Calculating Cost-Supply Curves of Wind Power and Photovoltaic Energy in North Africa using a Geographic Information System


Diplomarbeit, 2012

222 Seiten, Note: 1.1


Leseprobe


Table of Contents

2 List of Figures

3 List of Tables

4 List of Equations

5 List of Annexed Tables

6 List of Annexed Figures

7 Abbreviations

1 Introduction
1.1 Background
1.2 Objective and Scope
1.3 Methodological Approach and Structure of Assessment
1.3.1 Information on Data Processing and Area Correction Factors

2 The Power Sector in North Africa 7
2.1 Renewable Energy in North Africa
2.1.1 Morocco & Western Sahara
2.1.2 Algeria
2.1.3 Tunisia
2.1.4 Libya
2.1.5 Egypt

3 Introduction into Wind and PV Energy Technologies 18
3.1 Wind Energy
3.1.1 Introduction into Wind Energy Systems Technology
3.1.2 Expected Challenges by the North African Environment
3.2 Photovoltaic Energy
3.2.1 Introduction into Photovoltaic Systems Technology

4 Cost-Supply Assessment of Wind Energy in North Africa 30
4.1 Theoretical Potential of Wind Energy
4.1.1 Wind Resources in North Africa
4.1.2 Height Transformation and Roughness Length
4.1.3 Description of Wind Data
4.1.4 Wind Data Set Comparison
4.1.5 The Form Parameter k
4.2 Geographical Potential of Wind Energy
4.2.1 Area Restrictions for Wind Energy
4.2.2 Further Restrictions
4.2.3 Wind Power Installation Density
4.2.4 Results and Discussion of Geographical Potential
4.3 Technical Potential of Wind Energy
4.3.1 Hub Height
4.3.2 Losses
4.3.3 Reference Wind Turbines
4.3.4 Capacity Utilisation
4.3.5 Results for Technical Potential of Wind Energy
4.4 Economic Potential of Wind Energy
4.4.1 Financial Parameters
4.4.2 Cost-Supply Curve of Wind Energy in 2012
4.4.3 Results for Economic Potential of Wind Energy
4.4.4 Cost Development of Wind Energy until 2030 and 2050
4.5 Discussion of the Results

5 Cost-Supply Assessment of PV Energy in North Africa 77
5.1 Theoretical Potential of Photovoltaic Energy
5.1.1 Solar Irradiation
5.1.2 Acquisition and Transformation of Solar Irradiation Data
5.2 Geographical Potential of Photovoltaic Energy
5.2.1 Suitable Area for Ground Installation
5.2.2 Coverage Factor and Results for Ground Installed PV
5.2.3 Area for Building Integrated Installation
5.3 Technical Potential of Photovoltaic Energy
5.3.1 Performance of PV system
5.3.2 Results for Technical Potential of PV
5.4 Economic Potential of Photovoltaic Energy
5.4.1 Financial Parameters
5.4.2 Cost Development of Photovoltaic Energy
5.4.3 Resulting Cost-Supply Curves for Ground Mounted PV
5.4.4 Cost-Supply Curve for PV Systems Installed on Buildings
5.5 Discussion of the Results

6 Inclusion of Infrastructure into the Assessment
6.1 Background and Objective
6.2 Electricity Grid Infrastructure
6.2.1 The North African Power Grid
6.2.2 Cost of Grid Connection
6.3 Transport
6.4 Resulting Cost-Supply Curves Including Grid Connection
6.4.1 Photovoltaic Energy
6.4.2 Wind Energy
6.4.3 LCOE of North African Wind Energy on the European Market

7 PV and Wind Energy in a Broader Context
7.1 Comparison to CSP
7.2 Meeting Electricity Demand with PV and Wind Power Supply
7.2.1 Passive Balancing by Siting and Costs for Active Balancing
7.2.2 Electricity Storage
7.2.3 Vision of the Trans-Mediterranean Grid
7.2.4 Cost Effectiveness and Supply Security

8 Summary, Conclusion and Outlook
8.1 Summary and Conclusion
8.1.1 Conclusion
8.2 Outlook

9 Annex

9.1 Tables and Figures

10 References

Acknowledgement

This thesis would not have been possible without the supervision and support of Prof. Dr. Martin Wietschel and Dr. Mario Ragwitz.

I would like to show my gratitude to Inga Boie at Fraunhofer ISI for her valuable advice and guidance.

I am thankful to Gerda Schubert, Simon Kallfass and Dr. Frank Sensfuß for constructive discussions on GIS and ressource analysis.

Many thanks to David Lehmann, Nils Westerhaus, Ben Link and Jens Brokate for proofreading this diploma thesis.

