume, j) wind installed capacity, k) % Conventional power for pumping, l) wind energy exploited (%), m) cost of the plant per MW of

installed capacity, n) cost of the plant per MW of guaranteed power.

5.4.3.2 EVALUATION OF EXISTING RESERVOIRS In this paragraph the existing reservoirs for irrigation are evaluated by the point of view of energy

exploitation in the framework of a WPS integration. There are several reservoirs in Cyprus most of them constructed in fifties and sixties for irrigation

and watering. Due to the low wind potential, and the discouraging results of the pre-feasibility study of the WPS in Cyprus, the exploitation of the existing reservoirs seems to be the only way.

As regards the possibility to use the existing reservoirs for energy purposes, the topography in Cyprus always provide the opportunity for the construction of an upper reservoir. The main question is to provide the highest possible hydraulic head in the relatively lowest distance.

The initial overview of the existing reservoirs, together with the pre-feasibility study in Cyprus, shows that that there are several potential cases.

In Fig. 5.22and Fig. 5.23 all the existing reservoirs are presented. It is obvious that both the wind installed capacity and the hydro storage systems should be

geographically distributed in the island for several reasons. The wind installed capacity should be distributed for three main reasons:

First of all such a big installed capacity needs a big area to be installed. There is not a specific site in the island which could seat such a huge wind installed capacity.

The second reason introduces the advantages of the geographical distribution of wind installed capacity in the decrease of the wind power output variability

Finally, from the electrical point of view it is preferable to widely distribute the wind farms. By this way the electrical losses due to the transmission of wind power to the demand are decreased, and additionally, if a part of the grid is lost for any reason, it is preferable to lose only a part of the wind production.

k) l)

n)m)



119

The Pumped Storage System should be also divided in a couple of pumped storage systems for various reasons:

First of all due to the geographical distribution of wind installed capacity, aiming at the reduction of losses, the pumped storage units should be distributed in the island. Ideally, a pumped storage, with a pumping ability of the same order should be installed closely to the wind installed capacity.

Secondly, the centralized production of the hydro turbines should be also avoided for electrical stability reasons and reduction of the Transmission and distribution losses.

Thirdly, topographical reasons and restrictions impose the splitting of the plants. In general, it may be easier to find more sites which can host smaller reservoirs, than one site for a huge reservoir.

Finally, due to the size of the investment, and the significant risk – since it is an innovative application with few real examples worldwide – it is difficult to be realized by one individual investor. It would be easier to announce 5 to 10 plants which can be undertaken by different investors.

The sitting of a suitable site for the construction of the reservoirs seems to be more difficult than the sitting of the wind farms. A suitable site should fulfill the following criteria:

Suitable topography where two reservoirs can be constructed, providing a high hydraulic head between the two reservoirs, with small distance between them and a neighbored stream should be available for the filling of the reservoir in the beginning of

the operation, and the complement of water losses due to evaporation. So, it is preferable to construct a couple (3-6) of smaller WPS distributed in the island.

In the framework of this study and after the initial identification of order of the reservoirs’ capacity, only the large dams are examined as probable cases for WPS installation. Theoretically, the small dams, could also provide a hydro storage solution, but scale reduction cost does not occur.

Source:

Fig. 5.22 Siting of the existing reservoirs in Cyprus



120

Fig. 5.23 List of existing “large” reservoirs and their features in Cyprus [46]



121

Fig. 5.24 List of six largest reservoirs in Cyprus

Then after an initial evaluation of the topography around the existing reservoirs, four of them are considered as initially suitable and subject for further investigation.

According to the initial estimations, there are possible sites which provide a relatively significant hydraulic head in a relatively low distance (Table 5-7). According to the pre-feasibility study and the evaluation of the existing reservoirs, the dimensioning of the proposed system in the feasibility study could start for hydraulic head 350m, length of the penstock 7km and reservoirs capacity of the order of 30-40*106 m3. This is a acceptable reservoirs capacity considering both the hydro and wind potential.

Table 5-7 Evaluation of existing dams with potential for a WPS exploitation in Cyprus

Volume of the existing (lower) reservoir (106 m3)

Estimated Hydraulic head H (m)

Distance from the probable site of the upper reservoir L (km)

Kouris 115 350 7.5 Asprokeramos 52 250 5 Evretou 24 400 6.8 Dhypotamos 15.5 220 5.5



122

Fig. 5.25 Photo of the Dam of Kouris in Cyprus (capacity: 115,000,000m3)

Fig. 5.26 Photomap of the existing Dam of Kouris in Cyprus (capacity: 115,000,000m3)

Kouris Dam

Possible site for the upper reservoir (H≈350m, L≈7.5km)



123

Fig. 5.27 Photomap of the existing Dam of Asprokeramos in Cyprus (capacity: 52,375,000m3)

Asprokremos Dam

Possible site for the upper reservoir (H≈250m, L≈5km)



124

Fig. 5.28 Photomap of the existing Dam of Evretou in Cyprus (capacity: 24,000,000m3)

Evretou Dam




125

Fig. 5.29 Photomap of the existing Dam of Dhypotamos in Cyprus (capacity: 15,500,000m3)

5.4.3.3 FEASIBILITY STUDY The system of Cyprus has a special interest. Due to the lack of local energy sources, Cyprus is

totally dependent on the fuel oil imports. Additionally, Cyprus is not a typical windy island and suffers from water scarcity. Then any propose of sustainable development of the electrical system which will be based on wind with hydro, is expected to be more expensive than in other case-studies islands.

The dimensioning in this case, starts from the existing dams and the pre-feasibility study. Taking into consideration that the existing reservoirs are also used for irrigation, then only a part of the reservoir’s capacity can be used for energy use. Additionally, the topography of the site of the upper reservoir imposes an additional constrain. Then, useful capacity of the reservoirs has been estimated initially around to 30-40*106m3, to start with the simulation and the dimensioning procedure. The hydraulic head was received 350m and the distance between the two reservoirs 5km.

The rest components of the WPS has been then estimated: • Hydro turbine: 250MW,

• Wind Farms’ capacity: 450MW

• Upper reservoir: ~30,000,000m3 - Lower reservoir (existing): ~76,000,000m3

• Pumping station: 360MW

• Number of plants: 2

Dhypotamos Dam




126

The general design of the WPS is compromised: • double penstock

• use of the turbine as a spinning reserve to increase the direct wind absorption

• hydro-turbine’s supply of the daily demand peaks (stable daily energy production, distribution according to the needs)

The main effect on the contribution to 2010 of this plant is estimated to 11.3% of the electricity demand. A small part is derived by direct wind absorption (5.9%) and an other 5.4% is derived by the hydro turbine supply. The “Guaranteed power” is 250MW or the 25.4% of the annual peak demand.

In Cyprus, there is no legislation about hybrid systems. Then the tariffs for the hydro turbine production, the pricing of guaranteed power and the price for the conventional power for pumping are objectives of negotiation between investor, system operator and regulatory authority (CERA).

Table 5-8 presents a summary for the feasinility study for Cyprus while Table 5-9 presents the results of the energy features.

Table 5-8 Results of the simulation for Cyprus – development of two WPS plants (total hydro turbine capacity 250MW and wind capacity 450MW)

Island Cyprus

Year of simulation 2010

Average wind speed at the hub height (m/s) 6,3

Scenario of operation 7

Hydro turbine's share of the peak 25,4%

hydraulic head (m) 350 Length of the penstock (m) 2000

Main parameters of dimensioning Rated power of the turbine (MW) 250,0 Rated power of wind farms (in the hybrid) - MW 450,0 Rated power of wind farms (out of the hybrid) - MW 310,0 Rated power of pumps - MW 360,0

Capacity of the reservoir – 1,000,000m3 30

Energy calculations Peak - MW 74,8 Minimum demand - MW 11,8 Annual energy demand - GWh 246,3 Mean annual load - MW 28,1 Wind energy production (out of the hybrid) - GWh 60,7 Wind energy absorbed (out of the hybrid) - GWh 43,1 Wind energy curtailed (out of the hybrid) - GWh 17,6 Wind energy production (in the hybrid) - GWh 26,5 Wind energy absorbed (in the hybrid) - GWh 6,9

Wind energy curtailed (in the hybrid) - GWh 19,7 Cumulative wind energy produced 87,3 Cumulative Wind energy absorbed - GWh 49,9 Cumulative wind energy curtailed - GWh 37,3



127

Table 5-9 Energy features of the reverse hydro system

energy used for pumping given the number and rated of pumps 779,7 Wind energy rejected (out of the hybrid) 13,2% energy used for pumping given the number and rated of pumps, and the capacity of the upper reservoir 779,7 Wind energy rejected (in the hybrid) 65,1% energy used for pumping given the number and rated of pumps, and the capacity of the upper and the lower reservoir 779,7 Part of the rejected wind energy (in the hydrid), used for pumping 89,1%

cumulative final energy used for pumping 779,7 Part of the rejected energy (out of the hybrid), used for pumping 0,0% wind energy surplus (in the hybrid), used for pumping 491,3 Part of the energy used for pumping derived by conventional units 37,0%

wind energy surplus (out of the hybrid) used for pumping 0,0 Ratio of conventional power for pumping and of turbine's production 63,2%

conventional energy used for pumping 288,3 Theoretical Capacity factor of wind farms (out of the hybrid) 21,5% Turbine's production - GWh 456,3 Real Capacity factor of wind farms (out of the hybrid) 18,6% Renewable energy poroduction - GWh 167,9 Theoretical Capacity factor of wind farms (in the hybrid) 21,5% Conventional energy production - GWh 4047,2 Real Capacity factor of wind farms (in the hybrid) 7,5% Degree of turbine's setpoint achievement 100,0% CF of wind farms in the hydrib (absorbed and used for pumping) 20,0%

Turbine's Capacity Factor 20,8% Efficiency of the reverse hydro scheme (turbine's production / energy used for pumping) 58,5%

Wind farm's capacity factor 20,0% Degree of conventional for pumping set-point achievement 32,35% wind energy used (%) 92,9% Water moved through the pumps to the upper reservoir - 1000m3 565019 Capacity factor of the upper reservoir 17,5% Water moved through the turbine to the lower reservoir - 1000m3 575469 Capacity factor of the lower reservoir 82,5% Energy consumed by the pumps - GWh 779,7

Maximum conventional production / peak demand 100,5% Energy produced by the hydro turbine - GWh 456,3 Wind energy contribution (out of the hybrid) 10,1% Dynamic energy of water moved through the pumps - GWh 538,9

Wind energy contribution (in the hybrid) 5,9% Dynamic energy of water moved through the turbine - GWh 548,9 Turbine's contribution 9,1% Pumping operation efficiency 69,1%

WHPS contribution 15,0% Turbine's efficiency 83,1% Cumulative RES contribution 19,3% Efficiency of the reverse hydro scheme 58,5%



128

0

20

40

1

106 m

3


0

100

200

300

400

500

1

MW


-400

-200

0

200

400

600

800

1000

1

January

MW


0200400600800

1000

MW


Fig. 5.30 Simulation of the WPS in Cyprus (January)



129

5.4.4 Summary In this paragraph the summary of results for Cyprus are presented. The results are comparably

presented for the case with and without WPS in Table 5-10 Table 5-10 Summary of results for Cyprus

Current infrastructure (2010)

Adding RES without storage devices (2010)

Proposed operation (2010)

Demand (MWh) 5016960 5016960 5305276

Peak demand (MW) 984 984 984 Annual mean load (MW) 573 573 573-606

310 (outside the WPS) Wind installed capacity 0,00 310 450 (inside the WPS) Conventional capacity 1171,00 1171 921

506310 (wind outside WPS)

295540 (wind in WPS) RES Production (MWh) 506310 456250 (hydro turbine)

10% wind outside WPS 5,9% wind in WPS

RES supply (%) 10% 9,1% turbine 25% 77280 (wind outside WPS) RES curtailment

(MWh) 77280 551600 (wind in WPS) Thermal units (MWh) 5016960 4510650 4047176 Diesel (tn) 277187 249214 223607 HFO (tn) 831561 747640 670819 CO2 Avoided (tn) 428845 821407 SO2 Avoided (tn) 1812,59 3471,83

NOx Avoided (tn) 779,72 1493,47 PM10 Avoided (tn) 172,15 329,73 Resutls for 54$/b Estimated O&M cost (106 €) 54$/b 513,54 472,12 441,11 Estimated O&M Cost of Energy (COE)- €/kWh 54$/b 0,102 0,094 0,083 Installation cost pay -back (106 €) 129,59 170,04 219,63 Estimated change in Installation cost pay -back(€) 40,46 90,05 Total Cost of Energy (COE)- €/kWh - 54$/b 0,128 0,128 0,125 Resutls for 100$/b Estimated O&M Cost of Energy (COE)- €/kWh 100$/b 0,176 0,160 0,139 Total Cost of Energy (COE)- €/kWh - 100$/b 0,201 0,194 0,180



130

With the introduction of the WPS in the power system of Cyprus, the wind penetration is increased from 10%, which can be achieved with wind integration without storage, to 25%. Simultaneously the system’s EPC remains almost the same. The power system of Cyprus is entirely depended on oil, while it is characterized by a rather moderate wind potential. Then the wind power production or the proposed solution of WPS integration is a rather expensive solution, but it has to compete with a current high electricity production cost. Fortunately, the existing reservoirs constructed during the last three decades to collect water for irrigation and the Cyprus topography, permits the parallel exploitation of the existing infrastructure for energy purposes. The possible lack of water can be faced by the integration of desalination units which could operate with the wind surplus. Although Cyprus is not the ideal place for wind or hydro plans, there are several reasons -such as the existing feed-in tariffs, the lack of other renewable sources (hydro or geothermal) and the lack of any prospects for interconnections- which show that the WPS integration is a suitable solution, until the costs of solar thermal for electricity and photovoltaics drop.



131

5.5 Case study 3 results-Corvo The island of Corvo represents a very small autonomous system, with currently extremely high

electricity production cost. The choice of a hybrid system comprised by wind and pumped storage, may not be the best solution for the increase of RES penetration. Additionally, these solutions are rather complex for a small island without the required technical know-how for its operation and maintenance. The battery storage may be a more consistent technology for such small islands

5.5.1 Scenario 1- Wind power in Corvo without storage In this small case study-island, the identification of the wind power capacity, which could be

permitted without any storage devices, has only a theoretical importance, since the wind capacity will be very small and without attractiveness for any investor.

Using Fig. 5.1 for representative mean annual wind speed 9.0m/s, for allowed instantaneous wind penetration 30%xix and desirable real capacity factor at least 30% (initial capacity factor for 9.0m/s is 45%), the maximum allowed wind installed capacity without storage is estimated to the 60% of the annual mean load. The electricity demand is 1084MWh or annual mean load of 124kW. Then the maximum allowed wind capacity to be installed without storage is estimated to 74.4kW. The energy supply of this wind installed capacity could be 18% and the wind energy curtailment is 33.3%.

As a result, this is a very small wind capacity, and possibly it could be achieved with one wind turbine of the order of 70-80kW or with several small wind turbines of the order of 5-10kW integrated in buildings or rural areas in the island. In such cases the power supply system-market is always consisted of only one player. Then, if there is a political will for a significant RES penetration, the overall system (WHPS, local power station, grid and system operation) will be owned by one player. As a result no wind installed capacity outside of the hybrid is expected.

5.5.2 Scenario 2 - Pump Hydro Installation

5.5.2.1 PRE-FEASIBILITY STUDY The peak demand in this island is only 0.205MW. Then, the pre-feasibility study is realized only

for comparative reason. Pre-feasibility studies always provide an initial investigation on the dimensioning and on the different degrees of penetration that could be achieved.

Pre-feasibility study aims to examine the effect of the WPS integration into the autonomous power supply system. Then, in these charts the electricity production cost of the power system without pumped storage considering a Brent price of the order of 100$/b is presented in straight red line. This price is considered as a long term reference price.

Since the target for such a small case study island is a large WPS integration, then the philosophy of the operation which is considered has several differences from the previous case studies:

• First of all, conventional power for pumping is not considered here. The high wind potential and the order of capacity of the reservoir, permits a different design.

• Additionally, as soon as the WPS will play a primary role in the whole system, the stable daily turbine supply is not possible. The target here for the WPS is to provide the required power according to the needs of the demand. Then, the hydro-turbine’s energy output is modulated taking into consideration the daily energy demand.

• No wind installed capacity outside of the WPS has been considered.

• A fixed price for the wind production pw was considered 0.10€/kWh.

xix Corvo is a very small autonomous system and even a 30% allowed instantaneous wind penetration may

be an optimistic assumption.



132

The main conclusions of this approach in the Corvo island are: • For the large scale RES integration i.e. WPS electricity supply up to 100% (Fig. 5.31.a and

b), the required reservoirs volume is up to 90,000 m3 (Fig. 5.31.i) and the required wind installed capacity is up to 0.7MW (Fig. 5.31.j).

• The electrical system’s EPC could be reduced (Fig. 5.31d) as a result of the WPS integration (from the level considered with Brent price 100$/b).

• The conventional units EPC is increased (Fig. 5.31.e) as soon as the same units operate less after the WPS integration and their low load factor is further decreased (Fig. 5.31.g). In case of high WPS penetration, extremely high conventional units EPS occur, as soon as very limited operation is required.

• As regards, the turbine’s and hybrid’s EPC, the former varies around 0.25-0.45€/kWh and the latter around 0.30-0.40€/kWh (Fig. 5.31.c). This cost is very competitive against the peak current cost, even a lower Brent price is considered. The turbine’s electricity production cost is independent to the Brent price. Then, the profitability of this investment is strongly depends on the tariffs for the turbine’s electricity production. Additionally, the WPS cost is decreased as soon as its penetration is increased, clearly justifying the different target-approach in this case.

• Although, in terms of EPC, this hybrid solution seems to be attractive, requires a high investment cost. Actually, the estimated investment cost of the WPS is 4,000-6,000€/kW installed (cumulative wind and hydro turbine capacity) or 17,000-20,000€/kW of turbine capacity (Fig. 5.31.m, Fig. 5.31.n). Lower cost is estimated for larger WPS systems.

• The efficiency of the WPS plant is estimated to 70% (Fig. 5.31.h).

• The RES supply in Corvo can be significantly increased thanks to the WPS integration.

• For this achievement the wind installed capacity in the WPS should be increased. More than 60% of the wind energy is exploited (Fig. 5.31.l).

• No conventional power is used for pumping as explained previously (Fig. 5.31.k).



133

Turbine's EPC

Hybrid's EPC0,00

0,20

0,40

0,60

0,80

1,00

1,20

1,40

1,60

1,80

0% 20% 40% 60% 80% 100%


Turb

ine'

s EP

C (€

/kW

h) -

Hyb

rid's

EPC

(€/k

Wh) without WPS

0,000,100,200,300,400,500,600,700,800,901,00

0% 20% 40% 60% 80% 100%WPS share of the peak

Ele

ctri

cal s

yste

m's

EPC

(€/k

Wh)

0%

10%

20%

30%

40%

50%

60%

0% 20% 40% 60% 80% 100%


Load

Fac

tor

of c

onve

ntio

nal u

nits

Turbine's EPCrequired price

system's EPC without WPS

0,00

0,50

1,00

1,50

2,00

2,50

3,00


requ

ired

pri

ce fo

r IRR

=16%

- Tu

rbin

e's

EPC

(€/k

Wh)

Corvo

without WPS

0,00,20,40,60,81,01,21,41,61,82,0

0% 20% 40% 60% 80% 100%


Conv

entio

nal u

nits

' EP

C

(€/k

Wh)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%


effic

ienc

y of

rev

erse

hyd

ro

sche

me

Corvo

0%10%20%30%40%50%60%70%80%90%

100%

0% 20% 40% 60% 80% 100%


Hybr

id's

ene

rgy

supp

ly (%

)

0%10%20%

30%40%50%60%70%

80%90%

100%


% r

es s

uppl

y

010.00020.00030.00040.00050.00060.00070.00080.00090.000

100.000


Res

ervo

ir's

vol

ume

(m^3

)

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8


win

d in

stal

led

capa

city

in th

e W

PS

(MW

)

a) b)

c) d)

e) f)

g) h)

i) j)



134

0%10%20%30%40%50%60%70%80%90%

100%

0% 20% 40% 60% 80% 100%


% c

onve

ntio

nal p

ower

for

pum

ping

0%10%20%30%40%50%60%70%80%90%

100%

0% 20% 40% 60% 80% 100%


% w

ind

ener

gy e

xplo

ited

0

5000

10000

15000

20000

25000

0% 20% 40% 60% 80% 100%


cost

of t

he p

lant

per

MW

of

guar

ante

ed p

ower

(€/M

W)

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

0% 20% 40% 60% 80% 100%


cost

of t

he p

lant

per

MW

of

inst

alle

d ca

paci

ty (€

/MW

)

k) l)

m) n)

Fig. 5.31 Results of the pre-feasibility study in Corvo: a) hybrid’s energy supply, b) RES supply, c)

Turbine’s and hybrid’s, d) Electrical system’s EPC, e) Conventional units’ EPC, f) Required price for the turbines electricity production to provide IRR=16%, g) Load factor of conventional units, h) efficiency of the pumped storage system, i) Reservoir’s volume, j) wind installed capacity, k) % Conventional power for pumping, l) wind energy exploited (%), m) cost of the plant per MW of

installed capacity, n) cost of the plant per MW of guaranteed power.

5.5.2.2 EVALUATION OF MORPHOLOGY Corvo’s main feature is the extinct Monte Gordo volcano, whose crater (the Caldeirão) is the

predominant feature. This crater, with a 3.7 km perimeter and 300 metre depth, is populated by various lakes, volcanic cones and crowned by turbary fields (Fig. 5.32). The highest point on the island is the “Morro dos Homens” on the southern flank of the Caldeirão, at 718 m above sea level. One road leads to the top of the volcano.