My gratitude goes out to my family, my friends and my collegues at ISI Fraunhofer for their moral ecouragement and amusement.[1]

List of Figures

Figure 1: The North African region

Figure 2: General methodology in the Cost-Supply assessment

Figure 3: Electricity demand and population in North Africa until 2050

Figure 4: Percentage of power plants older and younger than 20 years as of 2005

Figure 5: Past and projected growth of installed wind power in Egypt

Figure 6: Components of a horizontal-axis wind turbine

Figure 7: Scheme of a variable speed with partial power conversion (DFIG)

Figure 8: Scheme of a variable speed with full power conversion (EESG)

Figure 9: Functionality of a solar cell

Figure 10: Interconnected solar-cells building up PV modules which connect to inverters form PV systems

Figure 11: Scheme of LCOE (Levelised Cost of Electricity) calculation methodology adapted to wind energy

Figure 12: The wind atlas methodology

Figure 13: Logarithmic wind profile with varying roughness lengths z 0 and hub heights

Figure 14: Wind speed data at 10 m height by CRU CL 2.0

Figure 15: Wind speed data at 50 m height by MERRA

Figure 16: Wind speed data at 50 m height by SWERA

Figure 17: Egyptian wind atlas with annual mean wind speeds at 50 m above ground

Figure 18: Tunisian wind atlas with annual mean wind speeds at 10 m above ground

Figure 19: Map of the annual mean wind speeds in Algeria at 50 m above ground

Figure 20: Moroccan wind atlas with annual mean wind speeds at 10 m in MO & EH

Figure 21: Distribution of form parameter k in North Africa

Figure 22: The geographical distribution of land categories and excluded areas

Figure 23: Wind speed at 80 m hub height; excl. areas by geographical assessment

Figure 24: Distribution [%] of turbines in NA according to their nominal power output

Figure 25: Share [%] of installed turbines by manufacturer in North Africa

Figure 26: Power curve, rotor diameter and capacity of wind turbines

Figure 27: Cost-Supply curves depicting marginal LCOE [ct € /kWh] over cumulative, annual power generation [TWh/a] for 2012; G-80, 80 m hub height, variable k

Figure 28: Cost-Supply curve depicting marginal LCOE for 2030; E-82, 120 m hub height, variable k

Figure 29: Simulated electric energy generation by Morocco wind parks

Figure 30: Scheme of LCOE calculation methodology adapted to PV energy

Figure 31: Irradiance spectrum with AM 1.5, AM 0 and black body 5762 ° K

Figure 32: Irradiance on latitude-tilted and vertical surfaces in the observed region at most northern and most southern locations with similar longitudes

Figure 33: Diffuse and direct solar irradiation on surfaces tilted according to latitude

Figure 34: Development in module efficiency of PV technologies starting in 2010

Figure 35: Temperature sensitivity of mono-cSi and thin-film PV modules

Figure 36: Forecasted initial investments for PV systems, LR=20%

Figure 37: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2012

Figure 38: Cost-Supply development for energy generated by PV ground installed systems in North Africa until 2050

Figure 39: Cost-Supply curve depicting marginal LCOE for PV rooftop installations with financial parameters for 2012

Figure 40: Cost-Supply curves for PV rooftop installations from today to 2050

Figure 41: Cost-Supply curves for PV facade installations from today to 2050

Figure 42: Map of the current electricity grid in North Africa; lines ≥ 225 kV only

Figure 43: Sketch of electricity transmission and generation infrastructure

Figure 44: Map of North African transport infrastructure including roads and rails

Figure 45: Cost-Supply curve for ground installed PV with prices for 2012, including distance to grid premium for medium sized PV parks

Figure 46: Map - LCOE of PV ground systems in 2012, including grid premium for midsize PV parks (20-90 MW)

Figure 47: Cost-Supply curve for ground installed PV with prices for 2012, including distance to grid premium for large PV parks

Figure 48: North African Cost-Supply curve for ground installed PV with prices for 2030, including distance to grid premium for midsized and large PV parks

Figure 49: Map - LCOE of wind energy in 2012; including grid premium for midsize wind park (20-90 MW)

Figure 50: Wind energy Cost-Supply curves for 2012, midsize wind parks

Figure 51: Wind energy Cost-Supply curves for 2030, large wind parks

Figure 52: Wind energy Cost-Supply curves for 2030, midsized wind parks

Figure 53: Wind energy Cost-Supply curves for 2012, incl. transmission costs to the European market for large wind parks

Figure 54: Forecasted Cost-Supply curve for wind energy including transmission cost to the European shore

Figure 55: Principle of the parabolic through CSP plant

Figure 56: Operational profile of a reference CSP plant with heat storage and co-firing during three days in August

Figure 57: Ratio between average wind speeds in July and January in Europe and North Africa

Figure 58: Ratio between average irradiance in July and January in Europe and North Africa

3 List of Tables

Table 1: Area of countries and spatial deviation of “ WGS84 ” and “ Albers Equal Area Conic ” from actual country size

Table 2: Electricity consumption and emissions in North Africa, OECD and the world