Fig. 5.32 The crater Monte Gordo volcano on Corvo island

Some small streams in the south-east part of the island Fig. 3.25 may provide the root towards the definition of a suitable site for the installation of a WPS. In this case a hydraulic head of the order of 250-300m could be easily provided in a distance of 600-800m between the two reservoirs (Fig. 5.33).



135

Fig. 5.33 Typical morphology of Corvo island.

The other case is to follow the example of Okinawa-Japan (Fig. 5.34) and construct a Seawater Pumped Storage Power Plant Operator. The plant of Okinawa is much bigger with a rated output of 30 MW, and an upper reservoir’s capacity 590,000 m3 despite the low effective hydraulic head of 136 m. Thanks to the morphology of Corvo island, there are several places suitable for this plant, which provide a significant hydraulic head. There are areas in the west side and the north east side of the island very close to the sea (distance of the order of 400-500m) with an altitude of 300-350m. The north part of the island provides also suitable sites with even higher hydraulic head, but these areas are considered as a natural beauty due to the volcano (Fig. 5.32).

Fig. 5.34 The example of sea water pumped storage in Okinawa.



136

5.5.2.3 FEASIBILITY STUDY The system of Corvo represents a totally different case study and has special interest, because it is

a very small island. It is obvious, that in such a small system with high potential and small energy needs, the achievement of a higher penetration directly could be the target. Corvo is the most isolated island in Europe, and then a solution which will provide energy independence is desirable.

According to the pre-feasibility study and the concept of the previous paragraph, a system which could provide the 100% of the energy needs is examined. To start with the dimensioning and the feasibility study of the Corvo’s WPS, a hydraulic head H=280m in a distance of 700m, is assumed

The rest components of the WPS has been then estimated: • Hydro turbine: 0.24MW,

• Wind Farms’ capacity: 0.9MW

• Upper reservoir: 90,000m3 - Lower reservoir (existing): 90,000m3

• Pumping station: 0.9MW (10 pumps)

The general design of the WPS is a little bit different than in the previous case studies. Again a double penstock is required to provide operational flexibility. Due to the small size of the power system of Corvo, and the small wind installed capacity, wind power is only used for pumping and the whole power demand is supplied by the hydro turbine.

The assumption that the wind power is only given for pumping is a safe-side assumption, which ensures the safe operation of this small isolated power system. In such cases, total loss of wind power may occur due to sudden fall of wind or due to a sudden storm over the one or two wind turbines of the island with wind speeds that exceed the operational limits of the turbines. Additionally, the variation of the wind power production may be significant and cause frequency or voltage fluctuation.

The hydro turbine supply is then estimated to 100% of the electricity demand in 2010. The “Guaranteed power” is 0.24MW or the 100% of the annual peak demand. Then, all the diesel power units of the system are set in cool reserve.

Table 5-11 gives a detailed summary of the results of the simulation. The former refers to the general electricity data while the latter refers to cost estimation issues.



137

Table 5-11 Results of the simulation for Corvo – development of one WPS plant (hydro turbine capacity 0.24MW and wind capacity 0,9MW)

General inforamtion Criteria of evaluation Island Corvo Degree of turbine's setpoint achievement 100.0% Year of simulation 2010 Turbine's set-point 60.3% Average wind speed at the hub height (m/s) 9.4 Wind farm's capacity factor 22.1% Scenario of operation 3 wind energy used (%) 49.2% Hydro turbine's share of the peak 100.0% Capacity factor of the upper reservoir 89.3% Main parameters of dimensioning Capacity factor of the lower reservoir 10.7% Rated power of the turbine (MW) 0.2 Maximum conventional production / peak demand 0.0% Rated power of wind farms (in the hybrid) - MW 0.9 Wind energy contribution (out of the hybrid) - Rated power of wind farms (out of the hybrid) - MW - Wind energy contribution (in the hybrid) 0.0% Rated power of pumps - MW 0.9 Turbine's contribution 100.0% Capacity of the reservoir - m3 90000 WHPS contribution 100.0% Energy calculations Cumulative RES contribution 100.0% Peak - MW 0.2 Minimum demand - MW 0.1 Wind energy rejected (in the hybrid) 100.0% Annual energy demand - GWh 1.3 Part of the rejected wind energy (in the hydrid). used for pumping 49.2% Mean annual load - MW 0.1 Wind energy production (out of the hybrid) - GWh - Part of the energy used for pumping derived by conventional units 0.0% Wind energy absorbed (out of the hybrid) - GWh - Ratio of conventional power for pumping and of turbine's production 0.0%

Wind energy curtailed (out of the hybrid) - GWh - Wind energy production (in the hybrid) - GWh 3.6 Wind energy absorbed (in the hybrid) - GWh 0.0 Theoretical Capacity factor of wind farms (in the hybrid) 45.0%

Wind energy curtailed (in the hybrid) - GWh 3.6 Real Capacity factor of wind farms (in the hybrid) 0.0% Cumulative wind energy produced 3.6 CF of wind farms in the hydrib (absorbed and used for pumping) 22.1%

Cumulative Wind energy absorbed - GWh 0.0 Efficiency of the reverse hydro scheme (turbine's production / energy

used for pumping) 72.0% Cumulative wind energy curtailed - GWh 3.6 Turbine's Capacity Factor 60.3% Energy features of the reverse hydro system energy used for pumping given the number and rated of pumps 3.1 Water moved through the pumps to the upper reservoir - 1000m3 1919 energy used for pumping given the number and rated of pumps. and

the capacity of the upper reservoir 1.7 Water moved through the turbine to the lower reservoir - 1000m3 1878



138

energy used for pumping given the number and rated of pumps. and the capacity of the upper and the lower reservoir 1.7 Energy consumed by the pumps - GWh 1.7

cumulative final energy used for pumping 1.7 Energy produced by the hydro turbine - GWh 1.3

wind energy surplus (in the hybrid). used for pumping 1.7 Dynamic energy of water moved through the pumps - GWh 1.5 wind energy surplus (out of the hybrid) used for pumping 0.0 Dynamic energy of water moved through the turbine - GWh 1.4 conventional energy used for pumping 0.0 Pumping operation efficiency 83.9% Turbine's production - GWh 1.3 Turbine's efficiency 87.6% Renewable energy poroduction - GWh 1.3 Efficiency of the reverse hydro scheme 72.0% Conventional energy production - GWh 0.0



139

0,000

0,050

0,100

1

106 m

3


0

0

0

1

1

1

1

MW


-1

-1

-1

0

0

0

0

0

1

JanuaryM

W


000111

MW


Fig. 5.35 Simulation of the WPS in Corvo (January)

In Fig. 5.35 the simulation of the WPS in Corvo for January is presented. The following remarks are drawn:

It is obvious that despite the high cost of installation of a WPS (15575 €/kW of guaranteed power or 9893 €/kW of cumulative installed capacity), the replacement of the current power supply system with the WPS is feasible. The electricity production cost of the system could be reduced to 0.3€/kWh from the current cost of 0.9€/kWh. Then, a definition of a tariff for the hydro production around 0.4€/kWh could provide the required incentive for the undertaken of this investment. It is a relatively small investment of the order of 3.7 million €, then a relatively higher IRR is essential.

Another important issue which should be taken into consideration is the operational complexity and the reliability of the whole system. In such isolated areas, the lack of know-how and of experts for the operation and maintenance of the energy systems, determines in a high degree the final decisions making.

Always the local power utility prefers technical solutions which are more reliable and less complicated. The integration of a hybrid system additionally to the current conventional system may increase the need for control and management of the whole system. On the other hand, the integration of a hybrid system, which will provide the total electricity supply maybe feasible and preferable in terms of operational complexity and energy independence

5.5.3 Summary

In this paragraph the summary of results for the case study of Corvo are presented. The results are comparably presented for the case with the current infrastructure and with the integration of WPS.



140

Table 5-12 Summary of results for Corvo

Current infrastructure (2010) Proposed operation (2010)

Demand (MWh) 1256 1256 Peak demand (MW) 0.24 0.24 Annual mean load (MW) 0.14 0.14

- (outside the WPS) Wind installed capacity - 0.9 (inside the WPS) Conventional capacity 0.56 0,24

- (wind outside WPS) 0 (wind in WPS)

RES Production (MWh) 0 1256 (hydro turbine)

0% wind outside WPS

0% wind in WPS RES supply (%) 0% 100% turbine 100%

- (wind outside WPS) RES curtailment (MWh) - 3550 (wind in WPS) Thermal units (MWh) 1256 0 Diesel (tn) 707 0 HFO (tn) 0 0 CO2 Avoided (tn) - 1064 SO2 Avoided (tn) - 4.50 NOx Avoided (tn) - 1.93 PM10 Avoided (tn) - 0.43 Resutls for 54$/b

Estimated O&M cost (M€) 0.74 0.08 Estimated O&M Cost of

Energy (COE)- €/kWh 0.587 0.059

Installation cost pay -back (M€) 0.09 0.36

Estimated change in Installation cost pay -back(M€) 0.27

Total Cost of Energy (COE)- €/kWh 0.656 0.348

Resutls for 100$/b Estimated O&M Cost of Energy (COE)- €/kWh 0.852 0.059 Total Cost of Energy (COE)- €/kWh 0.922 0.348

With the introduction of the WPS in the power system of Corvo, the wind penetration is increased

from 0% today up to 100%. Simultaneously the system’s EPC is significantly reduced (from the current 0.656 to 0.358€/kWh), providing in parallel fully independence from the oil price. Today,



141

the power system of Corvo is entirely depended on oil, while Corvo is an isolated power system without any prospect of interconnection. The oil is even more expensive due to the distance and transportation expenses. On the other hand, abundant wind potential remains the cheapest, local and unexploited source. The morphology of Corvo, permits the construction of the required infrastructure for the operation of a WPS which could supply the entire electricity demand. If the target is to make Corvo one of the first 100% RES islands in Europe, then the WPS seems to be an ideal solution. For more conservative targets, other solutions (for example PV with batteries) should be examined and may be proved more suitable or cheaper for such very small island-cases.

5.6 Conclusions of the case studies analysed In autonomous islands, the wind penetration is restricted due to technical reasons related with the

safe operation of the electrical systems. The combined use of wind power with pumped storage systems (WPS) is a mature solution to exploit the wind potential, increase the wind installed capacity and substitute conventional units operation and installed capacity. The simulation is based on the non-dynamic analysis of the electrical system, in order to calculate the energy contribution of the different power units. The aim is to analyse the prospects of Wind and Pumped Storage systems to increase the renewable energy penetration levels. Three islands with totally different features has been analysed as representative case studies:

• The island of Ios, which is part of autonomous system of ParoNaxia complex, Cyclades, Greece

• The island of Corvo, Portugal

• The island-country of Cyprus

Several differences are imposed in the design and operational strategy of the WPS due to the different features of each case study. The main features and the conclusions for the three case studies are comparably presented in the following table.

With the introduction of the WPS the wind penetration in autonomous systems can be increased, simultaneously in most cases decreasing the system’s EPC. As a consequence, the operation of conventional units and their required installed capacity can be significantly reduced.

The probable financial benefit from the introduction of the WPS should be shared between the ESO and the investor, by the definition of a suitable price. The main parameters which should be taken into consideration for the definition of the suitable price are the size of the plant, the size of the island, the current cost of the system and the duration curve of the demand.

Even in cases, where the system’s EPC is not decreased thanks to the integration of the WPS, an important advantage is that the production cost is to a large extent known in advance, contrary to the current cost which depends strongly on the oil price. Thus, the installation of WPS can provide both financial and environmental benefits and is strongly recommended for all the examined cases.



142

Ios Corvo Cyprus Short description of the case study

part of a medium size autonomous grid comprised by several islands

Very small and isolated island

Large country-island autonomous system

Prospects of interconnection

prospects of interconnection with the mainland

no prospects of interconnection

no prospects of interconnection in medium term

Why to introduce a WPS system in this system?

-Rather medium current electricity production cost -centralized system -existing reservoir -suitable topography for the construction of the upper reservoir

Very high current electricity production cost

- suitable topography for the construction of the WPS

- possibility of sea-water pumped storage

- Lack of other renewable sources (except of high solar potential, but expensive technologies)

- Existing reservoirs and dams constructed for irrigation

Drawbacks of WPS or main technical obstacles to be considered

- transport capacity of underwater cables

-operational complexity is increased with the integration of the WPS

- rather moderate wind potential

- lack of water

Operational design - wind power supplies the power demand with respect to the technical constraints of the autonomous system and the wind power surplus is used for pumping.

- wind power is used only for pumping

- wind power supplies the power demand with respect to the technical constraints of the autonomous system and the wind power surplus is used for pumping.

Target - a rather medium contribution

- even 100% electricity supply is possible

- a rather medium contribution

Other options - - use of batteries or hydrogen storage

- use of solar thermal or photovoltaics, when their costs are decreased.

Prospects of WPS to reduce the current electricity production cost

Neutral. The EPC will be slightly decreased thanks to the WPS integration

Positive. The EPC will be substantially decreased thanks to the WPS integration

Negative. The EPC will be increased due to the WPS integration.

Priority for the implementation

High High High



143

6 SIMULATION WITH HYDROGEN 6.1 General Description

In the analysis of Milos island case study we examined the introduction of hydrogen as an energy storage method in the overall power system. After some preliminary runs with HOMER software developed by NREL in the USA, it was decided to integrate a Proton Exchange Membrane (PEM) fuel cell with a nominal capacity of 1 MW, which accounts for a ca. 10% of the peak demand of the island. In the analysis of Corvo island case study it was decided to integrate a Proton Exchange Membrane (PEM) fuel cell with a nominal capacity of 100 KW, which accounts for a ca. 6% of the peak demand of the island.

Hydrogen is produced locally on the island through a water electrolysis unit, which uses excess energy produced by wind turbines. It is obvious that in order to produce adequate quantities of hydrogen for the operation of the fuel cell, additional wind turbines should be introduced in the power system of the island.

The proposed RES & hydrogen-based power systems of Milos and Corvo islands were optimized and simulated by using HOMER and then they were compared to the existing power systems of the islands in terms of cost of power generation, Renewable Energy (RE) penetration and emissions produced locally.

To perform the simulation of the existing power system of Milos and the optimization and simulation of the proposed hydrogen-based power system, actual load data derived from the Public Power Corporation (PPC) of Greece were used. Moreover, wind data measured on the islands were used as well. In the case of Corvo, the load data were obtained by the local electricity company EDA (Electricidade dos Açores) and the wind data used in the simulations are estimated data derived from data obtained from a meteorological station (10m) in the nearest island Flores.

Capital, O&M and replacement costs for thermal generator sets, wind turbines, water electrolysers, hydrogen storage tanks and PEM fuel cells used in the analysis were derived from previous studies conducted by members of the consortium of STORIES project [47] and from personal communications with equipment manufacturers. A detailed description of all the costs included in the analysis are presented in Deliverable 2.3.

6.2 Milos RES & Hydrogen results-Case Study 1 6.2.1 Simulation results of the existing power system of Milos island scenario 1

The results of the simulation of the existing power system of the island as described in subsection 3.2.1, revealed that the power generation cost is relatively high around 113 €/MWh and the wind energy penetration does not exceed 13%. The resulted cost of energy is derived using a price of 0.68 €/L for diesel fuel and 0.34 €/L for heavy oil fuel. Moreover, a subsidy of 30% for wind technology has been included. It is obvious that the electricity production on the island is based on imported fossil fuels (diesel and heavy oil), which results on the production of significant amounts of emissions (especially CO2).

The simulation results on the electricity production and demand of the island are summarized in Table 6-1.

The total fuel (heavy oil and diesel) consumption in Milos is 8,833,268 L/yr and the total CO2 emissions produced yearly on the island are 26,961,874 kg/yr. Moreover the annual operation, maintenance and fuel cost of the existing power system of the island is 4,056,099 €/yr. This amount also includes the cost of CO2 emission, which is calculated based on a 21 €/t value for CO2 emission rights and is equal to 566,200 €/yr.



144

Table 6-1 Presentation of the simulation results for the current system on Milos

Power Production Equipment Total Capacity Electricity production RES Share Thermal GenSets 11.25 MW 34,425,101 kWh/year 86.6 % Wind turbines 2.05 MW 5,316,007 kWh/year 13.4 % Yearly production 39,741,108 kWh/year 100 % Electric demand 39,728,620 kWh/year Excess electricity 12,544 kWh/year 0.032%

6.2.2 Scenario 2: Introduction of hydrogen (installed capacity of fuel cells accounts for a 10% of peak demand)

6.2.2.1 GENERAL DESCRIPTION In the analysis of Milos island case study we examined the introduction of hydrogen as an energy

storage method in the overall power system. After some preliminary runs with HOMER software developed by NREL in the USA, it was decided to integrate a Proton Exchange Membrane (PEM) fuel cell with a nominal capacity of 1 MW, which accounts for a ca. 10% of the peak demand of the island.

Hydrogen is produced locally on the island through a water electrolysis unit, which uses excess energy produced by wind turbines. It is obvious that in order to produce adequate quantities of hydrogen for the operation of the fuel cell, additional wind turbines should be introduced in the power system of the island.

The proposed RES & hydrogen-based power system of Milos island was optimized and simulated by using HOMER and then it was compared to the existing power system of the island in terms of cost of power generation, Renewable Energy (RE) penetration and emissions produced locally.

To perform the simulation of the existing power system and the optimization and simulation of the proposed hydrogen-based power system, actual load data derived from the Public Power Corporation (PPC) of Greece were used. Moreover, wind data measured on the island were used as well.

Capital, O&M and replacement costs for thermal generator sets, wind turbines, water electrolysers, hydrogen storage tanks and PEM fuel cells used in the analysis were derived from previous studies conducted by members of the consortium of STORIES project (Zoulias et al, 2007) and from personal communications with equipment manufacturers.

6.2.2.2 SIMULATION AND OPTIMIZATION RESULTS OF THE PROPOSED HYDROGEN-BASED POWER SYSTEM OF MILOS ISLAND

As it was discussed before, the main concept followed in the introduction of hydrogen as an energy storage medium in the power system of Milos, involves also the integration of additional wind turbines, which provide wind energy to serve the demand and when excess energy is available, it is used to produce hydrogen through water electrolysis as well.

The analysis conducted with HOMER software demonstrated that the optimum RES & hydrogen power system of Milos, schematically presented in Fig. 6.1, comprises the following equipment:

• 28 Vestas, model V-52 wind turbines with a nominal capacity of 850 kW each

• 2 Vestas, model V-44 wind turbines with a nominal capacity of 600 kW each (already in operation in the wind park of the island)



145

• 4 Thermal Generator Sets running on heavy oil with a nominal capacity of 4.9 MW in total (already in operation in the island)

• 1 Rental Thermal Generator Set (to be used only during summer months in order to cover peak demand) with a nominal capacity of 1032 kW.

• 1 PEM Fuel Cell with a nominal capacity of 1 MW

• 1 Alkaline Water Electrolyser with a nominal capacity of 2 MW, capable of producing ca. 400 Nm3/hr of hydrogen

• A hydrogen storage tank with a total capacity of 4,000 kg of hydrogen

Fig. 6.1 RES & Hydrogen Power System of Milos Island

The simulation results for the optimized RES & hydrogen power system showed that the power generation cost on the island decreases slightly to 112 €/MWh and there is a huge increase on RE penetration on the island of Milos, which can be as high as 86%. The simulation of the proposed system includes a subsidy of 30% for wind technology and 50% for hydrogen technologies. Moreover, the introduction of additional wind turbines and hydrogen as an energy storage medium results in a significant decrease on emission production on the island. The simulation results with respect to electricity production and demand showed are outlined in Table 6-2.

Table 6-2 Configuration of the proposed system

Power Production Equipment Total Capacity Electricity production RES Share

Thermal GenSets 4.9 MW + 1 MW (rental unit) 12,304,079 kWh/year 15 %

Wind turbines 25 MW 69,124,688 kWh/year 83 % PEM Fuel Cell 1 MW 2,353,161 kWh/year 3% Yearly production 83,781,928 kWh/year 100 % Electrolyser Load 7,352,470 kWh/year Electric demand 39,552,032 kWh/year



146

Excess electricity 36,877,792 kWh/year 44%

6.2.3 Comparison between the existing and the proposed hydrogen-based power system of Milos island

A comparison between the existing power system of Milos and the proposed RES & Hydrogen power system with respect to emission production, renewable energy penetration and power production cost is provided in Table 6-3.

Table 6-3 Presentation of the simulation results when Hydrogen & RES are installed on Milos

Emissions Existing Corvo Power System

Optimum RES & H2 Power System

Carbon dioxide 26,961,874 9,841,757 Carbon monoxide 57,416 21,809 Unburned hydrocarbons 6,360 2,416 Particulate matter 4,328 1,644 Sulfur dioxide 524,409 196,873 Nitrogen oxides 512,329 194,607 Parameters Cost of energy (€/MWh) 113 112 Renewable Energy Penetration 13.4% 86%

According to the results of the simulation of the proposed hydrogen-based power system of Milos

and the comparison to the existing power system of the island, the following conclusions can be made:

The introduction of hydrogen as energy storage method in Milos results in: • a decrease in the power generation cost of the island (ca. 1%)

• a huge increase on RE penetration on the island (from 13.4% to 86%)

• a significant reduction on the thermal generator sets operating on the island (from 11.25 MW to 4.9 MW)

• a considerable reduction in diesel and heavy oil fuel consumption (from 8,833,268 L/year to 3,209,769 L/year) (ca 64%)

• a substantial reduction in emissions produced (especially CO2) (ca 63%)

• providing the possibility to install additional wind parks on the island

It should be noted that a further reduction (as expected) on the cost of hydrogen energy equipment and the introduction of external costs in the analysis will make the hydrogen-based system economically competitive to the existing one. Moreover, since the proposed configuration of the optimized RES & hydrogen power system of Milos results in the production of significant amounts of excess energy (around 36,878 MWh/yr), the option of transforming excess energy to hydrogen in order to be used locally in the transport sector seems challenging and beneficial for the quality of the environment of the island. Another option is the use of this energy for heating purposes.



147

6.2.4 Summary of Milos Results

Milos is a Greek island located on the southwest part of the Cyclades. Currently, the annual electricity demand of Milos, which is around 39,729 MWh, is covered by 8 thermal generator sets with a total capacity of 11.25 MW and a 2.05 MW wind park. An analysis for Milos was carried out in order to examine the introduction of hydrogen as an energy storage means and the higher penetration of RE in the power system. The proposed RES & Hydrogen-based power system was optimized and simulated using HOMER. The proposed and existing power systems were compared in terms of cost of energy, renewable energy penetration and emission production. The results are presented in Table 6-4.

Table 6-4 Summary results for Milos

Existing power system Proposed RES & H2 power system

Demand (MWh) 39,729 39,729Peak demand (MW) 8.5 8.5Wind installed capacity (MW) 2.05 25Wind electricity production (kWh/year)

5,316,007 69,124,688

Conventional capacity (MW) 11.25 5.9Conventional electricity production (kWh/year)

34,425,101 12,304,079

Fuel cell capacity (MW) 0 1Fuel cell electricity production (kWh/yr)

0 2,353,161

Yearly production (kWh/yr) 39,741,108 83,781,928Electrolyser load (kWh/yr) 0 7,352,470Electric demand (kWh/yr) 39,728,620 39,552,032Excess electricity (kWh/yr) 12,544 36,877,792Renewable energy penetration 13.4% 86%Diesel fuel (L/yr) 704,548 154,906Heavy oil (L/yr) 8,128,720 3,054,863CO2 avoided (tn/yr) 0 17,120,117CO avoided (tn/yr) 0 35,607UHCs avoided (tn/yr) 0 3,944PM10 avoided (tn/yr) 0 2,684SO2 avoided (tn/yr) 0 327,536NOx avoided (tn/yr) 0 317,722Cost of energy (€/MWh) 113 112 The results of the simulations showed that the introduction of hydrogen as an energy storage

method and the increased penetration of wind energy deliver electricity at a lower cost. Apart from the cost of energy that was not drastically changed, the significant increase of renewable energy penetration of the proposed power system results in a substantial reduction in the diesel and heavy oil fuel consumption. The decrease of fossil fuel consumption has the dual effect of reducing the island’s dependency on imported fuels and the production of harmful emissions. Contrary to the



148

existing power system, the proposed system produces a considerable amount of excess electricity. However, this amount may be exploited by using it to produce either hydrogen fuel for transport applications or energy for heating purposes. These options make the proposed system even more beneficial for Milos island.



149

6.3 Corvo RES & Hydrogen results-Case Study 2 6.3.1 Scenario 1-Simulation of the existing power system in Corvo

The existing power system of Corvo includes four diesel generators with total capacity of 560kW. More specifically, the island’s annual demand of approximately 1084 MWh and peak of 182 kW is covered by two sets of 120kW and two of 160kW. Based on these figures, it is evident that the four generators are never operating simultaneously. Thus, the demand is met by two generators one of each group. The results of the simulation of the existing power system of the island revealed that the power generation cost is considerably high at 259 €/MWh. The resulted cost of energy is derived using a price of 0.816 €/L for diesel fuel. The expensive cost of energy in conjunction with the environmental impacts of a fossil fuel-based power system and the security of supply of an isolated small island make the need for RES exploitation more imperative. The simulation results on the electricity production and demand of the island are summarized in Table 6-5.

Table 6-5 Presentation of the simulation results of the existing power system of Corvo

Power Production Equipment

Total Capacity Electricity Production

Percentage

Diesel Generators 560kW 1,084,413 kWh/year 100% Yearly production 1,084,413 kWh/year 100% Electric demand 1,084,411 kWh/year Excess electricity 0 kWh/year 0% The total fuel (diesel) consumption in Corvo is 288,051 L/yr and the total CO2 emissions

produced yearly on the island are 758,532 kg/yr. Moreover, the annual operation, maintenance and fuel cost of the existing power system of the island is 269,964 €/yr. This amount also includes the cost of CO2 emission, which is calculated based on a 21 €/t value for CO2 emission rights and is equal to 15,929 €/yr

6.3.2 Scenario 2-Results from the proposed Hydrogen-based power system As it can be witnessed from the description of the power system of Corvo, the existing system is

based exclusively on fossil fuels with zero RE penetration. Thus, the introduction of hydrogen as an energy storage medium in the power system of Corvo includes also the integration of renewable electricity-generation technologies. As in the case of Milos, the selected technologies are wind turbines. The concept of the proposed system is to produce more green electricity to serve the demand by using wind turbines and diesel generators and when excess energy is available, to use it to produce hydrogen through water electrolysis

The analysis conducted with HOMER software demonstrated that the optimum RES & hydrogen power system of Corvo, schematically presented in Fig. 6.2, comprises the following equipment:

• 2 Fuhrländer, model FL100, wind turbines with a nominal capacity of 100 kW each

• 1 Diesel Generator Set running on diesel with a nominal capacity of 120 kW

• 1 Diesel Generator Set running on diesel with a nominal capacity of 160 kW

• 1 PEM Fuel Cell with a nominal capacity of 50 kW

• 1 Alkaline Water Electrolyser with a nominal capacity of 80 kW, capable of producing ca. 19 Nm3/hr of hydrogen

• A hydrogen storage tank with a total capacity of 200 kg of hydrogen



150

Fig. 6.2 RES &Hydrogen Power System of Corvo Island

The simulation results for the optimized RES & hydrogen power system showed that the power generation cost on the island decreases to 145€/MWh, but more than half the electricity is produced from wind energy, making a considerable increase of RE penetration from zero to 94%. Moreover, the introduction of RES and hydrogen as an energy storage medium in the power system results in a significant decrease in emission production on the island. The simulation results with respect to electricity production and demand are outlined in Table 6-6

Table 6-6 Presentation of the simulation results of the RES & Hydrogen power system of Corvo

Power Production Equipment

Total Capacity Electricity Production

Electricity share

Diesel Generators 280 kW 315,904 kWh/year 20% Wind Turbines 200 kW 1,137,987 kWh/year 73% PEM Fuel Cell 50 kW 104,003 kWh/year 7%

Yearly production 1,557,893 kWh/year 100% Electrolyser Load 320,005 kWh/year Electric demand 1,084,411 kWh/year Excess electricity 153,472 kWh/year 9.85%

6.3.3 Comparison between the existing and the proposed hydrogen-based power system of Corvo island

A comparison between the existing power system of Corvo and the proposed RES & Hydrogen power system with respect to emission production, renewable energy penetration and power production cost is provided in Table 6-7.



151

Table 6-7 Comparison of the existing and the proposed RES & Hydrogen power system of Corvo

Emissions Existing Corvo Power System

Optimum RES & H2 Power System

Carbon dioxide 758,532 234,323 Carbon monoxide 1,872 621 Unburned hydrocarbons 207 68.7 Particulate matter 141 46.8 Sulfur dioxide 1,523 471 Nitrogen oxides 16,707 5,537 Parameters Cost of energy (€/MWh) 259 145 Renewable Energy Penetration 0% 80%

According to the results of the simulation of the proposed hydrogen-based power system of

Corvo and the comparison to the existing power system of the island, the following conclusions can be made: The introduction of hydrogen as an energy storage method in Corvo results in:

• a considerable decrease in the power generation cost of the island (ca. 43%)

• a huge increase on RE penetration on the island (from 0% to 80%)

• a significant reduction in emissions produced (especially CO2) (ca 69%)

• a significant reduction in diesel fuel consumption (from 288,051 L/year to 89,009 L/year) (ca 69%)

As in the case of Milos, a further reduction (as expected) on the cost of hydrogen energy equipment and the introduction of external costs in the analysis will make the hydrogen-based system even more economically attractive.



152

6.3.4 Summary of the results for Corvo

The Portuguese island, Corvo, is one of the nine islands in the Azores archipelago. It is the smallest island of the archipelago with annual electricity demand equal to 1,084 €/MWh. The island is exclusively dependent on imported fuel in order to cover its electricity needs. Currently, it is has four diesel generators of 560 kW total capacity. The analysis of Corvo included the introduction of hydrogen as a storage means and wind energy as an additional electricity production source in order to replace conventional fuel. The simulations were carried out using HOMER. Table 6-8 presents the results of the simulations. As it can be witnessed the introduction of wind energy as an energy source and hydrogen as a storage method is a quite attractive proposal for the power system of Corvo island. The proposed system results in a remarkable reduction (43%) in the power generation cost. Moreover, with the proposed system the 80% of the electricity needs of the island would be covered by wind energy, which uses a free feedstock for the production of energy, decreasing the island’s heavy dependency on imported fuel and the production of harmful emissions and enhancing security of supply, taking also into account the long distance of this island from larger islnads and Continental Portugal.

Table 6-8 Summary results for Corvo

Existing power system Proposed RES & H2 power system

Demand (MWh) 1,084 1,084 Peak demand (kW) 182 182 Wind installed capacity (kW) 0 200 Wind electricity production (kWh/year)

0 1,137,987

Conventional capacity (kW) 560 280 Conventional electricity production (kWh/year)

1,084,413 315,904

Fuel cell capacity (kW) 0 50 Fuel cell electricity production (kWh/yr)

0 104,003

Yearly production (kWh/yr) 1,084,413 1,557,893 Electrolyser load (kWh/yr) 0 320,005 Electric demand (kWh/yr) 1,084,411 1,084,411 Excess electricity (kWh/yr) 0 153,472 Renewable energy penetration 0% 80% Diesel fuel (L/yr) 288,051 89,024 CO2 avoided (tn/yr) 0 524,209 CO avoided (tn/yr) 0 1,251 UHCs avoided (tn/yr) 0 138 PM10 avoided (tn/yr) 0 94 SO2 avoided (tn/yr) 0 1,052 NOx avoided (tn/yr) 0 11,170 Cost of energy (€/MWh) 259 145



153

6.4 Conclusions of the case studies analysed The introduction of hydrogen as an energy storage method in power systems was examined in the

case study of the Greek island Milos and the Portuguese island Corvo. In both cases, a proposed RES & Hydrogen-based power system was studied that used wind energy as an additional source of electricity production replacing part of conventional electricity and hydrogen as an energy storage method. A general conclusion that may be drawn from both case studies is that hydrogen energy storage may complement renewable energy sources as it has the potential to tackle their intermittent nature and thus to assist in achieving high level RE penetration. The combination and introduction of wind energy and hydrogen storage into the power system of Milos and Corvo showed that the reduction of fossil fuel dependency, the enhancement of security of supply and the decrease of the production of harmful emissions associated by fossil fuel consumption are feasible and can be achieved at a lower than the current power generation cost.

More specifically, in the case of Milos a 73% increase in wind energy penetration is achieved at a cost of energy slightly lower than the current energy cost and equal to 112 €/kWh. The significant RE penetration of the proposed system results in a significant reduction in diesel and heavy oil consumption (64%) and in the produced emissions, especially in the case of CO2 that is the predominant greenhouse gas (63%). This high penetration scenario, though, results in the production of a considerable amount of excess energy (around 36,878 MWh/yr). However, the use of this energy either for the production of hydrogen as a transport fuel for vehicles and fishing boats or for heating purposes may enhance the economic attractiveness of the proposed system.

The benefits of introducing wind energy and hydrogen storage into the power system of Corvo are manifold. The proposed system provides 80% renewable electricity and uses 69% less diesel fuel. The significant decrease in diesel needs leads to a major reduction in produced emissions. Apart from the harmful emissions production, the reduced fuel dependency also greatly affects the cost of energy. The proposed system is considerably more economically attractive than the existing system since the cost of energy has been reduced by 43% and is around 145 €/kWh.



154

7 SIMULATION WITH DESALINATION As described in section 3.3 of this report, the behaviour of the power system of 3 islands, namely

Milos, Mljet and Cyprus has been studied. First, the methodology of scheduling a desalination plant when its schedule should be linked with the foreseen RES installation, is described. Then results from the application of this methodology to Milos and Mljet are presented.

For Cyprus the approach followed is somewhat different. The scope is to simulate the operation of a desalination system and its impact on the power system of Cyprus when there is effort to reduce wind power curtailment on the island if the foreseen installations of wind power were currently realized.

7.1 Methodology followed for Milos and Mljet For Milos and Mljet, the methodology followed is described here shortly. The desalination plant acts as a consumer for the power company with the exception that its

operation program is known to the power company. The time programming is made in 20-minute steps but the check of the water level in the reservoir is on hourly base. This schedule is updated frequently, e.g. once per hour, to converge to the water demand and if applied to the wind power production.

Two modes of operation have been studied : • The desalination plant operates independently from the wind turbine (business as usual)

• The desalination plant takes into account the energy production of the wind-turbine and try to minimize on hourly basis the impact on the power system, by minimizing the exchange of power with it.

In both cases, the water level in the reservoir will affect the production plan of the desalination plant. Two events may take place

• the water level is above a maximum limit

• the water level is below a minimum limit

In the first case, desalination modules will shut down in order to prevent an overflow. The modules which will stop working will be those which are working the longer during the simulation period.

In the other case, the desalination company will buy energy from the power system. The buying time will be such, that the desalination company will have to buy the smallest amount of energy.

The water level, leveli will be checked every hour, so: min_ max_ilevel level level< < (1)

Where min_level is the low limit, max_level is the high limit and level is the water level every hour.

The maximum level is laid for safety reasons, in case of overflowing, here the 95% of the tank’s capacity. The minimum level is laid for water adequacy. In the Milos case, the lower level is calibrated in such way, so if there is a problem with the power lines, the water can last for about eight hours, enough time for the power company to fix the problem or inform the citizens for water shortage.



155

Also there is another constraint that is taken into account: a compulsory pause between two starts of the modules which it’s needed for maintenance and for safety reasons. The time pause is dependent by the size of the module’s machines (the bigger the machines, the longer the pauses).

The following paragraph describes the algorithm developed to simulate the operation of the desalination plant when this co-operates in determining the operation schedule with the wind turbine taking into account the following variables :

• The water level, before the start of the algorithm

• The wind production from the wind-turbine

• The water consumption

7.1.1 Algorithm explanation Taking into consideration the wind production, the quantity of the water that will be able to be

desalinated can be calculated. The wind power which is not used by the plant will be sold to the power company in a premium price, and in case of power needed the desalination plant will purchase energy as a common consumer.

The flow chart of the algorithm is shown next: For every time period (in our case, every 20 minutes), the algorithm will calculate the modules

which can work only with the wind production. After that, the calculated number will be compared with the number of modules already working. There are three different cases:

• The working modules are less than those which operate on time t. In this case, more modules will enter the production.

• The working modules are more than those which operate on time t. When this happens, some modules stop working.

• If the working modules are equal to those which can operate, the modules’ status remains the same.



156

Calculation of the modules that can work only with the wind production on time t

Is it more than on time t-1

Is it less than on time t-1

Is it the same as on time t-1

Increase time

Fig. 7.1 Flow chart of the proposed algorithm

7.1.2 Water level calculation

The next step is the calculation of the water level on the tank which is shown in the following figure. The required variable that must be known are the water level at the start of the simulation, the water production ability of the modules and finally the water consumption by the island. The loop runs until the end of the simulation horizon (the simulation time will be an input)

(Water level)i= (water level)i-1+(water_production)i-

water_consumption

i=i+1

Fig. 7.2 Water level Calculation

7.1.3 Level check subroutines Then it is checked whether the water within the tank is within the upper and lower limits level

specified for each step of the simulation horizon, according to the following flow chart. If there is no problem, the production plan is the same with the one that was produced with only input the wind production. If there is any violation of the tank limits, overflow or lack of water, two separate algorithms have been created described next.



157

i=1

Water_level> maximum

Water_level<minimum

i=i+1

Maximum algorithm

Minimum algorithm

YES

NO

YES

NO

Fig. 7.3 Facing violation of the water tank level (minimum and maximum)

7.1.3.1 VIOLATING UPPER LEVEL (MAXIMUM ALGORITHM) If the water level is greater than the upper safety limit in the end of an hour, the algorithm firstly

recognizes the exact hour when this occurs, tmax. Then, the program identifies the hour during which the water plant works the most in the simulation time before tmax. During that hour, t1 the required power will be the highest, so in that hour 1 desalination module will close, if technically possible. The level of water in the tank is updated and the level is checked again and if there is still danger of overflow, the algorithm is repeated. The flow chart of this is shown in Fig. 7.4.



158

Which hour is there the problem?

Calculate the new water level

RETURN

Stop one module for an hour

Which hour the plant consumes the maximum energy?

Is the problem solved?

YES

NO

Fig. 7.4 Facing the danger of potential overflow of the water tank.

7.1.3.2 VIOLATING LOWER LEVEL (MINIMUM ALGORTIHM) Again the algorithm identifies the hour when the problem is, tmin. The solution is the start of one

or more modules for the required period so that the problem is solved as the following flow chart describes. The start up of an additional unit is made during the hour t2, before tmin, when the excess wind turbine production is higher. Thus, the desalination company will buy the minimum amount of energy to face water shortage. So the energy cost will be smaller and the power company will not affected so much. In the end of the algorithm, it is checked whether the water shortage is faced



159

or not. In case water shortage still exists the algorithm is repeated until the problem is resolved. If all the desalination modules have been put into operation until tmin, an alarm which will be triggered.

Which hour is the problem?

Start one module for an hour

Check the new water level

Find out in which hour the unused energy is the greatest

Are all the modules working till that hour?

Is the problem fixed?

RETURN

ALARM

YES

NO

NO

YES

Fig. 7.5 Facing the low content constraint.



160

7.2 Case study 1 results-Milos The following scenarios have been evaluated for the island system of Milos regarding

• Addition of one wind turbine -SCEN 1

• Addition of one wind turbine +desalination with independent scheduling-SCEN 2

• Addition of one wind turbine +desalination with co-operation in scheduling SCE 3

• Addition of desalination plant only.-SCEN 4

7.2.1 Adding Wind power production without desalination-Scenario 1 In this scenario an additional wind turbine of 850kW is considered on the island at the existing

wind park. The peak demand of each month for the whole island remains intact, but the production of the thermal station, energy and peak changes as Table 7-1 describes. Wind power penetration reaches 16.35% and the production of the thermal power plant is reduced by 4.62%. The peak of the plant is reduced by about 70kW. The impact on the operating cost, fuel consumption and potential wind power curtailment is provided in Table 7-2.

Table 7-1 Summary of the operation of the thermal power plant for scenario 1

Month Thermal Energy produced (MWh)

Thermal station Peak Demand (kW)

Month Thermal Energy produced (MWh)


January 2821.4 7271 July 3459.7 7580 February 2555.9 7070 August 4405.2 9737 March 2679.7 6643 September 2963.3 7262 April 2277.6 5520 October 2086.2 5350 May 2375.3 5750 November 2535.7 5930 June 2761.4 6838 December 2317.7 6313 Total 33239.1 9737

Table 7-2 Summary of the cost-fuel consumption and wind power curtailment

Month Cost(€) Mazout Consumption(kg)

Diesel Consumption(kg)

Wind power curtailment (kWh)

January 212789.7 668307.1 67695.73 179906.5 February 202764.1 597859.8 61550.05 86054.09 March 207363.8 665019.1 32381.94 114442.7 April 209590.9 429581.8 140350.1 97697.37 May 206493.7 543277.2 84583.71 115918.1 June 223029.5 615358.3 89103.33 112739.4 July 307317.4 707638.4 173126.7 27947.58 August 413799.2 672957.5 391056.9 15372.75 September 212960.6 677724.4 74456.55 46584.75



161

October 136660.1 541729.8 7913.954 169537.3 November 172202.7 597066.1 58127.3 94196.14 December 167269.1 582431.4 58666.61 166180.1 Total(X1000) 2672.24 7298.95 1239.013 1226.58

The operation of the wind turbine affects mainly the units that consume diesel oil since they are

the most expensive. Mazout consumption is slightly decreased. Thus, the reduction in the operating cost is higher than the one of energy and reaches 106,637€ or 3.84% of the fuel cost of the island. However, the savings could have been increased if there could have been an energy medium to store the wind power curtailed which reaches 15.88% of the potential production. This is the scope of the following scenarios studied.

The percentage of wind turbine production compared to the total wind power production, around 31%, is taken into account when calculating the wind power curtailment and the final income loss for the owner. The curtailment for the additional wind turbine is 375.7MWh worth of 31785€. For the existing park, the amount of energy sold reduction is 373.2MWh with value of 31,572€. If there had not been the wind power curtailment, the additional income for the owner of the wind turbines would be 63,357€ higher.

7.2.2 Adding wind power production and desalination plant-Scenario 2

On the island of Milos, water is scarce and part of the local needs are met by water transported by ships from mainland Greece at high cost 8€/m3. For this reason, the installation of a desalination plant with four independent and identical desalination modules is considered to provide part of the water needs of the local population and avoid transferring water to the island from the mainland.

With the word module we mean everything that is needed in order to produce potable water from sea-water. Usually a module consists of :

• a pump which pumps water (sea water in this case),

• the desalination mechanism (reverse osmosis here) and

• a second pump which stores the potable water in reservoirs.