Table 3: Energy generation by RES in North Africa, OECD and the World

Table 4: Results of comparison by average at 50 m comparison height

Table 5: Pattern and RMS comparison between reference DB and global DB

Table 6: Areas distribution according to land use categories and usability; average power densities on included areas (standard density of 4 MW/km ² )

Table 7: Average, minimum and maximum wind speeds at hub height in incl. areas

Table 8: Installed capacity, annual AC power output and average annual FLH by country and installation type

Table 9: Assumptions for the LCOE calculation of wind energy valid for the year 2012

Table 10: Land use sensitivity analysis for wind energy

Table 11: Areas for PV ground installation, module area and received irradiation

Table 12: Suitable and fully covered rooftop and facade area

Table 13: Annual average irradiance on suitable rooftops and facade areas

Table 14: Installed capacity, annual AC power output and average annual FLH by country and installation type in the year 2012

Table 15: Financial input parameters for the LCOE calculation of PV installation types

Table 16: Financial parameters to calculate LCOE of 2012, 2030 and 2050

Table 17: Grid premium by distance classes for PV ground and WT installations at 3000 FLH no wheeling charges included[2]

4 List of Equations

Equation 1: Mechanical power of the wind converter

Equation 2: Power of undisturbed wind at rotor area

Equation 3: The power coefficient

Equation 4: Energy of a photon

Equation 5: The logarithmic wind profile and height transformation

Equation 6: Calculation of inherent roughness length

Equation 7: Calculation of average wind speeds weighted by polygon sizes

Equation 8: Elevating global atlas ’ average to reference atlas ’ average

Equation 9: Variance calculated weighted by polygon sizes

Equation 10: Root Mean Square error (RMSe) weighted by polygon sizes

Equation 11: Weibull distribution adapted to wind

Equation 12: Regrouped least square method linear regression

Equation 13: Transferring index and magnitude by median rank to logarithmic scales

Equation 14: Wind power installation density

Equation 15: Capacity utilisation (FLH) of wind turbines

Equation 16: Installed capacity considering capacity utilisation threshold

Equation 17: Annual power generation of wind turbine under various scenarios

Equation 18: General of LCOE calculation

Equation 19: Forecasted PV system costs employing the learning curve

Equation 20: Converting LTI to irradiation on vertical surfaces

Equation 21: Irradiation on rooftops and facades

Equation 22: Calculating the operation temperature of the solar-cell

Equation 23: Influence of temperature sensitivity on module efficiency

Equation 24: Annual capacity usage (full load hours) of a PV system

Equation 25: Installed capacity of ground installed PV systems

Equation 26: Installed capacity of rooftop and facade installed PV systems

Equation 27: Annual energy output of the PV system - technical potential

Equation 28: The annuity factor

Equation 29: Cost-supply calculation for PV energy depending on installation type

Equation 30: Forecasted PV system costs employing the learning curve

Equation 31: Grid premium for midsize PV or wind parks

Equation 32: Grid premium for large or very large PV or wind parks

5 List of Annexed Tables

Table (Anex.) 1: Electricity demand of the North African region 157

Table (Anex.) 2: Population of the North African region 157

Table (Anex.) 3: Wind Parks of North Africa 158

Table (Anex.) 4: Geographic restrictions and land use factors for WT & PV systems . 163

Table (Anex.) 5: IUCN Protected Areas Categories System 165

Table (Anex.) 6: BirdLife International - Global IBA Criteria 166

Table (Anex.) 7: Details on employed wind turbines and processing tools prior to the inclusion of the performance ratio 167

Table (Anex.) 8: Remaining area in km ² and as the percentage of the area included after the geographic assessment of wind energy 169

Table (Anex.) 9: Input values for LCOE calculation for wind turbine installation 170

Table (Anex.) 10: Calculation of average best market module efficiency 172

Table (Anex.) 11: Input values for LCOE calculation for PV ground, rooftop and facade installation 173

6 List of Annexed Figures

Figure (Anex.) 1: Wind power Cost-Supply curve for 2012 - baseline scenario; Enercon- E-82, 120 m hub height, variable k

Figure (Anex.) 2: Wind power Cost-Supply curve for 2050; Enercon-E-82, 120 m hub height, variable k

Figure (Anex.) 3: Wind power Cost-Supply curves for 2012, large wind parks

Figure (Anex.) 4: Map - LCOE of wind power in 2012; Gamesa G-80, 80 m hub height, variable k

Figure (Anex.) 5: Map - LCOE of wind power in 2030; Enercon E-82, 120 m hub height, variable k

Figure (Anex.) 6: Map - LCOE of wind power in 2012; including grid premium for large wind park (1000-3000 MW)

Figure (Anex.) 7: Map - LCOE of wind power in 2030; including grid premium for midsize wind park (20-90 MW)

Figure (Anex.) 8: Map - LCOE of wind power in 2030; including grid premium for large wind park (1000-3000 MW)

Figure (Anex.) 9: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2030

Figure (Anex.) 10: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2050

Figure (Anex.) 11: PV ground system FLH and excluded area

Figure (Anex.) 12 Map - LCOE of PV ground systems in 2012

Figure (Anex.) 13: Map - LCOE of PV ground systems in 2030

Figure (Anex.) 14: Map - LCOE of PV ground systems in 2050

Figure (Anex.) 15 Map - LCOE of PV ground systems in 2012; including grid premium for large PV parks (1000-3000 MW)

Figure (Anex.) 16: Map - LCOE of PV ground systems in 2030; including grid premium for midsize PV parks (20-90 MW)

Figure (Anex.) 17: Map - LCOE of PV ground systems in 2030; including grid premium for large PV parks (1000-3000 MW)

7 Abbreviations

illustration not visible in this excerpt

1 Introduction

1.1 Background

Evolution of mankind is strongly connected to its methods of obtaining and utilising energy [3]. Biomass, wind and solar energy have accompanied evolution by nurturing fireplaces, moving sailing ships and drying salines. Combining these not at all novel Renewable Energy Sources (RES) with cutting edge technologies opens up a new era in energy sourcing.

On a global level, the goal for RES is to mitigate greenhouse gas (GHG) concentrations thereby slowing down the global temperature upswing and the inherent climate change. Next to the Kyoto Protocol, which sets binding targets on the reduction of GHG emissions [4], the recent U.N. Convention on Climate Change in Durban, South Africa, decided to establish the Green Climate Fund dedicated to provide “support to developing countries to limit or reduce their greenhouse gas emissions and to adapt to the impacts of climate change” [5]. With 6.5 billion USD pledged by 14 developed countries, the Clean Technology Fund is a further programme to promote low-carbon technologies in 46 developing countries [6].

Regionally, the exploitation of RES gains strategic importance to reduce dependency on the external price and supply volatilities of fossil resources. Furthermore, aging power plants and steeply growing energy demands in North Africa require repowering and extension of the current power plant fleet. Similarly, the EU is experiencing a historic low in reserve power generation capacity. Between 2005 and 2030, generation capacities of 862 GW will have to be built for repowering and meeting the rising electricity consumption [7]. In all these cases, North African RES may be able to contribute and additionally render export revenue.

Locally, the deployment of RES may foster industries and create employment opportunities in manufacturing, construction, operation and maintenance, as Ragwitz et al. (2011) demonstrated for Concentrated Solar Power (CSP) in the MENA region [8].

However, “there is no single energy technology solution that can solve the combined challenges of climate change, energy security and access to energy” [9]. Consequently, RES technologies are not competing but rather complementing each other by their individual characteristics. Ultimately, RES’ economic competitiveness against fossil-fuelled electricity is the key for self-motivated, long-term RES capacity scale-up.

This thesis aims to assess - from an economic and technological perspective - the potentials of wind and PV energy to contribute to the current and future North African and European electricity supply. Available winds and solar irradiation as well as geographic, technological and infrastructure constraints will be included to determine favourable areas for PV and wind energy deployment. How can a power network cope with sizable shares of PV and wind energy? Light will be shed on the broader context of PV and wind energy integration as well.

Due to copyright issues, the map has been removed for publication.

Figure 1: The North African region [10]

illustration not visible in this excerpt

Situated between 37° North and 19° North as well as 16° West and 36° East, the North African region, mapped in Figure 1, encompasses - from west to east - Western Sahara

(EH), Morocco (MO), Algeria (DZ), Tunisia (TN), Libya (LY) and Egypt (EG). Country codes as stated in brackets are according to ISO 3166-1-alpha-2 code [11] and will be used throughout this study.

1.2 Objective and Scope

Considering the targeted energy supply for the North African and European region alike and its designated increase in the share of RES this study aims to answer the question:

How much electrical output is achievable through Photovoltaic and Wind Energy Generation in North Africa today and in the mid- to long-term future, and at what cost?

Thereby the core objective of this diploma thesis is to analyse the North African region, in regard to its economic potential for two renewable energy sources: photovoltaic energy and wind energy. Cost-Supply Curves for the analysed region and the respective energy sources provide Levelised Cost of Electricity (LCOE) in relation to the accumulated available generation potential.

Compared to previous studies, the objective of this analysis is not only to enhance the accuracy of the calculation with up-to-date, better resolution in data inputs as well as higher detail in site selection, but also to broaden the analysis: LCOE and the power generation potential for the years 2030 and 2050 will be estimated. The inclusion of grid infrastructure into the economic assessment shall refine the geographic distribution of potential electrical output. The impact of the North African environment on observed technologies and relevant developments to cope with this impact will gain importance: Concentrated PV - using the highly available direct solar irradiation in Northern Africa - and thin-film technologies will be appraised as well as mono- and polycrystalline PV technologies. The wind energy assessment will include both drive-train concepts - DFIG and EESG.

Finally, it is the goal of this paper to disclose the suitability of PV and wind technologies for a decisive share in energy supply in the North African region as well as possible contributions to the electricity demand of Europe.