There is secondary equipment like mechanisms that filter the water from the salt and mechanisms that throw the saline back to the sea.

The characteristics of each identical module are shown in Table 7-3. Table 7-3 Characteristics of the desalination plant.

Modules 4 Hourly water production/module 21m3 Power needed/module 150kW Tank capacity 3000 m3

Upper capacity level,(minleveli) 2800 m3

Lowest capacity level (maxleveli) 500 m3 The total demand is 406,000m3 which is expected to increase the demand of the island by about

2,900 MWh or 6.8%. In order to alleviate the impact of the increase in the demand, studied and thoroughly discussed in scenario 4, sub-section 7.2.4, the wind turbine of scenario 1 is considered to be installed. The difference in demand of the island before and after the installation of the wind turbine is shown in Fig. 7.6.

The operating schedule of the desalination plant, however, does not take into account the wind power estimations and the two operations are completely independent. Therefore, the algorithm



162

described in section 7.1 is applied only for the part described in 7.1.3 to avoid violating the level of water in the water tank. The Table 7-4 summarizes the demand data for the desalination plant.

Demand from the power plant is increased however as Table 7-5 presents. The peak of the island during August increases to 10330kW while the peak demand of the thermal power station is increased

to 10187kW. This increases mainly diesel consumption and thus the operating cost, as

Table 7-6. shows. However, significant reduction on the wind power curtailment is achieved and more wind power, by 322MWh is injected to the grid, although wind power penetration is somewhat decreased to 16%.

Table 7-4 Summary of the demand of the desalination plant-Milos

Month Desalination demand (MWh)

Peak Demand (kW)

Month Desalination demand (MWh)

Peak Demand (kW)

January 184.7 300 July 325.05 600 February 173.45 300 August 347.3 600 March 191.9 300 September 292.55 450 April 142.95 300 October 307.55 450 May 262.5 450 November 170 450 June 337.5 600 December 173.55 300 Total 2909 600

Percentage change of total load with and without the installation of the wind-turbine

-6-4-202468

101214

0 1 2 3 4 5 6 7 8 9 10 11 12

Months

(%)

Without wind-turbine With wind-turbine

Fig. 7.6 Change in demand of the island when desalination is installed with and without wind turbine installation

Table 7-5 Summary of the operation of the thermal power plant for scenario 2-Milos

Month Thermal Station Energy produced (MWh)


Month Thermal station Energy produced (MWh)




163

January 2973.3 7571 July 3770 8030 February 2707.2 7220 August 4744.4 10187 March 2848.6 6943 September 3241.6 7712 April 2401.4 5670 October 2342.6 5800 May 2602.5 6200 November 2684.4 6170 June 3047.2 7288 December 2462 6613

Total 35825.2 10187

Table 7-6 Summary of the cost-fuel consumption and wind power curtailment when a wind power turbine and a desalination plant are installed in Milos –Scenario 2

Month Cost(€) Mazout Consumption(kg)

Diesel Consumption(kg)


January 224628.8 688509.7 79454.09 147080.1 February 217320.9 615848 78531.17 63916.53 March 220637.6 694926.6 42471.75 91478.95 April 224075.6 435381.1 162682 78573.04 May 230761.5 557283.7 120210.2 80678.86 June 251890.7 644067.9 124365.4 61041.79 July 344661.2 714504.9 233681.9 13195.45 August 453754.5 683075.6 454831.1 7242.88 September 240905.3 705825.9 109947.3 32334.54 October 154794.5 595086.1 17811.78 118348 November 183960.1 617917.9 72750.63 72916.32 December 178330.9 604309.8 71049.24 136922.8 Total(X1000) 2925.72 7556.74 1567.79 903.73

Fig. 7.7 represents the power balance with the grid under this scenario. In this case, around 64%

of the times the balance is negative while the rest 36% positive. Moreover, around 13% of the time the balance is within ±50kW.



164

Power balance

0

2

4

6

8

10

12

-650

-600

-550

-500

-450

-400

-350

-300

-250

-200

-150

-100

-50 0 50 10

015

020

025

0300

350

400

450

500

550

600

650

700

750

80085

0

Pow e r(kW )

(%)

Power balance

Fig. 7.7 Power balance to the grid (Positive values selling to the Grid)

7.2.2.1 IMPACT ON THE ECONOMICS FOR THE OWNER OF THE DESALINATION PLANT& THE REST OF STAKEHOLDERS.

Here, an analysis for the economics for the owner of the desalination plant, the owner of the wind power turbine and the power system operator is provided. It is assumed that the desalination plant and the wind turbine have different accounting system.

The owner of the desalination plant is charged according to B1B tariff of the PPC as described in Table 7-7. The wind power curtailment is reduced to 277MWh reducing the income for the owner of the 4th wind turbine by 23437€, 8347€ less than scenario 1.

Table 7-7 Details for B1B Tariff-Milos

Power charge 8.4351€/kW

First 400kWh/kW 0.04990€/kWh Εnergy Charge

Rest kWh 0.03310€/kWh

Table 7-8 summarizes the monthly economical balance for the owner of the desalination plant if he is also the owner of the additional wind turbine. The cost for meeting the load is 5396€. The value of wind power curtailment would lead to income for the owner of this combination.

The fuels cost is increased for the PPC, on average by 98.01€/MWh, higher than the average fuel cost for the Island, compared to the operation under scenario 1.

The owner of the existing wind park increases the energy sold by 224 MWh, increasing his income by 18967.32€. However, he still sells less 149MWh to the grid compared to the current situation. This means that he still losses 12605.4€. If the whole park belongs to him, the additional benefit compared to scenario 1 reaches 27241.2€.

Table 7-8 Economics for the owner of the desalination plant-Milos

Month Energy expenses (€) Wind turbine income (€) Final cost (€) January 10660.10 22485.49 -11,825.4 February 10287.73 12486.15 -2,198.42 March 10898.42 11253.03 -354.61 April 9278.18 8806.574 471.61 May 15508.55 15305.76 202.79



165

June 20264.31 19901.34 362.97 July 19852.22 10608.44 9243.78 August 20588.69 13944.57 6644.12 September 16503.20 7361.29 9141.91 October 16999.70 15509.19 1490.51 November 12278.80 12320.72 -41.92 December 10291.04 18032.83 -7741.79 Total 173,410.91 168,015.4 5,395.56

7.2.3 Adding wind power production and desalination plant co-operation-

Scenario 3 Although wind power curtailment is reduced due to desalination plant in scenario 2, it still

remains high. At the same time, wind power is not used simultaneously with the increase in the loading of the power system. Scenario 3 assumes that the installation of a desalination plant is foreseen and the operating schedule is assumed to take into account the wind power estimations under the simple 4-hour persistence model. This means that the wind power forecast tool that the owner uses updates its forecast every 4 hours according to the last value of the previous 4 hours. Then the algorithm described in section 7.1 is applied. The total wind power requested is somewhat decreased but the peak of the thermal station production and the island remains the same with Scenario 2. However, what changes significantly is the reduction of the wind power curtailment, as described in Table 7-10. This affects both the operating cost and the fuel consumption provided in the same table.

The histogram of the power balance is shown in Fig. 7.8. As the chart shows, the power balance is negative most of the times. In fact, it is negative at around 54,07%, while at 46,93% of the time is positive. Around 22,40% of the time, the power balance is within ±50kW. Therefore the exchange of the combination with the grid is narrower compared to scenario 2 with all the benefits for the power system operation.

Table 7-9 Summary of the operation of the thermal power plant for Scenario 3-Milos





January 2956.6 7421 July 3763.1 8180 February 2690.4 7370 August 4742.3 10187 March 2810.2 6823 September 3227.6 7562 April 2364.6 5520 October 2331.3 5500 May 2570.1 6200 November 2663.8 6037 June 3026.1 6988 December 2457.7 6349 Total 35603.8 10187 Table 7-10 Summary of the cost-fuel consumption and wind power curtailment when a wind power

turbine and a desalination plant are installed in Milos, Scenario 3

Month Cost(€) Mazout Consumption(Tn)

Diesel Consumption(tn)


January 216849 680450 72970 124180



166

February 212770 617260 72920 53210 March 216976 686560 40870 39320 April 219381 422510 161410 55430 May 226742 537680 125440 48400 June 250493 622220 134840 39320 July 345439 702540 243290 9610 August 457239 658220 478370 5700 September 232649 698960 111620 22890 October 150410 585040 16730 98810 November 181629 614090 72480 63560 December 172755 602770 66260 126520 Total(X1000) 2883.33 7428.3 1597.2 686.95

Power balance

0123456789

10

-600-55

0-50

0-45

0-40

0-35

0-30

0-25

0-20

0-15

0-10

0 -500 50 10

015

020

025

030

035

040

045

050

055

060

065

070

075

080

085

0

Power (kW)

(%)

Power balance

Fig. 7.8 Power balance to the grid under scenario 3 (Positive values selling to the Grid)

7.2.3.1 IMPACT ON THE ECONOMICS FOR THE OWNER OF THE DESALINATION PLANT. The wind power curtailment is reduced, as Table 7-10 describes. The curtailment for the

additional wind turbine is 210MWh providing increased income by 5600€ compared to scenario 2 and 13947.59€ compared to scenario 1. If there had not been wind power curtailment, the income would have been 17837€ higher. The owner of the desalination plant is charged at B1B. However, the peak charge will be increased, since more often 4 units will operate. This is explicitly shown in Table 7-11.

Table 7-11 Economics for the owner of the desalination plant-Milos-Scenario 3

Month Energy expenses (€) Wind turbine income (€)

Final cost (€)

January 14584.48 23092.81 -8508.33 February 13038.78 12758.63 280.15 March 15325.49 12595.48 2730.01 April 11510.64 9401.28 2109.36



167

May 17766.54 16135.74 1630.8 June 20282.52 20457.30 -174.78 July 19744.64 10699.99 9044.65 August 20570.49 13983.79 6586.7 September 18625.86 7604.10 11021.76 October 19256.27 16022.58 3233.69 November 12460.84 12558.80 -97.96 December 13540.7 18305.01 -4764.31 Total 196707.3 173,615.50 23091.74

7.2.3.2 COMPARISON WITH SCENARIOS 1 & 2 If the owner of the desalination is owner of the additional wind turbine as well, he will have to

pay 23091,74€ as a balance for the water produced, 17696€ higher than the case of scenario 2. This increase can be faced if one takes into account that three stakeholders on the island are expected to have additional income.

The operator of the power system decrease the fuel cost increase by 42,389€ compared to scenario 2. The additional fuel cost for meeting the additional thermal power station due to the desalination plant (Compared to scenario 1) is 89.27€/MWh. A comparison of the monthly additional cost for meeting the desalination load for scenarios 2 and 3 is shown in Fig. 7.9. Only for July and August the additional cost is higher due to the increased water demand for this period and the fact that the additional energy requirements are closer in time with relatively high wind power production to maintain the balance. This leads to higher increase in production of the diesel fueled units. In all the other cases, the additional cost is lower and for the months with higher wind power penetration, significantly lower.

The retail seller increases its income compared to scenario 2 by the difference 196707-173410=23297€.

15

30

45

60

75

90

105

120

135

1 2 3 4 5 6 7 8 9 10 11 12

Month

Add

tiona

l Cos

t for

PPC

(€/M

Wh)

Scenario 2 Scenario 3 Fig. 7.9 Additional fuel cost for meeting the Desalination load demand compared to scenario 1

The owner of the existing wind park, if different from the owner of the 4th wind turbine, will sell 150.8 MWh more than in scenario 2 increasing his income by 12757.68€. The final curtailment for the existing wind park will remain the same with the current situation on the island.



168

If the owner of the 4th wind turbine is not the owner of the desalination plant, then the additional income for him will be 5600 €. The final increase in income of the wind park owners is 18,357.8€,compared to scenario 2. Compared to scenario 1, the additional income for wind park investors will reach, 45,652.42€ due to increasing wind power sales.

If the owner of the desalination plant is the owner of all the wind turbines, then the additional cost for scenario 3 is 4938.5€ higher than the one of scenario 2.

7.2.4 Adding desalination plant co-operation without wind-Scenario 4

This scenario evaluates the potential impact of a desalination plant on the power system of Milos when the wind turbine is not installed. Thus, the impact of combining wind power and a desalination plant can be assessed. The operation of the desalination plant is considered the same with scenario 2. The production of the thermal power plant is significantly increased as Table 7-12 describes. The peak demand of the island is the same with scenarios 2 and 3, 10330kW, but the peak demand of the power plant production is slightly increased. Moreover, the fuel consumption, especially diesel and hence O&M cost, are significantly increased as Table 7-13 shows. However, part of the increase of the demand is met by the reduction of the wind power curtailment that now is 5.96%, the lowest of all the scenarios studied.

It should be noted that the reduction of wind power curtailment is 157.9MWh and the additional income for the owner of the existing wind power park on the island will be annually 13,358€. Therefore, if the owner of the existing park installs the desalination plant the energy cost for him would be 160,052€.

Table 7-12 Summary of the operation of the thermal power plant for scenario 4-Milos





January 3182.5 7590 July 3912.1 8050 February 2832.4 7220 August 4906.1 10258 March 2950.6 7018 September 3316.8 7818 April 2486.2 5670 October 2486.9 5800 May 2775.8 6200 November 2807.2 6170 June 3253.2 7336 December 2685.1 6684 Total 37594.9 10258

Table 7-13 Summary of the cost-fuel consumption and wind power curtailment when only a desalination plant is installed in Milos

Month Cost(€) Mazout Consumption(Tn)

Diesel Consumption(tn)


January 239448 722435.3 91400.789 44350.6046 February 220478.3 664571.8 61631.882 22301.0046 March 227491.1 714578.2 44980.528 32649.781 April 231932.4 445933.9 170980.095 35375.0376 May 244833 571397 138272.914 23773.7923 June 269520.5 676783 139899.677 13249.1137 July 356246.1 727139 247679.478 2902.5875



169

August 470240.3 694536.6 478937.14 1973.1937 September 246665.3 715631.7 115660.27 10726.818 October 166309.8 634297.4 22367.1209 42631.819 November 192163.8 641001.3 79968.538 28123.656 December 189702.8 640652.3 77906.338 55520.867 Total(X1000) 3055.03 7848.96 1669.6848 313.5783

The owner of the desalination plant will be charged at B1B tariff scheme. The energy and the

peak of the desalination plant are the same with scenario 2 as provided in Table 7-4. The corresponding charge is provided in Table 7-8 under economic results column

However, the additional cost for the operator of the system will be significant under this scenario. The cost would increase by 9.85% due to the fact that the most expensive units, diesel fueled units would be affected and increase their production. The fuel cost for this additional amount of energy required would have been 99.46€/MWh significantly higher than scenario 3 and a bit higher than scenario 2. For 7 months, this additional cost would exceed 100€/MWh, and only for February, the additional demand would have been met with lower than 50€/MWh. Scenario 2 which has the same scheduling of the desalination unit but one additional wind turbine is by 126,801€ less expensive. Therefore, this is the value of adding one wind turbine for PPC to meet the additional demand caused by the wind turbine. If the desalination co-operates when determining its schedule with the wind turbine, scenario 3, the cost for PPC will be 169,191€ less expensive.

The value of the wind turbine production is 84.7€/MWh if the scenario 3 is applied and only 71.4€/MWh if scenario 2 is applied.

A detailed summary and comparison of all scenarios is provided in the following section.



170

7.2.5 Summary

The following tables and graphs summarize the impact on the fuel consumption, cost, and change in wind power production, curtailment and thermal power plant production for all the scenarios studied. First of all the impact per month in thermal power production and peak compared to the current situation is shown in Fig. 7.10. The increase in power plant production is higher in the scenario 2 compared to scenario 3 mainly due to the fact that there is co-operation between the desalination power plant and the additional wind turbine which reduces wind power curtailment. The impact on the peak of the thermal power production is shown in Fig. 7.11.

-12

-9

-6

-3

0

3

6

9

12

15

18

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge(

%)

Scenario 1 scenario 2 scenario3 scenario4

Fig. 7.10 Monthly change in power plant production-Milos

-1.5

0

1.5

3

4.5

6

7.5

9

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge(

%)


Fig. 7.11 Monthly change in peak of the power plant-Milos

For scenario 1, there is not always reduction of the peak demand because wind power and load are not coincident. For some months, and fortunately for August, the impact on the peak demand is reduced. Scenario 4 presents always higher peak than scenario 2 due to lack of the additional wind turbine. Both operating scenarios 2 and 4 increase the peak of each month.



171

Operation under scenario 3 manages to maintain the same peak demand for April and reduces the peak during December. Generally lower peak is maintained for most of the months with the exception of February and July compared to scenarios 2 and 4.

As a result of the change in thermal production, fuel consumption and thus the operating cost will be affected. As already discussed in the previous subsection, impact on units that consume mazout is rather limited as Fig. 7.12 shows. For winter months mazout consumption is reduced for scenario 1. For summer months the increase is due to the re-distribution of energy to the units that consume mazout due to shutting down for some periods one of the units that consume diesel. Very important is the change in diesel consumption shown in Fig. 7.13. For scenario 1, the diesel consumption is reduced for all months because of its cost.

For the rest scenarios, increase in thermal units production is noted for those units that consume diesel because the units that consume mazout, already operate almost at their technical maximum, before the units that consume diesel increase their production. Notable is the change during October when some of the base units are usually under scheduled maintenance and the increased required thermal production is met by units that consume diesel. Especially for Scenario 4, the consumption increase is 120%.

-6

-4

-2

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge(

%)


Fig. 7.12 Monthly change in mazout consumption-Milos

-30-15

0153045

607590

105

120135

1 2 3 4 5 6 7 8 9 10 11 12

Month

Chan

ge(%

)




172

Fig. 7.13 Monthly change in diesel consumption-Milos

Change in fuel consumption leads to change in cost for fuel as Fig. 7.15 shows. For all months in scenario 4 the cost is decreased. There is strong correlation between diesel consumption change and operating cost as the results for scenario 4 show. Scenario 3 manages to reduce cost for December and January when wind power conditions are rather favorable.

-8

-4

0

4

8

12

16

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge(

%)


Fig. 7.14 Monthly change in fuels cost-Milos

A comparison of the Wind power curtailment percentages per month is shown in Fig. 7.15. In all cases scenario 1 presents the highest wind power curtailment. For all months scenario 3 for the desalination plant helps so that the wind power curtailment percentage is lower than all the scenarios with the additional wind turbine. During some months, wind power curtailment percentage is lower than the current situation and especially for March scenario 3 presents the lowest wind power curtailment among the scenarios studied. High wind power curtailment impacts on the wind power penetration percentage especially during spring months as shown in Fig. 7.16.

0

3

6

9

12

15

18

21

24

27

1 2 3 4 5 6 7 8 9 10 11 12Month

Win

d po

wer

cur

tialm

ent (

%)

As is Scenario 1 scenario 2 scenario 3 scenario 4 Fig. 7.15 Wind power curtailment per month for each scenario studied-Milos

A summary of various indices for the power system of Milos is shown in Table 7-14 for quick evaluation. The percentage change for the various indices is shown in the figures that follow. Fig. 7.17 focus on RES and thermal plant production while Fig. 7.18 focuses on fuel consumption and cost. Fig. 7.19 presents the change in the emission levels. It is notable that for scenario3, although the thermal production is increased by 2.2%, CO2 emissions are increased by 1.8%. Taking into account that the desalination plant increases by 6.8% the total demand of the island, wind power helps in mitigating the increased emissions for meeting this demand.



173

3

6

9

12

15

18

21

24

27

1 2 3 4 5 6 7 8 9 10 11 12Month

Win

d po

wer

pen

etra

tion

(%)

As is Scenario 1 scenario 2 scenario 3 scenario 4 Fig. 7.16 Wind power penetration per month for each scenario studied-Milos

32.9

78.4

-0.7

30.1 31.5

2.8 3.9

34.2

-0.1 2.2 3.9

-3.8

7.9 4.6

-4.6

-33.0

-40-30-20-10

01020304050607080

RES penetration(%)

RES curtailmentshare(%)

Thermal units production (MWh)

Peak DemandThermal Station

(kW)

Cha

nge

(%)

Scenario 1 Scenario 2 Scenario 3 Scenario 4

Fig. 7.17 Comparison of the various scenarios with current operation regarding RES and power plant consumption

-1.3

-13.79

2.2

9.1

5.3

0.5

11.1

3.86.0

9.9

-3.84

16.1

-15-12

-9-6-30369

121518

MazoutConsumption

Diesel Consumption Estimated fuel cost

Cha

nge

(%)


Fig. 7.18 Comparison of the various scenarios with current operation regarding fuel consumption and cost



174

Table 7-14 Comparative table of all the scenarios studied for Milos

Current situation Scenario 1 Scenario 2 Scenario 3 Scenario 4 Demand (MWh) 39737.574 39737.574 42646.574 42641.574 42646.57 Peak Demand (kW) 9880 9880 10330 10330 10330 RES Production (MWh) 4887.6 6498.5 6821.4 7038.1 5045.5 RES penetration (%) 12.3 16.35 16 16.51 11.83 RES curtailment (MWh) 477.7 1226.6 903.7 687 319.8 RES curtailment share(%) 8.9 15.88 11.7 8.89 5.96 Thermal units production (MWh) 34849.974 33239.074 35825.174 35603.474 37601.07 Peak Thermal Station (kW) 9808 9737 10187 10187 10258 Mazout Consumption Tn 7395.2 7299 7556.7 7428.3 7841.5 Diesel Consumption (klt) 1437.2 1239 1567.8 1597.2 1668.9 CO2 Avoided (tn) 27178.59 26386.155 28014.711 27675.714 29173.26 SO2 Avoided (tn) 423.56 417.91 432.89 425.58 449.24 NOx Avoided (tn) 143.31 137.35 148.7 147.45 155.26 Particulate Avoided (tn) 15.47 15.05 15.77 15.72 16.3 Estimated Fuel cost (k€) 2778.9 2672.2 2925.7 2883.3 3052.5 Estimated Fuel Cost of Energy (COE)- €ct/kWh 7.974 8.039 8.167 8.098 8.118



175

-2.9

-1.3

-2.7

3.12.2

3.8

1.91.8

0.5

2.9

1.6

7.3

5.4

-4.2

8.3

6.1

-5-4-3-2-10123456789

CO2 SO2 NOx PM-10

Cha

nge

(%)


Fig. 7.19 Comparison of the various scenarios with current operation regarding emissions



176

7.2.6 Conclusions

For the island of Milos the operation with and without the combination of an additional wind turbine with a desalination plant has been studied.