Not included within the boundaries of this thesis is a detailed country-specific resource scheduling in the constraints of existing power plant fleets. Favourable regions for PV and wind energy installations will be defined in maps. However, advice on specific locations for possible installations are not included. Socio-economic effects on industries and workforce in the region also fall outside the scope of this analysis.

1.3 Methodological Approach and Structure of Assessment

Subsequent to Chapter 1, which presents the North African power industry, in particular the renewable energy sector, Chapter 2 will introduce photovoltaic and wind power technologies accompanied by a brief historic depiction and imminent technological tendencies.

Chapter 3 and 4 provide a detailed explanation of the Cost-Supply assessment wind power and PV energy, respectively. The assessment methodology follows the structural approach adopted by Hoogwijk (2004).

illustration not visible in this excerpt

Figure 2: General methodology in the Cost-Supply assessment

A spectrum of geographic, technical and economic constraints condense the theoretically available potential of wind and irradiance in North Africa to produce country-specific information on magnitude and cost of the feasible wind and PV electricity output. At the beginning of both, Chapter 4 and 5, an adapted version of the methodological structure, illustrated in Figure 2, will sketch the steps for the upcoming assessments. Results from these steps are directly compared to outcomes of similar studies and put into context with the current and projected electricity demand.

The inclusion of transmission infrastructure in Chapter 6 refines these results. Chapter 7 discusses the implications when employing photovoltaic and wind energy for electricity supply, as well as the options to overcome the inherent challenges. Finally Chapter 8 concludes by summarising the results of the findings and providing an insight into the future of renewable energies in North Africa reassessing seminal ideas of this paper.

1.3.1 Information on Data Processing and Area Correction Factors

The information required for theoretical and geographical potential assessment is gathered and blended in a Geographic Information System (GIS), in this case ArcGIS Version 9.3 and 10 by the Environmental Systems Research Institute (ESRI). GIS allows spatial analysis and manipulation of information enabling the creation of maps with high information density [12].

Datasets are provided either vector based, as raster files - in a multitude of formats - or as individual data points. Cartographic projections define the method of projecting the geographic data, given as a three-dimensional spheroid, to a two-dimensional plane. Since the shape of the earth is complex and the deviation between data point position and actual position on earth is greater with more general cartographic projections, commonly only global datasets employ the “World Geodetic System 1984” (WGS84). Country-specific datasets are given in country-specific projections. When no projection is specified, as e.g. with scanned maps, the raster, which in this case is a picture, can be georeferenced by linking characteristic features, e.g. administrative boundaries, on the sourced map with the same characteristic features in a correctly projected target map. For data assembly and processing, all datasets are converted to “WGS84”. While the performance of genuine computers does not support the geographic processing of the joint, high resolution data for the entire region, the same was conducted country by country.

Table 1: Area of countries and spatial deviation of “ WGS84 ” and “ Albers Equal Area Conic ” from actual country size [13]

illustration not visible in this excerpt

To calculate area sizes and distances, the three-dimensional spheroid is projected again to a two-dimensional plane. All cartographic projections are related to a specific projection origin. With growing distance to this origin distortions grow as well. While “WGS84” exhibits largest distortions, other projections are also not fully true to length and area size [14]. Several cartographic projection methods are evaluated regarding lowest possible distortion when projecting the data of the North African region to a two-dimensional plane. Actual country sizes provided by the Federal Statistics Bureau of Germany [15] and the CIA[16] are utilised as reference.

Of the compared cartographic projections “WGS84”, “Lambert”, “Africa Sinusoidal” and “Albers Equal Area Conic” the latter exhibits smallest areal deviation from reference country sizes. Table 1 depicts this deviation exemplarily for “WGS84” and “Albers Equal Area Conic”.

Consequently, prior to data export and further assessment external to the GIS software, the assembled and blended data is first converted to “Albers Equal Area Conic” and subsequently adjusted by country- and projection-specific correction factors in order to match country sizes in GIS with actual country size figures as provided by DeStatis (2009).

For geographic presentation, outcomes are imported to GIS and subsequently merged.

2 The Power Sector in North Africa

Economic growth, progressing urbanisation and demographic changes are causing sharp increases in electricity demand all over North Africa. Generation capacities are adapted in different ways but with a consensus for the integration and harmonisation of national electricity markets into the European electricity market as well as the need for more competitive market rules. State owned suppliers like O.N.E. in Morocco, Sonelgaz in Algeria and STEG in Tunisia have surrendered monopolies giving independent power producers (IPP) a chance to enter the market [17].

The picture of the North African countries in regard to fossil resources is ambiguous. It ranges from countries with an abundance of fossil resources - Algeria as world’s fifth biggest exporter of gas in 2009 [18] and Libya with world’s ninth biggest crude oil production capacity in 2010 [19], to countries with no fossil resources at all, e.g. Morocco, where the development of a decisive share in power generation from RES is one of today’s fundamental energy security interests [20]. Table 2 depicts that currently Morocco is the only net importer of electricity in North Africa besides its general importing position on the fossil resource market. Egypt and Morocco are clearly leading the path towards the development and integration of renewable energies generation. Concerning CO2 emission, Libya has the highest per capita pollution thereby exceeding the world average, though still below the OECD average.

Table 2: Electricity consumption and emissions in North Africa, OECD and the world

illustration not visible in this excerpt

Source: Net Electricity [21], all others [22], all values for 2009

Based on population growth Trieb et. al (2009) forecast a steeply increasing electricity demand in all North African countries with a total demand of 646.9 TWh/a predicted for 2030 and 1225.2 TWh/a for 2050 - 600 % of the value of 2010. Figure 3 illustrates this development including forecasted population figures by various sources. Compared to the forecast North African demand values by the IEA (2030: 700 TWh/a; 2050: 1500 TWh/a) [23], the more conservative estimates by Trieb et al. (2009) will be utilised in the study in hand.

The European Union is confident of accomplishing 20 % of its total demanded energy of 4000 TWh/a to be supplied by RES by 2020. For 2030 the EU plans its demand of 4560 TWh/a to be predominantly generated by RES (41 %) leaving 30 % to fossil and 29 % to nuclear sources [24]. According to Scholz (2010), the electricity demand of North Africa and the European continent will add up to 9,560 TWh/a in 2050.

illustration not visible in this excerpt

Figure 3: Electricity demand and population in North Africa until 2050

Sources are found next to the absolute values given in Table (Anex.) 1 and Table (Anex.) 2.

Quoting data for 2005 provided by the “Observatoire Méditerranéen de l’Energie”, Hilgers (2010) states 22 to 73 % of the operating plants in North Africa to be aged more than 20 years. The country-specific evaluation is provided in Figure 4. In addition to the required power plants to meet the increasing demand, the need for repowering is especially urgent in Libya.

illustration not visible in this excerpt

Figure 4: Percentage of power plants older and younger than 20 years as of 2005 [25]

2.1 Renewable Energy in North Africa

Along with other MENA countries Morocco (and Western Sahara), Algeria, Tunisia, Libya and Egypt are founding members of the “Regional Center for Renewable Energy and Energy Efficiency” (RCREEE) a think tank promoting favourable policies, encouraging cooperation and fostering R&D in the field of renewable energy and energy efficiency. Up until 2012 the RCREEE will be supported with financial and technical assistance from the EU, Egypt, Denmark and Germany [26].

All of the observed countries have plans to increase the share of energy generation from RES within their power mix within the next decade; the rollout of wind energy and PV is parts of these plans. Hydro power is already utilized in most of the North African countries, with decisive shares in Morocco and Egypt, as given in Table 3. However, in terms of actually existing and projected wind parks, the situation in each country is highly diverse. For detailed information on existing and projected wind parks see Table (Anex.) 3. According to Table 3 the development of PV power in Northern Africa is currently in all North African countries well behind the size and quantity of wind power projects resulting in few utility scale PV parks and basically off-grid, rural applications. The North African share in global energy generation by RES is yet marginal. This section provides a country specific analysis of the renewable energy sector.

Table 3: Energy generation by RES in North Africa, OECD and the World

illustration not visible in this excerpt

Source: [27] all values for 2009

A commonly practiced strategy in promoting renewable energy deployment is the establishment of feed-in tariffs, which guarantee the RES power producer a price, at which the electricity is bought. Long term tariff structures, coupled with low interest loans and tax incentives, enable proprietors to establish sound financing structures, reaching economic viability of the project. With a feed-in premium, the RES power generator receives revenue additional to the market price [28]. Egyptian feed-in tariff schemes are scheduled for introduction in 2012 [29]. Since 2004, Algerian feed-in premiums top up the Algerian retail price by 200 % for PV and wind energy fed into the grid [30]. No other incentivising feed-in pricing schemes are known to exist or be planned within the North Africa region.

2.1.1 Morocco & Western Sahara

The Moroccan government currently administers most of the West Saharan territory. The efforts being made in regard to renewable energy deployment in Western Sahara are also driven by the state of Morocco. The two countries will therefore be analysed conjointly.

In 2008 Morocco was generating ca. 21.268 TWh by an installed capacity of 5292 MW and importing additionally 4,261TWh to satisfy a consumption of 21.711 TWh [31] - an overrun of 3818 TWh. Electricity consumption grew at an average rate of 7.5 % between 2002 and 2008 [32] reaching 23.90 TWh in 2009 [33].

The objective of the Moroccan government is to reach a 12 % share of energy supplied from renewable sources by 2012 [34]. In terms of installed capacity Morocco had a 26 % share of RES in 2008, with a major contribution by hydroelectric (22 %) and a small contribution by wind power (2 %). By 2020 the share of renewables is targeted at 42 %, with solar and wind energy meant to contribute 14 % each to the overall installed Moroccan capacity of 14,580 MW [35]. Though not particularly observed in this study Morocco is also notably active in the field of CSP with 160 MW installed in 2009 [36] and 2 GW targeted for by 2020 [37].