Installation of an additional wind turbine will reduce the fuel consumption and hence both the emissions and the operating cost of the island. However, this will increase significantly the wind power curtailment on the island reducing the value of the investment on wind power compared to investing on wind on another island, with lower potential but without such problems. The cost decrease is sufficient, under circumstances, to pay-back the investment of wind turbine by the owner of the thermal plant. Especially if the investment on the wind turbine is made by another investor than the one owning the existing wind park, the wind power sales of the current owner of the wind park are significantly reduced.

Adding a load that can be controlled and whose product can be relatively easier stored than electricity, will significantly reduce wind power curtailment. Such a load is the desalination load. Thus the value of wind power on the island can be increased for all the stakeholders on the island. Moreover, in the scenario where the desalination plant schedule takes into account the wind power production estimations, the annual average wind power penetration is increased a bit compared to adding only a wind turbine with significant increase from 10% to 15% during Spring months.

If it assumed that neither the additional wind turbine, nor the desalination plant belong to the owner of the existing park, his income loss compared to adding a wind turbine only is mitigated. There will be zero impact in his income if the wind power estimations are used for providing the schedule of the desalination plant.

Under all circumstances, it is much more profitable for the owner of the thermal power plant to have co-operation of the wind turbine with the desalination plant in terms of scheduling its output according to the proposed algorithm. The value of the wind turbine when a desalination plant is to be added reaches 169,191€ contrary to 106,000€ with just adding a wind turbine without increasing Milos demand. Moreover, the additional emissions for the additional demand of the island are significantly lower compared to just adding a wind turbine and a desalination plant on the island.

A potential new investor on Milos island increases the wind power sales income if he also adds a desalination plant. However, co-operation of wind power and the desalination plant has higher operational cost for him taking into account the current tariff scheme on the island. The other stakeholders, existing wind park owner and the power plant owner have benefits which are higher and can cover this additional cost by simply reducing their benefits. The additional cost for the co-operation between the desalination plant and the wind turbine is the lowest if this potential investor is the owner of the existing wind park.

Adding only a desalination plant, increases by 6.8% the total demand of the island, and addition of wind power helps in mitigating the increased emissions for meeting this demand and reduce the cost of fuel on the island. The owner of the existing park, even if he is not the owner of the desalination plant will increase his sales because the demand will increased during these night hours but the wind power penetration will be generally reduced.

The Municipality of Milos with the addition of a desalination plants avoids transporting water to the island the water cost is well below 2€/m3 a quarter of the transport price of 8€/m3. Allowing installation of wind turbine instead of having the desalination plant running would decrease the additional emissions for meeting the water needs of both the locals and the visitors proving some little income from the increased wind power sales. The benefits and costs of the scenarios studied taking into account externalities such as social and environmental costs are explicitly discussed in the deliverable D 2.3.

In task 3.3 the issue of tariffs is under investigation. For Milos case study, the existing tariff scheme increases the operational cost of the investor if the schedule of the desalination plant is



177

based on wind power estimations. However, such a co-operation is more profitable for the rest stakeholders on the island.

Suming up

Desalination plant installation provides water at significantly lower prices (1/4th of the current cost) and if installed at the same period with RES under medium –high penetration reduces wind power curtailment with benefits to the owner of the existing wind park even if he does not make any additional investment.

Moreover, such an installation decreases significantly compared to no RES addition, the additional demand for the power system and thus, both the power system fuel cost and the emissions.

Co-operation of RES and desalination during the scheduling, according to the current tariff scheme for loads in MV in Greece may not be favorable but,

• The fuel cost for the operator of the island can be reduced.

• The company of the additional or existing wind park (depending on who invests) increase their profits.

• The operator of the island has more profit compared to the income loss of the desalination plant owner.

• The municipality meets the water demand at lower emission levels.Therefore, the results from tasks 2.3 and 3.3 are expected to have much interest regarding which scenario is more profitable for the society and which tariff scheme could help in better co-operation of RES and the desalination plant.



178

7.3 Case study 2 results-Mljet Two main scenarios with various sub scenarios have been considered for Mljet Island Scenario 1: Each one of the 3 existing desalination plants, installs RES to meet its own energy

demand. Scenario 2 : The Hotel Odisej, major energy and water consumption on the island couples RES

with a desalination plant.

7.3.1 Scenario 1 The monthly electricity demand for desalination for each of the existing plant is provided in

Table 7-15. The total demand represents 2.86% of the island demand. Its maximum penetration reaches 4.2% during June.

Table 7-15 Monthly Electricity demand for each desalination plant on Mljet island(kWh)

Month Kozarica Sobra Blato Total January 330.7 1976.5 3451.9 5759 February 494.7 1149 3117.8 4761.5 March 480.5 2492 2191 5163.5 April 465 1881 2120.3 4466.3 May 484.6 3390.4 4377.2 8252.2 June 1643 7203 8481 17327 July 1697.8 8281.1 9411.6 19390.5 August 2502.7 12534.3 12030.1 27067.1 September 1218 6153 8151.1 15522.1 October 1061.2 2800.3 3013.2 6874.8 November 1157 1489 2916 5562 December 1008.5 1792.8 3013.2 5814.6 Total(MWh) 12.54 51.14 62.27 125.96

If for every desalination plant a RES unit is to be installed in order to meet at least the annual

electricity demand, then the electricity demand from the upstream network would change. Two sub-scenarios have been considered one with wind and another with PV.

7.3.1.1 IMPACT ON ENERGY BALANCE

7.3.1.1.1 Wind energy Sub-scenario The wind turbine installed considered is a Fuhrlander 30 of 33kW installed capacity. The wind power produced for each wind turbine, the total wind power production and the

penetration in the demand of the island is provided in Table 7-16. Due to the fact that if wind power is going to be installed on the island will be installed in the eastern part, the wind power penetration in its eastern part is also provided.

The electricity demand after this installation is provided in Table 7-17 for Kozarica,

Table 7-18 for Blato and Table 7-19 for Sobra.



179

Table 7-16 Wind power production and penetration levels for Mljer

Wind Turbine (kWh)

Wind Turbines (kWh)

Penetration total island (%)

Penetration eastern part (%)

January 5222.155 15666.47 5.56 10.5 February 8345.902 25037.71 10.41 20.54 March 10128.45 30385.35 11.68 23.06 April 6853.423 20560.27 7.17 14.23 May 7682.375 23047.13 7.24 16.79 June 5450.806 16352.42 3.99 9.83 July 8420.413 25261.24 4.32 10.48 August 6436.993 19310.98 2.69 6.45 September 6867.526 20602.58 4.56 10.84 October 6331.414 18994.24 5.52 13.1 November 10444.81 31334.43 13.15 25.2 December 5172.084 15516.25 5.78 10.94 Total(MWh)/Average penetration (%) 87.356 262.07 5.95 13.15

Table 7-17 Updated energy balance in Kozarica with wind power

Month Energy Purchased (kWh) Energy Sold (kWh) Net Purchases (kWh)

January 138 5053 -4,916

February 146 8,562 -8,416

March 318 8,691 -8,374

April 94 6,832 -6,738

May 1248 5,426 -4,179

June 1057 4,573 -3,516

July 350 7,175 -6,825

August 386 5,013 -4,627

September 364 6,214 -5,850

October 408 6,513 -6,105

November 240 9,487 -9,247

December 647 3,515 -2,868

Annual 5394 77,055 -71,661

For all desalination plants there is flow from mainland Croatia during August.



180

Table 7-18 Updated electricity balance in Blato with wind power

Energy Purchased (kWh) Energy Sold (kWh) Net Purchases (kWh)

January 2,003 4,029 -2,026

February 1,068 6,583 -5,515

March 674 8,673 -7,998

April 689 5,542 -4,854

May 615 6,287 -5,671

June 4,875 2,251 2,624

July 3,930 4,202 -272

August 6,855 1,510 5,345

September 6,856 2,626 4,230

October 1,607 4,937 -3,330

November 970 8,515 -7,545

December 1,798 3,918 -2,119

Annual 31,941 59,072 -27,131

Table 7-19 Energy balance after WT installation for Sobra


January 1,145 4,465 -3,320

February 410 8,177 -7,767

March 967 7,808 -6,842

April 823 5,680 -4,857

May 855 5,975 -5,120

June 1,004 4,604 -3,600

July 887 6,558 -5,670

August 8,677 1,569 7,108

September 3,593 4,004 -412

October 3,791 4,432 -641

November 2,575 6,595 -4,020

December 4,930 2,287 2,643

Annual 29,657 62,154 -32,497



181

7.3.1.1.2 Solar Energy Using HOMER software, the optimal sizing of PV plants to meet the demand of each desalination

plant as described in Table 7-15 has been made. The installed capacity, the installation slope and the Azimuth Angle for each case is provided in Table 7-20.The total capacity is 95.12kW.

Table 7-20 Characteristics of the PV installations at each Desalination plant

Blato Sobra Kozarica Installed Capacity (kW) 44.9 40.8 9.42

Installation Slope (South) (°) 36 35 43

Azimuth Angle (°)(West) 12 1 10

Table 7-21 PV production and penetration levels for Mljet-Scenario 1

PV Production (kWh)

Penetration total island (%)

Penetration eastern part (%)

Penetration western part (%)

January 6706.2 2.38 1.78 3.11 February 7350.2 3.06 2.46 3.79 March 10457.0 4.02 3.3 4.92 April 10296.0 3.59 3 4.34 May 12135.8 3.81 3.79 4.11 June 13680.8 3.34 3.55 3.46 July 15533.2 2.66 2.78 2.78 August 15199.2 2.12 2.16 2.26 September 13157.1 2.91 2.86 3.17 October 11007.7 3.2 3.11 3.51 November 7551.4 3.17 2.48 4.02 December 5443.7 2.03 1.5 2.67 Total(MWh)/Average penetration (%) 128.52 2.92 2.7 3.3

The electricity demand after this installation is provided in Table 7-22 for Kozarica, Table 7-23 for Blato and Table 7-24 for Sobra.

Table 7-22 Updated energy balance in Kozarica-PV


January 221 855 -634

February 315 778 -463

March 641 857 -216

April 274 917 -643

May 2,332 201 2,132

June 1,515 287 1,228

July 634 827 -193

August 587 1,043 -456



182

September 579 1,171 -592

October 623 1,016 -392

November 685 688 -3

December 782 567 216

Annual 9,189 9,206 -17

Table 7-23 Updated electricity balance in Blato-PV

Energy Purchased (kWh) Energy Sold (kWh) Net Purchases (kWh)

January 2,050 1,941 109

February 1,639 2,225 -586

March 941 3,689 -2,748

April 852 3,696 -2,844

May 715 4,493 -3,777

June 3,883 2,354 1,528

July 3,698 2,960 738

August 6,206 1,603 4,603

September 6,387 1,388 4,999

October 1,417 3,469 -2,053

November 1,835 2,415 -579

December 2,181 1,613 568

Annual 31,804 31,846 -42

Table 7-24 Energy balance after PV installation for Sobra


January 1,176 1,915 -739

February 658 2,536 -1,879

March 1,406 3,209 -1,803

April 1,215 3,204 -1,989

May 1,041 3,897 -2,856

June 911 4,445 -3,534

July 898 5,201 -4,303

August 7,533 1,144 6,389

September 3,189 2,149 1,040



183

October 3,473 1,471 2,002

November 4,267 1,125 3,142

December 5,123 681 4,441

Annual 30,890 30,978 -88

7.3.1.2 IMPACT ON THE LOSSES OF THE NETWORK

7.3.1.2.1 Wind energy case. Wind power on the island will be on its eastern part and thus the demand will be reduced on the

eastern part of the island and thus the power flow, especially on the line Croatia-Zaglavac. The minimum impact on the losses will be if the wind power is installed at the end of the under sea cable, i.e. at Zaglavac busbar since the undersea cable losses will be reduced. Increase on the voltage of Zaglavac busbar will have some slight effect on reducing the losses on the rest of the eastern part of the island.

Assuming that the eastern and western part are running independently, then the only impact on the losses will be on the eastern part of the island. In such a case it was noted that for 0.29% of the year, i.e. 25 hours, there will be flow from Zaglavac to Croatia with maximum flow of 27.6 kW neglecting the losses.

As expected, during months with high demand, especially during peak hours, the impact of installing wind power on the eastern part is significant as the comparison between August and December show in Fig. 7.20. The difference in monthly losses avoided and its percentage is shown in Fig. 7.21. The losses avoided amount to 6656kWh per year or 10.35% of the losses of the eastern part of the island or 7.53% of the total losses for meeting Mljet demand. The value is 67.2kWh/kW installed.

0

10

20

30

40

50

60

70

80

90

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour

sum

of l

osse

s pe

r hou

r ty

pe (k

Wh)

August Decenmber

Fig. 7.20 Comparison of losses avoidance due to wind turbine in August and December



184

0

200

400

600

800

1000

1200

1 2 3 4 5 6 7 8 9 10 11 12

Month

(kW

h)

0

4

8

12

16

20

24

(%)

Avoided Losses (kWh) Percentage avoided (%)

Fig. 7.21 Comparison of Monthly losses avoidance due to wind turbine

7.3.1.2.2 Solar energy Since PV will be installed in both Eastern and western part of the island the losses in both parts of

the island will be reduced. On the eastern part, the losses reduction is 3.4% of the corresponding losses, 2183kWh. The total losses avoidance in both parts of the island reach 3293kWh, 3.72% of the total losses of the island.1108MWh are avoided on the western part, 4.6% of the total losses of this part of the island. The value of losses avoidance is 34.6kWh/kWp and in terms of produced energy 25.6kWh/MWh a little bit higher than wind. However, this indice for the eastern part of the island reaches 53.53kWh/kWp. This is lower than the average of wind power-67.1kWh/kWp, but for June and August, the value is higher than wind 11.41kWh/kWp compared to 10.81kWh/kWp for August and 5.55kWh/kWp compared to 4.35kWh/kWp. Fig. 7.23 shows that the losses avoidance clearly follows the PV production pattern. It should be noted that the impact of PV during August and noon hours is significantly higher compared with the losses avoided by the wind turbine shown in Fig. 7.20.

0

100

200

300

400

500

600

700

800

1 2 3 4 5 6 7 8 9 10 11 12Month

kWh

avoi

ded

0

0.7

1.4

2.1

2.8

3.5

4.2

4.9

5.6

Loss

es p

erce

ntag

e(%

)

Losses Avoided (kWh) Percentage avoided (%) Fig. 7.22 Comparison of Monthly losses avoidance due to PV for the whole island



185

0

15

30

45

60

75

90

105

120

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour

sum

of

av. l

osse

s pe

r hou

r ty

pe

(kW

h)

August Decenmber

Fig. 7.23 Comparison of losses avoidance due to PV in August and December



186

7.3.2 Scenario 2

Table 7-25 provides the monthly water demand for the touristic season proportionally to the tourists arrivals, taking into account that the average daily water consumption for this season is 100m3. The energy efficiency of the desalination plant is assumed similar to the one of the existing desalination plants i.e.4.4 kWh/m3.

For the non operating months the assumptions on the water demand are also provided in the same table. During March, when preparations for the season are made it assumed that the average daily consumption is half the one of April. For the rest months the water demand is assumed around 1/10th of the one in April being 2m3 on average. All this demand is assumed during 8 hours 08.00-16.00 as Table 7-26 describes.

Table 7-25 Water demand and corresponding Energy consumption on Hotel Odissey (estimations)

Month Water Demand (m3) Average Daily consumption (m3)

Energy Consumption (kWh)

January 62 2 272.8 February 58 2 255.2 March 372 12 1636.8 April 728.2 24.273 3204.08 May 1546.5 49.885 6804.6 June 3182.9 106.099 14004.76 July 5014.7 161.764 22064.68 August 5963.1 192.357 26237.64 September 3926.2 130.872 17275.28 October 1038.5 33.499 4569.4 November 60 2 264 December 62 2 272.8 Total 22014.1 96862.04

Table 7-26 Water Demand during the months when the Hotel does not operate.

Month Water Demand (m3) 08.00-16.00 Power demand (kW) 08.00-16.00 January 0.25 1.125 February 0.25 1.125 March 1.5 6.75 November 0.25 1.125 December 0.25 1.125

For the hotel operating months, it is assumed that there are two periods of consumption within the day. The period 00.00-06.00 when the water consumption is assumed to be one third of the average daily demand. For the rest hours of the day, the rest of the daily demand is distributed evenly. The water demand per period is provided in Table 7-27.

It assumed that the water tank has capacity of providing 3 successive days of water during August, i.e. 577m3. The minimum content is assumed to be the amount of water to supply for one day the average demand of the current month. In this way, if there is a problem with the desalination plant, there is sufficient time to order water and reduce in the meantime the water consumption.

In order to meet the water demand 3 identical un8its with energy consumption 19kW and production capacity 4.32m3/h have been assumed.



187

Table 7-27 Water demand and corresponding power demand for the desalination plant during the

operating months

Month Water Demand (m3) 00.00-06.00

Water Demand (m3) 0600-23.59

Power demand (kW) 00.00-06.00

Power demand (kW) 06.00-23.59

April 0.337 1.236 1.5165 5.562 May 0.693 2.54 3.1185 11.43 June 1.474 5.403 6.633 24.3135 July 2.247 8.238 10.1115 37.071 August 2.672 9.796 12.024 44.082 September 1.818 6.665 8.181 29.9925 October 0.465 1.706 2.0925 7.677

First, the impact of installing desalination plant without RES is examined and then for various

scenarios of installing RES to compensate for this increase in the demand.

7.3.2.1 DESALINATION WITHOUT RES Table 7-28 describes the summary of the results of the analysis of desalination without RES. The

monthly energy consumption, the peak demand for the desalination plant are provided. The maximum and minimum level of the water within the tank is also provided. One 250m3 tank would be sufficient for meeting the water demand. The utilization factor of the desalination units is relatively low, 43.7%, but for August reaches one unit operates for the whole month and 2 units for the 2/3 of the month.

The simulation starts with the season when there is not need for much potable water, November, in order to start from a season with low content of the storage tank.

Table 7-28 Summary of the demand and water production due to the desalination plant.

Number of Unit-Hours

Peak number of units

Peak demand (kW)


Water produced (m3)

Min Content (m3)

Max Content (m3)

November 14 2 38 247 56.14 24.8 35.2

December 13 1 19 285 64.77 24.8 35.3

January 87 2 38 266 60.45 24.9 48.7

February 172 1 19 247 56.14 24.9 71.8

March 372 2 38 1653 375.68 52.7 138

April 750 3 57 3268 742.73 112 183

May 1169 1 19 7068 1606.36 165 234

June 1371 2 38 14250 3238.64 145 232

July 885 2 38 22211 5047.95 37.7 175

August 241 2 38 26049 5920.23 33.9 65.4

September 13 2 38 16815 3821.59 24.8 36

October 15 1 19 4579 1040.68 24.9 36.8

Total/Max/Min 5102 3 57 96938 22031.36 24 242.762



188

This desalination plant is going to have impact on both the energy and the peak demand of both

the island and Hotel Odisej node as Table 7-29 shows. The total demand of the island increases by 2.2%. However, the maximum change in peak, is significantly lower than the peak demand of the desalination plant, as Fig. 7.24 shows. Therefore, the proposed operation scheme which takes advantage from night tariffs, alleviates the impact on the power system of the island. Moreover, 48.7% of the additional demand of the island due to desalination is distributed to off-peak hours. During the period October-April the demand is distributed only to hours 00.00-07.00 increasing the demand during hours with minimum impact.

Table 7-29 Impact on the Croatian power system due to desalination without RES

Island Hotel Odisej Peak (kW) Energy (kWh) Peak (kW) Energy (kWh)

January 587 281854 60 13186

February 542 240764 60 19194

March 610 261838 60 22390

April 698 289937 91 27206

May 800 325550 170 68091.5

June 982 424396 288 108510

July 1590 606832 363 150790

August 1610 742546 363 178065

September 1150 468872 300 111123.5

October 853 348581 220 75051.5November 585 238596 48 13958December 606 268929 49 13255Total/Max 1610 4499 MWh 363 800.8 MWh

05

1015

2025

3035

40

1 2 3 4 5 6 7 8 9 10 11 12

Hotel Odisej total island Fig. 7.24 Peak demand change on both Odisej and island compared to operation without the

desalination plant

7.3.2.2 DESALINATION COMBINED WITH WIND POWER One Fuhlander 30 and one Tulippo 2.6kW wind turbine are assumed to be added to the network

to compensate for the increase in the demand, providing 95.1MWh annually. The following scenarios of operation have been assumed :



189

a)There is just addition of a wind park and the desalination plant operates as described in sub-

section 7.3.2.1. b)The desalination plant takes into account the estimation of wind power production for

determining the number of desalination units to be used.