Morocco’s EnergiPro plan is further promoting the development of renewable energy installations by assuring power transmission and power purchase for renewable energies.[38]

Morrocco will mobilise 1.5-2 billion USD for its “Fond de Developpement de l’Energie” promoting low-carbon technology deployment. This institution is further supported by the Clean Technology Funds with 150 million USD [39].

The Moroccan electricity market has been liberalised since 1994. Especially for power plants up to a capacity of 50 MW the access to the electricity market was facilitated in 2008 with plans to split the generation market into a regulated and an open market segment [40].

Wind Energy

Having several medium to large sized wind parks, Morocco is not at all new to wind energy deployment. The government of the Kingdom of Morocco is conducting the plan “L’initiative 1000 MW”. Within this initiative Morocco is promoting a diversification in energy supply and the use of national energy resources, notably renewable energies with a focus on wind energy. Fourteen wind parks - existing, planned or under construction - scattered over the Moroccan and Western Saharan territory will amount to an installed capacity of over 1000 MW by 2012 [41] with 2 GW targeted for 2016 [42].

Photovoltaic Energy

Several rural electrification projects accomplished the supply of PV energy for 40,000 rural households with a further 110,000 households targeted, which are not available for grid connected electricity supply [43]. Grid connected utility scale PV applications commenced in 2007 with 50 kWp being inaugurated at Tit Mellil followed up by 150 kWp at the Airport Med V [44]. By 2012, Morocco plans to integrate 10 MW of “decentralised applications”, thus rooftop PV systems, into the low voltage grid. Until 2020 the installed capacity of these “decentralised systems” is meant to reach 80 MW [45].

2.1.2 Algeria

Algeria still has a very low share of renewable energy sources, basically hydroelectricity (see Table 2), but is entering the renewable energy scene parallel with CSP, geothermal, biomass, wind and PV technologies. In 2007 Algeria planned to reach a six percent share of renewable energy generation by 2015. These goals were enforced by laws for promotion and bonuses depending on the amount of cogeneration in hybrid CSP-fossil plants.[46] Nonetheless in 2008 targets for electricity generation by RES were adapted to five percent until 2017 with a long term target of 20 % by 2030, of which 70 % should be provided by CSP, 20 % by wind and 10 % by PV power. Since its liberalisation in 2002, several independent electricity producers have been emerging on the Algerian electricity market [47].

Wind Energy

In 2007 the installed capacity of wind turbines stood at a negligent value of 73 kW [48]. Algeria’s first industrial scale wind park (10 MW) will be situated in southern Algeria, beginning operation in 2012 [49]. Furthermore Algeria is planning on substituting and hybridizing Diesel power systems in the Grand Sud region with wind turbines [50].

Photovoltaic Energy

2352 kW of PV systems were installed until 2007, mainly being used for electrification, telecommunication and pumping [51]. Sonelgaz, the Algerian electricity provider, is distributing a total capacity of 453 kW from stand-alone PV systems for 1000 households in 20 off-grid villages in the Grand Sud area [52].

2.1.3 Tunisia

With few fossil resources on hand, Tunisia’s increasing electricity demand is still predominantly generated by fossil fuels. The growing energy deficit [53] shall be encountered with the development of RES exploitation encompassing CSP and PV capacity of 120 MW combined and wind capacities of 330 MW until 2016. The long term goal for renewable energy installed capacity is set at 1200 MW in 2020 and 1800 MW by 2030. Since 1996 independent power producers can supply the liberalised Tunisian electricity market [54]. Until 2016 Tunisia aims to reduce fossil fuel consumption by 22 % of the fuel consumption level of 2009. The revenue associated to the Carbon Emission Reduction (CER) is estimated with 13 million USD per year [55].

Wind Energy

Tunisia gained experience in the development and operation of wind parks. By extending its first wind park at Sidi Daoud to almost 54 MW and with scheduled commissioning of an additional 120 MW, Tunisia planned to reach an installed wind power capacity of 174 MW by the end of 2009. Up until 2011 the overall installed grid-connected WT capacity was limited to 200 MW due to grid security constraints [56].

Photovoltaic Energy

Since 1980 generally 100W-PV-Kits have been deployed to electrify rural and isolated settlements and border control posts in Tunisia. Other applications are pumping and telecommunication totalling an estimated installed power of 2 MW in 2009 [57]. The Tunisian public electric energy provider STEG is further targeting to equip 1000 solar buildings with 1.5 MW in PV systems [58]. In 2001 France and Tunisia signed a declaration of intention to set up a concentrated photovoltaic system of the Concentrix type, owned by the French semiconductor company Soitec, and to cooperate in the development of utility scale electrical energy storage systems, notably lithium-ion batteries [59].