7.3.2.2.1 Scenario a) independent operation In this case the wind turbine produces energy provided to the network and the scheduling of the

desalination unit is not affected. What is affected is the economics of the water production and the energy balance. Table 7-30 summarizes the impact of installing RES to meet the demand of the desalination plant. The annual additional demand is 1847kWh or just 0.04% of the annual demand. The Hotel peak is slightly decreased. The percentage change in demand of the island with and without RES when adding the desalination plant is shown in Fig. 7.25.

Table 7-30 Change in peak demand and energy due to Scenario 1-Independent Operation

Island Hotel Odisej Peak (kW) Energy (kWh) Peak Energy (kWh)

January 564 276151 60 7495.8February 519 231736 60 10125.2March 576 250846 55 11445April 698 282457 75 19708.3May 799 317105 170 59716.5June 974 418382 269 102492July 1590 597646 358 141532.2August 1610 735409 362 170756September 1150 461349 319 103373October 853 341726 220 68653.9November 585 227413 49 2820.3December 606 263298 30 7638.8Total(MWh)/Max 1610 4403.52 362 705.76

-6

-4.5

-3

-1.5

0

1.5

3

4.5

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge

in D

eman

d (%

)

With RES Without RES

Fig. 7.25 Change in demand of the island from mainland with and without wind power



190

Wind power penetration in this case is 2.11%, while 98% of the desalination plant demand is met by wind power. However, the lack of coincidence between the desalination plant and the wind power production leads to the need of buying electricity from the grid. The balance of the combination Desalination plant & Wind turbine is shown in Fig. 7.26.

02468

101214161820

-55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35

Power to the grid (kW)

(%)

Fig. 7.26 Balance of power exchange with the grid –scenario 2a-Wind turbines

7.3.2.2.2 Scenario b)-Combined operation In this case it is assumed, that the wind power production estimation is known to the operator of

the desalination plant who tries to operate the plant according to the wind power production and the water demand. The scope is to produce water when larger production of RES is apparent, reducing the impact in the power system operation.

The results regarding the water and power demand of the desalination plant are shown in Table 7-31.

Table 7-31 Summary of water and demand when W/T program helps in determining R/O plant schedule.



Peak demand (kW)


Water produced (m3)

Max Content (m3)

Min Content (m3)

November 125 1 19 2375 539.37 519 32

December 15 1 19 285 64.73 519 38.2

January 0 0 0 0 0 518 456

February 0 0 0 0 0 456 400

March 107 1 19 2033 461.71 519 404

April 168 1 19 3192 724.92 518 364

May 271 2 38 5149 1169 516 82.5

June 750 2 38 14250 3236 185 111

July 1170 3 57 22230 5048 246 164

August 1382 3 57 26258 5963 241 192

September 873 2 38 16587 3767 207 37.1

October 241 2 38 4579 1039.92 288 33.9

Total/Max/Min 5102 3 57 96938 22015.6 519 32



191

The water, energy consumption and number of operating units remain the same for the whole

year but there are monthly variations. During November, month with high wind power capacity and low water consumption, the combination of wind and desalination plant can lead to significantly increased water production which requires double tank capacity compared to the previous scenarios. Since the water consumption is low for the period January-February and the amount of excess water produced during November is significant, it is decided that the desalination plant does not operate for theses months. Thus, less personnel is required for that period. It is characteristic, that during November such an operation leads to positive energy balance, i.e. the demand of the island is always decreased, despite the fact that the total demand of the island is generally higher compared to the previous scenarios.

Moreover, the significant content in the water tank till April, helps so that the demand during May, when the wind power production is lower, is reduced by about 50%. Again there is increase in the demand for the period June-September.

The demand of both the island and the Hotel Odisej is shown in Table 7-32. The change in the energy demand compared to independent operation is shown graphically in Fig. 7.27 and Fig. 7.28 for the hotel and the island. The impact in the power balance with the upstream network is shown in Fig. 7.29.

Table 7-32 Change in Demand and Energy for the Island and Hotel Odisej when co-operation is foreseen

Island Hotel Odisej

Peak (kW) Energy (kWh) Peak(kW) Energy (kWh) January 564 275885 60 7229.7February 519 231489 60 9878March 576 251226 55 11824.7April 698 282381 75 19632.2May 799 315186 173 57791.1June 974 418382 287 102490.3July 1590 597666 358 141549.2August 1610 735609 362 170965.4September 1150 461120 319 103144.7October 853 341726 220 68612.7November 585 229541 40 4947.9December 606 263298 30 7638.7Total(MWh)/Max 1610 4403.51 362 705.7



192

-100

1020304050607080

1 2 3 4 5 6 7 8 9 10 11 12

Month

Con

sum

ptio

n ch

ange

(%)

Fig. 7.27 Impact in the Hotel Odissej demand vs independent operation

-0.8-0.6-0.4-0.2

00.20.40.60.8

11.2

1 2 3 4 5 6 7 8 9 10 11 12

Month

Con

sum

ptio

n ch

ange

(%)

Fig. 7.28 Impact in the total island demand vs independent operation

0

5

10

15

20

25

-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35


(%)

Fig. 7.29 Balance of power exchange with the grid –scenario 2b

7.3.2.2.3 Comparison of the scenarios with wind In Table 7-33 it is summarized how often the combination of the desalination plant and the wind

turbine buys and sells energy to the network. Practically this table summarizes the characteristics of the change in the power imported from Croatia. For around 50% of the time the energy demand from Croatia is reduced with some little differences on how often this happens. If desalination schedule is based on the wind power production, the periods for buying energy from the network



193

are reduced by 5%. More important difference between the scenarios assumed is how often the change in the imported demand is within specific limits or confidence interval. These two indices are important for the operator of the local substation which provides power from mainland to Mljet.

Co-operation of the wind turbines with small desalination units manages the narrowest values of power balance. It is characteristic that for 80.53% of the time, the demand from Croatia mainland will change within the ± 20kW. The change in the energy flow from Croatia is shown in Fig. 7.30, for all the considered scenarios. The difference is smaller for 95% confidence interval but again this is narrower for the case of co-operation.

Table 7-33 Statistical values of the balance of the Desalination &Wind combination

No co-operation With co-operation Negative Values (Buying from network) 36.4% 31.9%

Zero balance 13.7% 16.6%

Positive Value (Selling to the grid) 49.9% 51.5%

Confidence interval 95%, [2.5%, 97.5%] [-37.9,34.2] [-36.32,33.68]

Frequency within [-20kW, 20kW] 72.9% 80.53%

0

4

8

12

16

20

24

-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35


(%)

SC2aW Sc2bW

Fig. 7.30 Impact in the power exchange with the upstream network

The impact on the peak and the energy demand of the island are shown in Fig. 7.31 and Fig. 7.32 respectively. Both peak and energy demand are reduced for winter months and are increased for summer months.

450600

750900

1050

12001350

15001650

1 2 3 4 5 6 7 8 9 10 11 12Month

Isla

nd P

eak

(kW

)

Desalination w/o RES Scenario a Scanario b



194

Fig. 7.31 Impact on the peak of Mljet

150

250

350

450

550

650

750

1 2 3 4 5 6 7 8 9 10 11 12Month

Mon

thly

Dem

and

(MW

h)

Current status Desal w/o RES Scenario a) Scenario b) Fig. 7.32 Impact on the energy demand of Mljet

The maximum change in peak demand is 30kW. For the case without RES, the demand increase is higher than 3% for the period June-September while RES reduces this increase to lower than 3%. For period November-May RES help so that the energy imported from Croatia is reduced by 2.13%.

Regarding the losses for the various scenarios these are shown in Fig. 7.33. Wind power helps so that the losses are significantly reduced compared to the base case scenario of desalination without RES. For all the off-season months the losses of the power system are reduced compared to the current situation. A summary of the annual change in losses in both actual and percentage form is provided in Table 7-34. Although in all cases the total island demand is increased, the fact that the demand is increased in the western part and the production is increased on the eastern part, which generally presents higher losses, compensates for this change of island demand reducing losses. This is the reason for the higher losses decrease when wind turbine and the desalination plant do not co-operate-Scenario 2a.

Table 7-34 Change in losses for the scenarios studied with wind-Scenario 2

No RES No co-operation With co-operation Increase in Demand (kWh) 96938 1847 1875 Losses Change(kWh) 2024.83 -739.6 -657.6 Losses Percentage Change(%) 2.29 -0.84 -0.74

-250

-125

0

125

250

375

500

625

750

875

1 2 3 4 5 6 7 8 9 10 11 12

Month

Loss

es C

hang

e (k

Wh)

w/o RES Sc2aW Sc2bW



195

Fig. 7.33 Impact on the losses of the power system for the various scenarios with Desalination and wind

7.3.2.3 DESALINATION COMBINED WITH PVS The same 3 operating scenarios with wind are also studied for PVs. The characteristics of the

suggested PV plant on the area of the Hotel Odissej, e.g. roof etc are: • 73.9kWp,

• 39o Slope and azimuth angle

• 14o to the west.

The expected production for the whole year is 99.3MWh.

7.3.2.3.1 Scenario a) independent operation Similarly with paragraph 7.3.2.2.1, Table 7-35 describes the impact of PV on the island power

system. The percentage change in demand of the island with and without PV when adding the desalination plant is shown in Fig. 7.34. The total demand from Croatia mainland is reduced by 0.06% compared to no desalination plant with RES penetration of 2.21%. The PV meets 12.4% of the total Hotel and desalination plant demand and 102.44% of the desalination plant demand. However, the demand is not constantly met and the histogram of the power demand from the grid is shown in Fig. 7.35.

Table 7-35 Change in peak demand and energy due to Scenario 1-Independent Operation

Island Hotel Odisej

Peak (kW) Energy (kWh) Peak (kW) Energy (kWh) January 587 276649 60 7976.6February 541 235048 60 13449.78March 605 253743 58.5 14332.1April 698 281993 75 19222.8May 800 316099 170 58680.4June 982 413795 263 97917July 1590 594900 339 138770.6August 1610 730859 360 166127.2September 1150 458805 319 100852.8October 853 340127 220 67038.4November 585 232704 49 8091December 606 264740 30 9080.4Total(MWh)/Max 1610 4399.46 360 701.54



196

-3

-2

-1

0

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10 11 12

Month

Dem

and

Cha

nge(

%)

No PV With PV Fig. 7.34 Change in demand of the island from mainland with and without PV

0

5

10

15

20

25

30

35

-55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75power to the Grid (kW)

(%)

Fig. 7.35 Exchange of Desalination plant &PV

7.3.2.3.2 Scenario b)-Co-operation with PV estimations In this case it is assumed, that an estimation of the PV production is known to the operator of the

desalination plant who tries to operate the plant according to both this estimation and the water demand. The scope is to produce water when larger production of RES is apparent, reducing the impact in the power system operation. The results regarding the water and power demand of the desalination plant are shown in Table 7-36. It should be noted that 18,250m3 are produced during peak hours of the day and that the peak demand of this operating scenario never exceeds 38kW. Moreover, significant amount of energy is bought during evening hours during which the electricity prices are lower, and no PV production is available.

Table 7-36 Summary of water and demand when PV estimation helps in determining R/O plant schedule.



Peak demand (kW)


Water produced (m3)

Max Content (m3)

Min Content (m3)

November 125 2 38 2375 539.4 519 35.8

December 13 1 19 247 56.1 519 509

January 0 0 0 0 0 509 447

February 0 0 0 0 0 447 391

March 115 3 57 2185 496.2 519 385

April 167 3 57 3173 720.6 519 409



197

May 314 3 57 5966 1354.9 512 260

June 702 3 57 13338 3029.1 316 115

July 1169 3 57 22211 5044.2 231 165

August 1368 3 57 25992 5902.9 231 132

September 897 3 57 17043 3870.6 180 76.6

October 299 3 57 5681 1290.2 345 51.4

Total/Max/Min 5169 3 57 98211 22304.2 519 35.8 Table 7-37 presents the impact of this combination in both energy demand of the island and the

Hotel. Fig. 7.36 presents the impact in the energy demand of both the hotel and the island compared to previous scenario. The demand is significantly increased during November for the hotel while change in the island demand remains within ±1%.

Table 7-37 Change in Demand and Energy for the Island and Hotel Odisej when co-operation is foreseen

Island Hotel Odisej Peak (kW) Energy (kWh) Peak (kW) Energy (kWh)

January 587 276383 60 7710.6February 541 234801 60 13202.8March 605 254273 59.4 14864.7April 698 281894 75 19128.2May 800 314996 170 57580.4June 963 412880 263 97014.4July 1590 594861 344 138782August 1590 730780 360 166085September 1130 459026 319 101089October 853 341239 220 68141.4November 585 234838 40 10218.9December 606 264708 30 9043.1Total(MWh) /Max 1590 4400.68 360 702.86

-4

0

4

8

12

16

20

24

28

1 2 3 4 5 6 7 8 9 10 11 12

Month

Cha

nge

(%)

island Hotel Odissej Fig. 7.36 Change in demand of island and Hotel Odisej when desalination plant schedule is

determined by PV vs independent operation.



198

0

5

10

15

20

25

30

35

40

45

50

-40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65


Freq

uenc

y(%

)

Fig. 7.37 Balance of power exchange with the grid –scenario b

7.3.2.3.3 Comparison of the scenarios with Solar energy Table 7-38 presents a summary of the values obtained in the previous paragraphs. Combination of

PVs and desalination plant can significantly increase the percentage of exchange with Croatia to be within ±20kW. These values are significantly higher than the ones with wind turbines due to the better correlation of PV power and water demand and the more discrete nature of wind power compared to PVs. It should be noted that energy from the grid is bought mainly during off-peak hours, since for 29.8% of the peak hours, energy is required from the grid. Only for 2.42% this is higher than 20kW. It should be noted that for 5.9% of the peak hours the power injected to the grid, rest of the hotel, is higher than 20kW, providing significant relieve for the system during these hours.

Table 7-38 Statistical values of the balance of the Desalination &PV combination

No co-operation With co-operation Negative Values (Buying from network) 33.7% 26.1

Zero balance 30% 42.1

Positive Value (Selling to the grid) 36.3% 31.8

Confidence interval 95%, [2.5%, 97.5%] [-38,51.9] [-22.7, 42.61]

Frequency within [-20, 20] 74.5% 91.17

The expected change in the demand injected from Croatia is shown in Fig. 7.38 at 5kW steps. Quite often, around 50%, the flow from Croatia to Mljet remains practically intact. Small desalination plants help so that the peak of the island changes very little. This combination presents the lowest peak demand, 1590kW, among all the cases studied wind and PV. The updated peak for Mljet island is shown in Fig. 7.39.

Moreover, such a combination helps so that the increase of the demand injected from Croatia is only for 2.5% of the year higher than 22.7kW, which is the lowest among all the cases studied. As for the monthly demand from Croatia, the same combination reduces significantly the demand from Croatia compared to all the other scenarios to levels close to the ones before installation of the desalination plant as Fig. 7.40 shows.



199

05

101520253035404550

-40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65


Freq

uenc

y(%

)

Sc2aPV Sc2bPV

Fig. 7.38 Balance of power exchange with the grid –comparison of the scenarios

450

600

750

900

1050

1200

1350

1500

1650

1 2 3 4 5 6 7 8 9 10 11 12

Month

Isla

nd P

eak

(kW

)

DEs w/o RES Scenario a Scanario b Fig. 7.39 Peak for the various scenarios with PV

150

250

350

450

550

650

750

1 2 3 4 5 6 7 8 9 10 11 12Month

Mon

thly

Dem

and

(MW

h)

Current status Desal w/o RES Scenario a) Scenario b) Fig. 7.40 Energy imported in Mljet for the various scenarios studied with PV.

Regarding the losses, Table 7-39 summarizes the change in the demand of the island and the change in the losses. The PV is placed near the load, next to it, actually, but the western part presents significantly lower losses compared to the eastern part due to the significantly lower resistance. This has as impact reduced losses reduction for the months when some additional PV power is injected to the grid and significantly higher losses compared to wind power case when additional demand is required as Fig. 7.41 depicts. Thus, although the demand requested from the



200

upstream network is generally reduced, the losses, a non-liner magnitude, are increased. Due to the increased water production for the case of co-operation of PV with the desalination plant, the losses are increased. Decrease by 50% in the energy export to mainland Croatia leads to 10% increase of the network losses.

-250

-125

0

125

250

375

500

625

750

875

1 2 3 4 5 6 7 8 9 10 11 12

Month

Loss

es C

hang

e (k

Wh)

w/o RES Sc2aW Sc2bW

Fig. 7.41 Impact on the losses of the power system for the various scenarios with Desalination and PV

Table 7-39 Change in losses for the scenarios studied with PV

No RES No co-operation With co-operation Increase in Demand (kWh) 96938 -2480 -1090 Losses Change(kWh) 2024.83 736.09 808.37 Losses Percentage Change(%) 2.29 0.83 0.91

7.3.2.4 COMPARISON OF WIND AND PV FOR THE HOTEL ODISSEJ PV helps so that the flow from mainland is maintained within ±20kW and providing significant

increase in around zero values compared to wind power. Also PV when combined with small desalination units maintains lowest peak than Wind power.



201

7.3.3 Summary

Table 7-40 presents the results from the evaluation of scenario 1 where the current situation is compared with the addition of wind and PV. Table 7-41 summarizes the results for the installation of a desalination plant for the Hotel Odissej, with and without RES, either wind or solar. Fig. 7.42 provides a summary of the results for scenarios under the various configurations studied. Since, there are some differences in the water production, the CO2 emissions level for each m3 of desalinated water was used as an index of comparison as presented in Fig. 7.43. It should be noted that although the demand from the mainland power system is decreased, the emissions are increased. This is due to the fact that the demand for the desalination plant during the summer months, who present higher emission values, as Fig. 3.33 shows, cannot be met by the RES production. Thus, there is increase of the demand during high emission values and higher network losses and decrease of the demand during low emission values. Therefore, reduced demand from the upstream network cannot compensate for the increased emission values. However, the avoided emissions value of produced kWh for Scenario 2 especially when RES co-operate with the desalination schedule, is significantly higher as shown in Fig. 7.44. It should be noted that if no RES is to be installed but only the desalination plant, the increase in emissions will be significantly higher than the one shown in Fig. 7.44, reaching 816.7kg per MWh requested.

Table 7-40 Summary of the scenario 1 results

Current situation

Scenario 1-Wind Scenario 1 PV

Total Island Demand (MWh) 4401.92 4401.92 4401.92 RES Production (MWh) 0 262.07 128.52 RES penetration (%) 0 5.95 2.92 Imported energy to the island (kWh) 4401.92 4139.85 4273.4 Peak Imported (kW) 1580 1574 1576 Exported energy from the island (MWh) 0 357.3 0 Peak Exported 0 27.6 0 Losses (MWh) 88.44 81.78 85.14 CO2 Avoided (tn) 200.8 99.6 NOx Avoided (kg) 365.41 176.9 SO2 Avoided (kg) 1212.59 600.32 Particulate Avoided (kg) 97.64 48.8 Estimated O&M cost Avoidance (€) 11824.1 5800.08



202

Table 7-41 Summary of the scenario 2 results

Wind PV Current situation

Scenario 2-NO RES Scenario 2a Scenario 2b Scenario 2a Scenario 2b

Total Island Demand (MWh) 4401.92 4498.858 4498.858 4498.858 4498.858 4500.131 RES Production (MWh) 0 0 95.1 95.1 99.3 99.3 RES penetration (%) 0 0 2.11 2.11 2.21 2.21 Imported energy to the island (MWh) 4401.92 4498.695 4403.518 4403.509 4399.46 4400.679

Peak Imported (kW) 1580 1610 1610 1610 1610 1590 Exported energy from the island (kWh) 0

Losses (MWh) 88.44 90.46 87.7 87.78 89.17 89.24 CO2 Avoided (tn) -79.166 -6.544 -5.642 -5.958 -4.86 SO2 Avoided (kg) -140.956 -9.557 -5.837 -10.5 -5.209 NOx Avoided (kg) -550.116 -115.208 -105.921 -121.226 -102.065 Particulate Avoided (kg) -44.161 -9.039 -8.533 -9.251 -8.133 Estimated O&M cost Avoidance (€) -4900.23 -42.5568 -46.0784 85.808 21.8736



203

1.9

-0.75

0.02

0.630.9

-0.01

-0.9

-0.6-0.3

0

0.3

0.60.9

1.2

1.5

1.82.1

2.4

Peak Increase(%) Losses Increase (%) Total Demand Increase (%)

SC2 Sc 2aW Sc 2bWSc2aPV Sc2bPV

Fig. 7.42 Summary of the results for Scenario 2.

0400800

12001600200024002800320036004000

SC2 Sc 2aW Sc 2bW Sc2aPV Sc2bPV

CO

2 em

issi

ons

(g/m

3 )

Fig. 7.43 Emissions comparison for the various configurations of Scenario 2-Croatia.

725

730

735

740

745

750

755

760

Sc 1W Sc1PV Sc 2aW Sc 2bW Sc2aPV Sc2bPV

CO

2 avo

idan

ce (k

g/M

Wh)

Fig. 7.44 CO2 avoidance comparison due to RES production-Croatia



204

7.3.3.1 SENSITIVITY ANALYSIS- PV FOR SCENARIO 2 ON THE EASTERN PART OF MLJET Due to increased losses and the increased emission values compared to current situation although

the flow from Croatia is reduced, an additional study for the PV of Scenarios 2, 73.9KW installed on the eastern part of the island was made. Even if the whole capacity is installed at Zaglavac busbar, the one that can provide the lowest decrease on the losses of the undersea cable, the losses are decreased to reach 86.9MWh, the lowest among the configurations of scenario 2. The reduction of the cost for the DNO, even if the average cost for the Croatian power system (49.6€/MWh) is used can reach 193.59€, or 2€/MWh higher compared to installation of PV on the western part at the Hotel bus bar. This significant decrease on the losses, 2.27MWh compared to the scenarios with PV, and the time it takes places, helps in decreasing the emissions for meeting the updated Mljet demand by 1.6tn per year. However, this is not enough to provide finally total reductions on the emissions.

If the PVs are equally distributed to the main buses of the eastern part, the losses are further reduced by 573kWh, further reducing CO2 emissions by 430kg and reducing cost by 28.42€.

7.3.4 Conclusions From the above results some conclusions can be drawn. Addition of RES on the current configuration of the Mljet power system-Scenario 1 can help

reducing both losses and emissions. For similar capacity however, 99kW of wind and 95kW of PV, the value of wind for the power system is significantly higher. This is due to much higher capacity ration of the wind compared to the PV and the fact that wind must be installed on the eastern part that presents significantly higher losses. Generally, it is desirable for the Distribution Network Operator (DNO) of the mainland to have installations of RES on the eastern part of the island. Therefore, the pricing or the subsidies should be slightly higher on the eastern part due to this fact. However, generally the DNO should prefer wind power on the eastern part of the island.

If the major consumer on the island, Hotel Odissej, is going to install a Desalination plant to meet their own water needs and no RES is foreseen, then the cost will increase but most importantly both network losses and emissions will be increased. Increase in emissions will be 816.7g/kWh due to increase of the demand during summer months.

Installation of RES will decrease both emissions and losses of the network. Especially for the wind power the losses will be lower than the losses of the grid currently, due to its installation on the eastern part of the island which presents much higher losses. In terms of network it is more desirable to install RES to compensate the increase in the demand on the eastern part and not on the consumption bus. The highest benefit for the DNO comes if PV is installed without co-operation with the desalination plant in the scheduling of the module (Sc2aPV). The demand in such a case is increased during hours with rather low load and the production is increased during hours with high load, reducing in this way the emissions avoided. The decrease in the energy cost of producing water will reach 0.5€ct/m3. The lowest emission increase is for the co-operation of PV large desalination units-(Sc2bPV). In this case the CO2 emissions is as low as 217.8g/m3 of produced water, much lower than the 3.6kg/m3 if no RES is to be installed. Some small differences in the cost between the cases of PV will be alleviated due to the fact that the amount of energy requested by the desalination plant will be higher. With energy selling prices lower than 5€ct/kWh, the scenario of co-operation of desalination and PV becomes the most profitable of all. In the case of wind, the co-operation of wind power with desalination increases slightly the cost for the DNO but with around 1 tn lower CO2 emissions. Therefore with a cost of emissions trading at 3.52€/tn, the co-operation of wind with desalination is the most profitable scenario for wind power.



205

Generally co-operation of RES with desalination plant decreases the emissions for each m3 of produced water, compared to no co-operation. PV presents lower emissions than using wind power for each m3 of produced water. The sensitivity analysis on the location of the PV for scenario 2-eastern or western shows that the value of PV production is 2€/MWh higher if this is installed on the eastern part, reducing the losses by 2.27MWh and the CO2 emissions by 1.6tn annually. Scattering the same capacity of PV to the eastern part of the grid of the island would reduce the losses 0.57MWh and the CO2 emissions by additionally 430 kg annually.



206

7.4 Case study 3 results-Cyprus Here a different approach for the analysis of the impact of desalination is followed. It is examined

what can be the impact of a Desalination plant on the island of Cyprus when it is partly utilized to reduce the wind power curtailment. It is assumed that for year 2005 the authorized wind power capacity of that period, 289MW is in operation. The fact that wind power should co-operate with steam turbine units leads, especially the periods of low demand, to significant wind power curtailment. The estimated amount of wind power curtailment is shown in Table 7-42. This amount of wind power curtailment, if it was possible to be utilized, would produce significant amount of water, as presented in the same table. However, wind power curtailment cannot be fully exploited because this necessity will exist for no more than 900 hours, and significant amount of power would be required to exploit the amount of wind. This would lead to extremely low utilization factor of a costly device like the Desalination RO plant. Therefore, a compromise should be made so that the same amount of water is produced by reasonable number of units with satisfactory utilization factor.

Table 7-42 Estimated wind power curtailment, water to be produced and constant number of units to operate

Months Estimated wind power curtailment(MWh)

Estimated water to be produced.

Desalination units operating for the whole month

January 2007.4 446,088.9 2 February 2358.6 524,133.3 2 March 4402.2 978,266.7 5 April 5459.6 1,213,244.4 6 May 5233.4 1,162,977.8 6 June 2155.3 478,955.6 3 July 234.8 52,177.8 1 August 438.3 97,400 1 September 1444.8 321,066.7 2 October 3024.4 672,088.9 4 November 3680 817,777.8 5 December 3626.1 805,800 4 Total 34604 7,569,978 The following scenarios have been simulated and compared for the whole year Scenario 1: Business as usual (no addition of desalination) Scenario 2: Use of 14 desalination units of 1 MW. These units would be able to produce water,

using amount of energy equal to the expected wind power curtailment. For each month the number of units that could produce 75% of the monthly water demand was identified and decided to operate for the whole month while the rest would operate when wind power curtailment would be expected. The amount of units that could operate for the whole month for this scenario is shown in Table 7-42. For instance during March 5 units would operate during the whole month and the rest 9 for the hours when wind power is expected to be curtailed.

Scenario 3: Production of the annual quantity of water of Table 7-42 using 4 units all the time irrespective of the month.



207

Scenario 4: Production of water of Table 7-43 using constant number of units, updating the scheduling of the units every about 120 hours without increasing the number of operating units when curtailment takes place.

Scenario 5: Production of water of Table 7-43 using constant number of units, updating the scheduling of the units every about 120 hours but increasing the number of operating units to 7 when wind power curtailment is foreseen.

Further explanation for scenarios 4 and 5 is provided in the following paragraphs. 7.4.1 Explanation of scenarios 4 & 5.

The reason for testing and analyzing scenarios 4 and 5 is the fact that significant amount of units are required for scenario 2 with low utilization. More specifically, 8 out of 14 units will operate only during the hours when wind power curtailment occur, about 900 hours. 6 units will operate continuously for additional 3 months.

Moreover, during summer period, when more potable water is required, the amount of water produced according to Table 7-42 is very low. To face these both disadvantages of scenario 2, it assumed that half of the annually water demand is distributed equally to all months and the rest is distributed proportionally to wind power curtailment as provided in Table 7-42. Thus, the amount of water to be produced and the number of desalination units required to produce it is provided in Table 7-43.

Table 7-43 Suggested number of desalination units & water production per month

Months Number of units Water January 3.4 555,434 February 4 594,456 March 5 821,523 April 5.9 939,012 May 5.5 913,879 June 3.6 571,868 July 2.2 358,479 August 2.3 381,090 September 3.1 492,923 October 4.2 672,089 November 5.1 817,778 December 4.4 805,800 Total 3.9 7,570,000

Due to the fact that the units of Table 7-43 are not an integer number and the ability of increasing

the number of operating units when curtailment is foreseen-Scenario 5, it is assumed that the water demand of Table 7-43 should be met with in one month and the schedule for the operating desalination units for each month is updated at the following time steps:

[0 120 240 360 480 600 660] and submitted to the system operator. The initial number of operating desalination units is sufficient to meet 75% of the water demand

to account for the expected wind power curtailment for the month that is considered a priori unknown for the simulation program developed. Then at each of the above time-steps, the number of the desalination units is calculated by (2), where month_demand is the monthly water demand according to Table 7-43, curr_produced is the water produced till the time of update, while the denominator expresses the number of hours remaining till the end of month.



208

10005.4_

___ ⋅⋅⎟⎟⎠

⎞⎜⎜⎝

⎛−−

−=

passedhrshrsmonthproducedcurrdemandmonthroundunitsDes

(2)

Practically an average level of production is suggested for the best period. If there is any deviation, then in the next update this will be settled.

For the last update, hour 660, formula (3) is used to ensure that the monthly water demand will be met.

⎥⎥

⎤⎢⎢

⎡⋅⋅⎟⎟

⎠

⎞⎜⎜⎝

⎛−−

−= 10005.4

____passedhrshrsmonth

producedcurrdemandmonthunitsDes

(3)

Any surplus of water produced is transferred to the next month at 50% till the end of the summer,

and 75% for the rest of the months. 7.4.2 Results

In this section details on the wind power curtailment, the additional cost for EAC and the water produced are provided for each month. For scenario 2 the number of constant operating units and part-time operating units is provided, e.g. for January (2+12)

Table 7-44 Results –January-Cyprus

Shedding (MWh)

Water produced (m3)

Desalination load met by RES (%)

Wind power shedding reduction (MWh)

EAC additional cost (€/MWh)

Max Units Number

Scenario 1 2008.771

Scenario 2 (2+12) 1376.0394

501333 28.05 632.0 22.49 14

Scenario 3 1817.24 661333.3 6.44 191.5 52.057 4

Scenario 4 1844.895 568000 6.41 163.9 52.128 4

Scenario 5 1621.491 568666.7 15.13 387.3 42.363 7

Table 7-45 Results –February-Cyprus

Shedding (MWh)

Water produced (m3)



EAC additional cost (f/MWh)

Max Units Number

Scenario 1 2361.67

Scenario 2 (2+12) 1524.35 601111 30.95 837.3131 11.094 14

Scenario 3 2028.61 597333 12.39 333.058 48.728 4

Scenario 4 2018.613 589333 12.93 343.0583 48.556 6

Scenario 5 1917.356 600000 16.46 444.3154 41.918 7

Table 7-46 Results –March-Cyprus

Shedding Water Desalination Wind power EAC Max



209

(MWh) produced (m3)

load met by RES (%)

shedding reduction (MWh)

additional cost (€/MWh)

Units Number

Scenario 1 4402.2

Scenario 2 (5+9) 3254 1022600 24.94 1148.2 29.881 14

Scenario 3 4094 661333.3 10.36 308.17 47.936 4

Scenario 4 4014.9 832000 10.3444 387.2945 47.948 6

Scenario 5 3840 831000 15.08785 564 42.645 7

Table 7-47 Results –April-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 5460

Scenario 2 (6+8) 4128 1184000 24.99 1331.36 32.261 14

Scenario 3 5237 640000 7.72 222.2456 51.317 4

Scenario 4 5109 946666.7 8.21 349.7952 51.215 7

Scenario 5 4971 940888.9 11.54 488.7952 47.549 7

Table 7-48 Results -May-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 5233

Scenario 2 (6+8) 4077 1185778 21.65997 1155.776 36.444 14

Scenario 3 5030 661333.3 6.820339 202.9733 52.847 4

Scenario 4 4940 912000 7.140197 293.0337 53.138 6

Scenario 5 4749 918000 11.71655 484.0108 47.960 7

Table 7-49 Results -June-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 2155.264 Scenario 2 (3+11) 1374.18 629111.1 27.59022 781.0791 26.823 14

Scenario 3 1948.56 640000 7.177292 206.706 51.963 4

Scenario 4 1975.62 573333.3 6.962884 179.6424 52.271 4

Scenario 5 1764.24 584222.2 14.87358 391.0264 43.382 7



210

Table 7-50 Results -July-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number


Scenario 3 232.7 661333.3 0.07 2.17 60.163 4

Scenario 4 213.8 349333.3 1.34 21.03 58.675 3

Scenario 5 178.5 341777.8 3.67 56.37 55.103 7

Table 7-51 Results -August-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 438 Scenario 2 (1+13) 259 542,000 7.36 179.4 13.870 14

Scenario 3 382 661,333 1.88 55.99 57.261 4

Scenario 4 402 376,000 2.13 35.99 57.369 3

Scenario 5 337 392,000 5.74 101.3 53.120 7

Table 7-52 Results -September-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number


Scenario 3 1289.5 640000 5.39 155.35 52.837 4

Scenario 4 1322.5 493333.3 5.51 122.35 52.640 4

Scenario 5 1167.6 492000 12.52 277.23 44.673 7

Table 7-53 Results -October-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number



211

Scenario 1 3024.4 Scenario 2 (4+10) 2020.3 782222.2 28.52 1004.07 33.798 14 Scenario 3 2862.2 661333.3 5.45 162.13 54.462 4 Scenario 4 2872.1 621333.3 5.45 152.30 54.361 4 Scenario 5 2623.2 610888.9 14.59 401.15 43.011 7

Table 7-54 Results -November-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number


Scenario 3 3483.3 640000 6.73 193.77 54.273 4

Scenario 4 3480.2 666666.7 6.56 196.82 54.484 5

Scenario 5 3213.2 664222.2 15.52 463.86 44.322 7

Table 7-55 Results -December-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 3626.1 Scenario 2 (4+10) 2482.3 861333.3 30.82 1194.72 25.577 14 Scenario 3 3251.7 661333.3 14.29 425.37 48.312 4 Scenario 4 3252.7 680000 13.87 424.37 48.705 5 Scenario 5 3015.9 669333.3 21.95 661.11 38.374 7

Fig. 7.45 presents a comparison of the wind power curtailment for the various scenarios studied.

The lowest wind power curtailment is achieved for scenario 2 and then for scenario 5. Constant operation of the desalination plant helps the least in avoiding wind power curtailment mainly in Spring. Inverse is the analogy for the RES share in meeting desalination load as Fig. 7.46 shows. Scenario 2 manages to maintain significant RES share for meeting the desalination load significantly higher than the average penetration of wind power on the island. In all scenarios RES share is decreased during summer months and is increased especially during November-December. Scenario 5 manages to maintain significantly higher RES share than scenarios 3 and 4. RES share in scenarios 3 and 4 however, is a benefit from increasing the demand since part of the otherwise curtailed wind power is now exploited.

This affects the additional cost for EAC to meet the additional desalination load as Fig. 7.47 shows. Higher values of RES share reduce the additional cost since lower quantities of energy are required from the oil-fired units. Especially for scenario 2, there are months with lower than 10€/MWh for the additional production from EAC’s units. The additional cost is in all the cases apart from July lower than 40€/MWh. Scenario 5 manages to maintain lower additional cost but always above 40€/MWh and below 56€/MWh. Generally, the additional fuel cost for all scenarios



212

is lower than the average fuel cost, 60.42€/MWh for EAC without wind power, leading to slightly lower average cost of fuel compared to scenario 1 as Table 7-58 shows.

0

800

1600

2400

3200

4000

4800

5600

1 2 3 4 5 6 7 8 9 10 11 12Month

Cur

tailm

ent (

MW

h)

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

Fig. 7.45 Expected wind power curtailment for the scenarios studied-Cyprus

Fig. 7.48 presents the monthly water production. Scenarios 4 and 5 present much smoother water production compared to scenario 2 reducing the need for size of the storage tank.

0

5

10

15

20

25

30

35

1 2 3 4 5 6 7 8 9 10 11 12Month

RES

Sha

re(%

)


Fig. 7.46 RES share for Desalination under scenarios 2-5-Cyprus.



213

08

16243240485664

1 2 3 4 5 6 7 8 9 10 11 12Month

Add

ition

al C

ost f

or E

AC

(€/M

Wh


Fig. 7.47 Additional cost for EAC to meet the additional load-Cyprus.

0

200

400

600

800

1000

1200

1400

1 2 3 4 5 6 7 8 9 10 11 12Month

Wat

er P

rodu

ced

(x10

00m

3)


Fig. 7.48 Monthly water produced scenarios 2-5.

The duration curve of operation for the desalination plants is provided in Fig. 7.49. 4 units will operate for more than 50% of the year. The operating hours for the rest 3 units are drastically decreased and 6th or 7th unit is required for less than 20% of the time. Scenario 5 presents better utilization of units 6 and 7 compared to scenario 4, which requires for few periods of February-May 6 units operating, and only during April, the month with the highest wind power curtailment, the operation of the 7th units is required for 180 hours only.



214

0102030405060708090

100

0 1 2 3 4 5 6 7Number of desalination units

Dur

atio

n (%

)

Scenario 4 Scenario 5

Fig. 7.49 Duration curve for the operation of the desalination units for scenarios 4 and 5.

7.4.3 Summary -Cyprus

Table 7-56 presents a summary of tables in sub-section 7.4.2 for the whole year for quick comparisons. More details are provided in Table 7-58, regarding the impact on the power system of the island, the cost and the emissions avoided. The change in the production of each type of the units and the impact in the fuels cost is provided in Fig. 7.51. The more expensive units are mostly influenced by the introduction of wind power on the island of Cyprus. Gas turbines and more expensive Moni power station units reduce their output significantly. Fig. 7.52 presents the increase in emissions due to the operation of the desalination plant. The percentages are quite close but in all case when wind power curtailment is exploited, the percentage increase is lower –Comparison of Scenario 5 to the rest of the cases.

Table 7-56 Summary for the scenarios studied for the whole year-Cyprus

Shedding (MWh)

Water produced (m3)




Max Units Number

Scenario 1 34066

Scenario 2 24233 8,854,000 24.81 9,883 27.951 14 Scenario 3 31658 7,786,666 7.02 2,459 52.714 4 Scenario 4 31448 7,608,000 7.8 2,670 52.088 7 Scenario 5 29396 7,613,333 13.78 4,721 44.9 7



215

14.7

5.1

14.8

3.6

24.8

14.6

4.77.0

14.6

7.8

14.67

4.37

13.78

0.0

4.7

0

4

8

12

16

20

24

28

RES pentration RES curtailment RES share onDesalination

(%)

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Fig. 7.50 Comparison of various scenarios regarding RES shares-Cyprus

10.9220.29

61.98

90.26

15.66 15.010.0219.31

61.75

90.26

15.08

90.3

61.6

19.210.0

0102030405060708090

100

Vaslikos Dekheleia Moni Gas Turbines Fuel cost

Red

uctio

n (%

)

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Fig. 7.51 Comparison of the various scenarios with current operation regarding change in

production and cost-Cyprus

4.4 4.4 4.44.7 4.7

3.9

4.84.6 4.6

3.6

4.74.3 4.4

3.84.2

3.3

00.5

11.5

22.5

33.5

44.5

5

CO2 SO2 NOx PM-10

Cha

nge

(%)


Fig. 7.52 Increase of emissions for the various scenarios compared to Scenario 1



216

The annual costs and benefits for various stakeholders regarding addition of the desalination system are provided in Table 7-57. The investors on wind parks would prefer the scenarios with least wind power curtailment to reduce the income loss. Both EAC and the municipalities would have benefits, since water is produced by lower impact on the operation of the local units reducing both operational costs and emissions. However, the energy cost for the owner of the desalination units is higher for scenarios 2 and 5. This is mainly due to the fact that more units are required and the peak demand for these two scenarios is increased. The cost difference between scenario 4 and 5,which produce almost the same amount of water is mainly due to the peak charge which for scenario 5 is 103,000€.

Table 7-57 Impact of Desalination system for the various stakeholders in Cyprus

Scenario 2 Scenario 3 Scenario 4 Scenario 5 W/T investors benefit (€) 621,625.1 152,229.6 165,505.4 295,229.2 EAC additional cost (€/MWh) 27.95 52.71 52.08 44.90 Desalination owners Energy cost (€)

4,622,667 3,774,960 3,653,045.04 3,754,940.4

Energy cost for water (€/m3) 0.522 0.485 0.480 0.493 Cyprus municipalities CO2 emissions(kg/m3 water) 2.59 3.17 3.13 2.91



217

Table 7-58 Comparative table of all the scenarios studied for Cyprus

Current situation Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

Demand (MWh) 4354381 4354381 4394224 4389421 4388617 4388641

Peak Demand (MW) 849.9 850.9 853.9 851.9 851.9

RES Production (MWh) 0 639342.6 649175.6 641751.2 641961.2 644012.6

RES penetration (%) 0 14.68 14.77 14.62 14.63 14.67

RES curtailment (MWh) 0 34066.43 24233.46 31657.84 31447.83 29396.47

RES curtailment share(%) 0 5.06 3.6 4.7 4.67 4.37

Thermal units production (MWh) 4354381 3715038 3745049 3747670 3746656 3744629

Peak Thermal Station (MW) 849.9 841.47 842.47 845.47 843.47 843.47

Vasilikos units (MWh) 2939044 2641484 2658867 2661907 2661820 2661043

Dekheleia Units (MWh) 1262575 1015389 1027510 1027144 1026324 1025217

Moni steam Units (MWh) 151243 58016 58528 58469 58362 58219

Gas Turbines (MWh) 1520 148 142 148 148 148

CO2 Avoided (tn) 00 520803 497861 496144 497027 498671

SO2 Avoided (tn) 0 1082.9 1035.33 1031.8 1033.65 1037.07

NOx Avoided (tn) 0 3688.735 3548.69 3546.49 3554.67 3566.98

Particulate Avoided (tn) 0 87.34521 83.5385 83.1624 83.2828 83.5373

Estimated Fuel cost (k€) 266132.483 224453.5 225567.1 226300.5 226236.8 225991.7

Estimated Fuel Cost of Energy (COE)- €ct/kWh 6.112 6.042 6.023 6.038 6.038 6.035

.

218

7.4.4 Conclusions of the case studies -Cyprus If Cyprus is to be operating under high RES, wind in this case, penetration, some of the produced

energy 5.1% will be curtailed. This amount of energy is sufficient to provide 7.6 million m3 from a desalination plant of average efficiency of 4.5kWh/m3. However, the fact that the wind power curtailment takes place around 900 hours per year and the fact that the installed capacity should be high means that desalination plant cannot eliminate wind power curtailment. Thus, various scenarios of producing the same amount of water were studied under various levels of exploiting wind power curtailment information

If the amount of water produced followed the wind power curtailment pattern, and some of the desalination units were operating only to meet the requirement for additional demand during wind power curtailment hours, 14 units of 1MW would be required. Such an operation provides the maximum reduction for the wind power curtailment and the maximum percentage of RES production for the desalinated water among the studied scenarios. Thus, such an operation has the minimum impact for the EAC and the additional emissions. However, it increases the energy part of the water cost by about 8%. Moreover, 8 desalination units present very low capacity factor, around 10% which is prohibitive for further cost benefit analysis.

If the same amount of water was produced by constant number of operating units, then 4 units of 1 MW would be required. The increase in demand during hours when wind power curtailment takes place helps in reducing a bit the amount of wind power curtailment. The capacity factor of the operating units is very good, but the additional cost for the EAC and the emissions for producing water are the highest among the scenarios studied.

If the 50% of this amount of water was distributed according to the monthly variation of the wind power curtailment, then both wind power curtailment and cost for the EAC are decreased and so are the emissions for each m3. However 7 units would be required, instead of 4, with 2 of them having capacity factor below 20%.

If the operation of the desalination units is updated every 120 hours but the operator of the island cannot make any suggestions for change according to wind power curtailment, the wind power curtailment avoidance is a result of the variation of the water production and not of actions of the operator himself. On the other hand, if the same number of desalination units are installed and the operator of the island has the ability to give order for increasing water production when wind power curtailment is expected, decreases wind power curtailment by 6.52%. Moreover, the fuel cost for meeting the additional demand is significantly reduced by 13.7% while the additional emissions are reduced by 7%. The slight increase in the energy part of the water cost is due to the increased peak demand, which is however is during off-peak hours. This increase can be alleviated if all the rest stakeholders reduce a bit their benefits and provide some kind of subsidy to the owner of the desalination power plant.

Generally, the wind power penetration level remains at the same level, close to 15%. Taking into account the size of the island and the relatively low wind power curtailment, the desalination plant cannot increase wind power penetration drastically. In all cases studied the demand is met at lower average fuel cost than currently and the emissions are reduced. Increasing the demand during valley hours can help the power system in Cyprus reduce the wind power curtailment and maintain the significantly high technical minima of the operating units of the island. Therefore, motivation to the customers for increasing the demand during these hours, especially during spring should be provided additionally to installing desalination plants and give incentives to increase production during periods when wind power curtailment is expected.

219

7.5 Conclusions of the case studies analysed All the case studies analyzed, Milos,Mljet and Cyprus share two common characteristics, scarcity

of water resources and favourable or very favourable RES potential conditions. Adding solely a desalination plant in order to meet partly their water needs will increase the electricity demand of the island, the fuel cost and the emissions levels of the upstream network even at higher percentage level than the demand, such as the case of Milos. Installing RES will decrease this effect and the local people will have potable water at significantly lower emission levels.

For two of the islands however, Milos and Cyprus, if only additional wind power is to be installed, significant amount of wind power should be curtailed due to difficulties in co-operation with the local grid thus reducing the value of wind power on these islands. Even worse for Milos, if the additional wind turbine considered to be installed belongs to a stakeholder other than the owner of the existing wind park, the latter will face significant income loss. Therefore, RES development even for medium wind power penetration on islands requires a storage method, or at least a demand management method for loads of significant demand such as the desalination plants.

Taking this fact into account, ways of co-operation between RES and desalination have been examined for all the islands studied and especially for the above mentioned two islands as a means of alleviating wind power curtailment.

For the islands that may face problems with RES curtailment even at medium penetration levels, even the simple addition of the desalination plant without any control strategy for exploiting the excess wind power production, would decrease curtailed RES production. RES curtailed can be even more reduced if a strategy for taking into account RES production or RES curtailment when scheduling the desalination plant is taken into account. For the case of Milos wind power estimations are used for the desalination plant scheduling, increasing wind power penetration slightly higher compared to wind power addition only. Thus, the economic and operational impact on the power system for the increased demand due to the desalination plant is limited. At the same time, water to the municipality at significantly lower costs is provided at lower costs compared to the transportation price from mainland. RES penetration in such a case may remain practically the same but the absorbed wind power increases significantly, providing additional income to the potential private RES investors.

For the case of Cyprus, when the TSO is given the permission to ask for increasing water production by the desalination power plant, the impact on the power system of Cyprus is lower compared to the case of the same number of desalination units operating only with production pattern that partially follows the expected monthly wind power curtailment.

In both cases however, the operating cost for the owner of the combination desalination plant and wind power according to the tariff schemes on these islands is higher when these two plants co-operate closely compared to considering these two installations running completely independent. The financing benefits for the rest stakeholders on the island, such as power plant operators, independent wind power investors and the municipalities seem sufficient to finance this shift of more active co-operation of the desalination plant and RES. The externalities for the cost of energy and the cost benefit analysis performed in Task 2.3 will shed more light on this issue. Therefore, the reader is encourage to study them as well within the Deliverable D2.3.Investigation of potential tariff schemes to give incentives for such an operation, are an issue that Task 3.3 will answer. Thus, the reader is also encouraged to study the Deliverable D3.3 as soon as it is available on the STORIES web site.

Finally, for the Croatian island, due to its small size, the interconnection with mainland and the relatively low RES capacity considered, the impact of combination of RES and desalination plant is more limited. However, the benefits by simply adding RES can be significant compared to operation of desalination plant only, especially if wind power is installed on the eastern part of the island. In such networks, where RES curtailment is not an issue, the desalination addition will not provide any other significant impact other than meeting the local water needs. However, the above mentioned deliverables will show which is the most desirable operating mode and RES to be

220

installed wind or PV power for such an island. Moreover, relatively limited wind power exploitation with rather small wind turbines strategically placed in the network grid and taking into account environmental constraints can help significantly the Croatia island networks. The results have shown that the value in terms of emissions avoided per installed kW is higher for wind than for PV due to higher capacity factor of this technology. Therefore, a total ban of installing wind parks on Croatian islands should start being revised into a more flexible framework which takes into account the fragile environment of these islands.

221

8 GENERAL SYNOPSIS &COMPARISONS This deliverable provided results from the simulation of 7 different islands from 6 different

countries and different sizes and electricity demand, from as small as Corvo in the Azores Archipelago to as big as Cyprus in the Mediterranean for various levels of RES penetration combined with energy storage. More precisely, 3 different storage methods, namely batteries, Pump Hydro storage and use of Hydrogen produced by electrolysis driven by RES for electricity production via a Fuel Cell were considered. Additionally a demand side management methodology for desalination plants via Reverse Osmosis was also examined since water scarcity is a common characteristic not only for some of the islands studied, but also of other islands targeted by the STORIES action. The provided results have focused on the impact that the combination of RES and storage means can have on the operation of an island power system mainly in terms of system economics and emissions. Further analysis on the results of the cost benefit analysis for the society as well, will be provided by the Deliverable D2.3.

For the case of batteries, two islands have been studied assumed that they operate without their interconnections to other islands. Thus, the optimum configuration of the proposed power system with as high as possible wind power penetration was identified. For both islands, La Graciosa and San Pietro, scenarios that lead to 80% RES penetration were simulated. The cost for La Graciosa is significantly lower due to its smaller size, and the higher correlation between load and RES especially during the periods of high demand compared to San Pietro. For levels of penetration above this percent, the excess electricity is geometrically increased as well as the installation cost. Unless other methods of exploiting excess energy are exploited, such as hydrogen used for transportation or desalination for potable water, RES penetration cannot be further economically increased for the cases studied. A scenario with lower RES penetration target, e.g. 50% for La Graciosa will provide lower benefits for the grid in the emissions levels and fuel consumption, but with significantly lower excess energy. Such a configuration can be used a first step in introducing RES on this Spanish island and during the construction phase to consider ways of exploiting additional excess electricity and apply demand side measures for smoothening the peak of the island. Then the additional works than can lead to much higher penetration, always considering storage, can be constructed increasing the expected benefits.

Pump hydro is an interesting option for larger scale applications of energy storage, wherever the morphology of the island can help. The case studies of Ios, Cyprus and Corvo have covered all the range of demand and various levels of wind power potential. In all of these cases Pump hydro storage can help in increasing significantly the wind power penetration on these islands. Especially for Corvo, with very favourable RES conditions and good correlation with the load, penetration level of 100% can be easily achieved reducing the operating cost to ½ of the current operating cost.

Pump hydro storage can help in even rather moderate wind power conditions such as Cyprus to increase wind –power penetration and provide a viable solution for wind power development even in such cases. Finally even if any of these islands are interconnected to greater networks, like the case of Ios, the pump hydro will provide significant aid to the weakly interconnected area without decreasing the value of the hybrid plant created. In the meantime, a WHPS (wind hydro power station) can help in further exploitation of the wind potential with significant fuel avoidance.

A general conclusion that may be drawn from both case studies simulated with hydrogen energy storage is that hydrogen may complement renewable energy sources as it has the potential to tackle their intermittent nature and thus to assist in achieving high level RE penetration. The combination and introduction of wind energy and hydrogen storage into the power system of Milos and Corvo showed that the reduction of fossil fuel dependency, the enhancement of security of supply and the decrease of the production of harmful emissions associated by fossil fuel consumption are feasible and can be achieved at a lower than the current power generation cost, more specifically it results in :

• decrease in the power generation cost of the island (ca. 1% for Milos 43% for Corvo)

222

• a huge increase on RE penetration on the island (from 13.4-84% for Milos and from 0% to 80% for Corvo

• a significant reduction in emissions produced (especially CO2) (63- 69%)

• a significant reduction in diesel fuel consumption (from 288,051 L/year to 89,009 L/year) (ca 64%-69%)

For Milos, the thermal units capacity can be also reduced. In none of the cases the impact of using hydrogen as a means of transportation has been studied. The excess electricity could help in producing more hydrogen for this purpose. The results become more favourable as the cost of hydrogen infrastructure decreases.

Desalination cannot produce electricity as the other methods studied. It incorporates storage in the form of potable water. However, the output of such a plant can be scheduled to match as possible RES production and demand as a means of Demand Side Method. The islands selected have scarcity of water and either have or are about to install desalination plants.

The results from the applications show that RES can mitigate the demand in increase of the islands due to the desalination plants and thus the emissions and fuel cost for the power system. This mitigation can be even higher if the schedule of the desalination plant is based on RES estimations or RES curtailment estimations, with significant benefits for both the power system and the owners of the RES installations on the islands.

More specifically for Milos and Cyprus if RES only are installed, wind in specific, for penetrations around 14-16% there will be significant wind power curtailment 5.1% for Cyprus and 15.9% for Milos. This can be reduced by adding desalination plant down to 4.7% and 11.7% respectively. If there is co-operation in the schedule of the desalination plant and the wind power estimations or curtailment, wind power curtailment is reduced to 4.37% and 8.89% respectively. Wind power penetration remains practically the same within ±0.5%, due to increase in demand but the wind power that can be absorbed is significantly higher. Moreover, the locals can respectively have 7.6million m3 and 406 thousand m3 of potable water.

For Mljet, the impact is not as high due to no problems of RES curtailment and relatively lower penetration. Strategic placement of RES can help in reducing losses for the energy transported to the island from the mainland.

Thus desalination cannot only provide a method for meeting the water demand of population with limited access to potable water but also for islands with RES curtailment can help with appropriate management to reduce the amount of RES energy curtailed.

A general conclusion of this Deliverable is that in order to achieve significant penetration of RES on an island system, energy storage or demand management methods or even combination of them is required. Utilizing energy storage will maximize the value of RES for the island, decrease fuel dependency and reduce emissions. At the same time, storage devices can help in achieving higher capacity factor for RES installations for the same potential. There were cases where the cost of energy can be decreased significantly utilizing a combination of storage and RES achieving very high RES penetration levels at the same time.

Therefore, energy storage is a vital component for managing grid issues in Autonomous power system when high RES penetration is considered and efforts in eliminating barriers to installing storage in such networks should be intensified.

223

9 BIBLIOGRAPHY 1 A.F.Wals, R.H.Hendriks, “Economics οf Energy Storage -An analysis of the administrative consequences

of electricity storage”, March 2004, [on-line] www.ecn.nl/docs/library/report/2004/c04006.pdf 2 « ENHANCING SECURITY OF THE UK ELECTRICITY SYSTEM WITH ENERGY STORAGE »,

DTI, CONTRACT NUMBER: K/EL/00319/00/00, 2006. 3 “The future value of storage in the UK with generator intermittency” DTI contract DG/DTI/00040/00/00

(2004). 4 Kelsey B. Lynn, “Estimating the Value of Distributed Energy Storage to California in Combination with

Solar Photovoltaics, Market Investigation, Preliminary Analysis and Recommendations for Further Research”, Electric Power Research Institute, March 2006

5 A.G.Bakirtzis,P.S Dokopoulos “Short term generation scheduling in a small autonomous system with unconventional energy sources” IEEE Trans on power Systems,Vol3, No3, August 1988

6 John Kabouris, “Renewble Energy Sources Penetration in Power generation systems”,PhD Dissertation,School of Electrical and Computers Engineering,NTUA 1992 (n Greek).

7 A.Tsikalakis, I.Tassiou, N.Hatziargyriou, “Impact Of Energy Storage In The Secure And Economic Operation Of Small Islands”, Proc. of MedPower04, Lemessos, Nov 2004, MED04/CH33

8 Manos Psychogiopoulos, Antonis G. Tsikalakis, Nikos D. Hatziargyriou, “Short-term scheduling of Reserve-Osmosis Desalination Loads in Hybrid Systems”, 4th European PV-Hybrid and Mini-Grid Conference Glyfada, Greece, May 29th/30th, 2008, pp 344-349.

9 A.Ter-Gazarian,”Energy storage for Power systems”,Peter Peregrinus IEE Energy Series 6,UK, 1994 10 Edgardo D. Castronuovo,J. A. Peças Lopes, “On the Optimization of the Daily Operation of a Wind-

Hydro Power Plant”, IΕΕΕ Trans On Power Systems, Vol. 19, No. 3, August 2004,pp1599-1606. 11 G.N. Bathurst, G. Strbac “Value of combining energy storage and wind in short-term energy and

balancing markets”, J. Electric Power Systems Research 67 (2003), pp1-8 12 Rau, N.S. Tayor, B, “A central inventory of storage and other technologies to defer distribution

upgrades-optimization and economics”, IEEE Trans. on Power Delivery, Jan 1998,Vol. 13, Iss. 1 pp.194-200 13 Ph.Strauss,W.Kleinkauf,J.Reekers,G.Cramer, G.Betzios “Kythnos Island –19 Years Experience of

Renewable Energy Integration.”.Proceedings of Renewable Energies for Islands-towards 100% RES Supply 14-16 June 2001, Chania, Crete

14 Caralis G., Zervos A. (2006) “Prospects of Wind and Pumped storage systems’ integration in Greek islands”, EWEA 2006, 27February-2March 2006, Athens, Greece.

15 Papantonis D. (2007) “A hybrid system for Ios island”, Presentation in RESTMAC WORKSHOP - BEST PRACTICES OF RES APPLICATIONS IN ISLANDS 31 AUGUST 2007 – Kos island – Greece.

16 Zoulias EI, Lymberopoulos N, Glockner R, Tsoutsos T, Vosseler I, Gavalda O, Mydske HJ, Taylor P, (2006). Integration of hydrogen energy technologies in stand-alone power systems: Analysis of the current potential for applications, J.Renewable and sustainable energy reviews 10: 432–462.

17 G.E. Marnellos, C. Athanasiou, S.S. Makridis and E.S. Kikkinides, (2008). Integration of Hydrogen Energy Technologies in Autonomous Power Systems. Chapter in Book: Hydrogen-based Autonomous Power Systems: Techno-economic Analysis of the Integration of Hydrogen in Autonomous Power Systems, edited by E.I Zoulias and N. Lymberopoulos, Springer-Verlag Ltd, London

18 Floch P–H, Gabriel S, Mansilla C, Werkoff F, (2007). On the production of hydrogen via alkaline electrolysis during off-peak periods. International Journal of Hydrogen Energy 32(18): 4641–4647

19 Grigoriev SA, Porembsky VI, Fateev VN, (2006). Pure hydrogen production by PEM electrolysis for hydrogen energy. International Journal of Hydrogen Energy 31(2): 171–175

20 Michael Felderhoff, Claudia Weidenthaler, Rittmar von Helmolt and Ulrich Eberle. “Hydrogen storage: the remaining scientific and technological challenges”, J.Phys. Chem. Chem. Phys., 2007, 9, 2643 - 2653

21 US DOE, (2002). Fuel cell handbook (Sixth edn). EG&G Technical Services, Inc. Science Applications International Corporation. DOE/NETL–2002/1179

22 Gasteiger H, Kocha S, Sompalli B, Wagner F, (2005). Activity benchmarks and requirements for Pt, Pt-alloy and non-Pt oxygen reduction catalysts for PEMFCs”, Applied Catalysis B: Environmental, 56: 9–35

224

23 Dubravko Sambrailo, Jure Ivic, First land-based plant for RO desalination in Croatia, Desalination,

Volume 132, Issues 1-3, 20 December 2000, Pages 329-335 24 Center of Renewable Energy Sources (CRES), 2001. Development of an information system for the

assessment of the technically and economically exploitable potential of Renewable Energy Sources. Operational Programme for Energy - measure 3.4.3

25 Antonis G.Tsikalakis “Methods for the economic evaluation of Renewable Energy Sources in Island power systems. Application to the island of Crete for the year 2000”, Diploma thesis, School of Electrical and Computers Engineering,National Technical Univrsity of Athens, July 2002.

26 G.C.Bakos and M.Soursos, “Technical Feasibility and economic viability of a grid-connected PV installation for low cost electricity production.” Energy and Buildings J., vol. 34, pp. 753–758, July 2002.

27 Public Power Corporation-Renewables web site http://www.ppcr.gr/ 28 Loucas Savvides, Gerald Dörflinger, Kyriakos Alexandrou, “The Assessment Of Water Demand Of

Cyprus”, Nicosia – Cyprus, October 2001 29 A.G. Tsikalakis, N.D. Hatziargyriou, “Environmental Benefits of Distributed Generation With and

Without Emissions Trading”, J.Energy Policy, Vol. 35, Iss. 6, June 2007, pp 3395–3409. 30 http://www.ucte.org/services/onlinedatabase/_xls/17762P193P198P176P226P1488.xls 31 Maja Bozicevic Vrhovcak, “External costs of electricity production: case study Croatia”, J.Energy

Policy, Vol 33, Issue 11, July 2005,pp 1385-1395 32 Papathanassiou St, and Boulaxis N, 2006. Power limitations and energy yield evaluation for Windfarms

operating in island systems. Renewable Energy, Vol 31, Issue 4, 457-479 33 Katsaprakakis D, Christakis D, 2004. The Wind power penetration in the island of Crete. RES & RUE

for islands international conference. Cyprus. 30-31 August 2004 34 Kaldellis JK, 2002. An integrated time-depending feasibility analysis model of windenergy applications

in Greece. Energy Policy J 30, 267–280 35 Kaldellis JK, Kavadias KA, Filios AE, Garofallakis S, 2004. Income loss due to wind energy rejected by

the Crete island electrical network - the present situation. Applied Energy 79, 127–144. 36 Caralis G. 2008. “Analysis of wind energy with pumped storage” PhD Thesis, National Technical

University of Athens, March 2008. 37 G.Caralis, A.Zervos, “Prospects of Wind and Pumped Storage systems’ integration in Greek islands”,

EWEC 2006, 27th February – 2nd March 2006, Athens 38 Regulatory Authority of Energy (RAE), 2003. Methodology for the assessment of wind penetration in

non-interconnected islands. http://www.rae.gr/k2/apepenetration.pdf, http://www.rae.gr/K2/deliberation-ape.html (in Greek)

39 Manolakos D. Papadakis G. Papantonis D. Kyritsis S. (1998) “A simulation-optimisation programme for designing hybrid energy systems for supplying electricity and fresh water through desalination to remote areas. Case study: the Merssini village, Donoussa island, Aegean Sea, Greece”, Energy 26 (2001) 679–704, July 1998.

40 C Caralis G, Zervos A. (2005) “Assessment of the wind penetration in Autonomous Greek islands. 3rd National Conference for RES, RENES, Athens, 23-25 February 2005. (in Greek)

41 Caralis G., Zervos A. (2003) “A pumped storage unit to increase wind energy penetration in the island of Crete” International Conference: RES for island tourism and water, Crete, Greece.

42 G.Caralis, A.Zervos (2007) “Analysis of wind power penetration in autonomous Greek islands” Wind Engineering, Volume 31, No. 6, 2007, pp.487-502

43 EWEA, 2004. Wind Energy The Facts: An Analysis of Wind Energy in the EU-25, EWEA, Brussels. 44 G.Caralis, Y.Perivolaris, K.Rados, A.Zervos, (2008) “On the Effect of Spatial Dispersion of Wind

Power Plants on the Wind Energy Capacity Credit in Greece”, Environmental Research Letters, Volume 3, 015003 (13pp), January-March 2008

45 Alafouzos V., Vougiouka A., Perivolaris Y., Mourikis D., Zagorakis V., Theodorakakos A. “A Method for Estimating Wind Speed Profile in Complex Terrain Based on an Advanced CFD Tool” EWEC Conference, Athens 2006.

46 Ministry of Agriculture, Natural and environment – Water development department, “Dams of Cyprus” Dec 2001.

225

47 E.I. Zoulias, N. Lymberopoulos, “Techno-economic Analysis of the Integration of Hydrogen Energy

Technologies in Renewable Energy-based Stand-Alone Power Systems”, Renewable Energy Journal, 2007, Vol. 32 (4), pp. 680-696.

market applications for energy storage methods and res ... · contract n°: eie/07/159/si2. 466845...

Documents