2.1.4 Libya

With its vast reserves in oil and natural gas Libya will remain a major player on the future fossil resources market. The Libyan economy is heavily dependent on these commodities with oil revenues generating 70 % of the GDP in 2009. Electricity generation capacities are meant to increase from 5.5 GW in 2007 to 10 GW in 2015 with tendencies to substitute oil by increasing the current ten percent share of primary energy stemming from gas. Energy prices are heavily subsidised, giving no economic incentives for renewable energy deployment. Consequently, the high RES potential - in particular solar - remains untapped. Nonetheless the Renewable Energy Authority of Libya (REAOL) envisioned elevating the zero percent share of generation from RES in 2008 (see Table 2) to 25 % in 2025 and 30 % in 2030 with intermediate targets of 6 % in 2015 and 10 % in 2020 [60]. The rollout of WT and PV capacities is planned along with CSP capacities of 100 MW. Nonetheless these targets were not commonly agreed on in 2010 [61] and might have been changed due to the new administration in 2011. Another source is quoting a national plan with RES contributing ten percent to the electricity demand by 2020 [62].

Wind Energy

REAOL’s plans for 2008 to 2012 included projects of 850 MW in wind parks, thereof 360 MW in the north east, 250 MW in the west, 120 MW in the south east and 120 MW in the south western region [63]. These plans might gain new momentum now.

From 2000 on, a Danish-German consortium of consulting engineers and research institutions developed a wind atlas, and since 2004 prepared feasibility studies on potential sites. However, the 25 MW pilot wind farm at Dernah in north-east Libya, with a first stage upgrade to 60 MW, was never realised. According to Martina Dabo, Wind Assesment Group Leader at C.U.B.E. Engineering, wind turbine equipment had been delivered to a Libyan port coinciding with the bombing of the same during the 2011 Libyan civil war [64].

Photovoltaic Energy

Through governmental programs, by 2005 Libya reached an installed PV capacity of 875 kW. utilised mainly in remote communication networks, rural electrification and pumping [65]. REAOL’s plans furthermore envisioned PV energy to be contributing 5-10 MW of grid connected capacity along with 2 MW off-grid, stand-alone PV in the electrification of remote areas [66]. By the 1000-roofs-programm a total capacity of 2 MW in PV systems was targeted to be installed on rooftops. Assembly capacities for 50 MW of PV systems where meant to be established in Libya [67]. However, no information on the degree of realisation of these plans could be found.

2.1.5 Egypt

In 2008, 12.6 % of Egypt’s consumed electricity was generated by RES, of which 11.8 % were contributed by hydro and the remaining 0.8 % by wind power (please refer to Table 2 for further information). Egypt’s energy strategy aims at boosting this share to 20 % of electricity generation by RES in 2020. The lion’s share of 12% will be contributed by 7200 MW of grid connected wind parks, followed by hydropower and solar energy. These ambitious targets are accompanied by sound policies. International tenders are promoting the BOO scheme (build, own, operate) [68] while the upcoming feed-in-tariffs [69] are meant to further encourage the private sector. A future share of 33 % in state-owned and 67% in privately owned wind parks is targeted [70].

Wind Energy

After a steady growth in wind park sizes with the Hurghada and Zafarana projects (542 MW), Egypt’s New and Renewable Energy Authority (NREA) is implementing several large-scale projects on a 150 km stretch along the desert plains west of the Gulf of Suez, reaching from Zafarana to the Gulf of El-Zayt. The first (540 MW) and second stage (580 MW) are scheduled for commissioning between 2013 and 2014, with many more projects currently in the pipeline (please refer to Table (Anex.) 3 and Figure 5 for more information). The deployment of wind energy is supported by the Clean Technology Fund with 1056 million USD [71].

illustration not visible in this excerpt

Figure 5: Past and projected growth of installed wind power in Egypt [72] [73]

To capitalise on the new capacity in the Giga Watt range Egypt is seeking to develop local manufacturing of WT blades and towers with yearly outputs equivalent to 400 MW of new WT capacity. Consequently, in 2008 the largest Arab producer of electrical equipment and cables, El Sewedy, entered the wind power market. Already in 2010 capacities of 240 MW in WT assembly, 300 MW in tower production and 300 MW in manufacturing of rotor sets were established [74]. Today, 30 % of wind power equipment is manufactured domestically. By 2020 this share is planned to be raised to 70 % [75].

New tariff schemes by the Egyptian government exhibit large tariff reductions for all wind turbine components by paying a reduced tariff of 2-10 % on imports. Wind turbine towers are excluded and pay 30 % import tariffs. While incentivising wind energy tariff reductions are only for WT “specific” components, they also appear to protect the emerging “non-specific” WT tower manufacturing industry in Egypt [76].

Photovoltaic Energy

In 2011 the installed PV capacity in Egypt amounted to 5.2 MW of decentralised, grid- independent operating devices. A 43 kW utility project for rural electrification was completed in 2010. The five year plan “2012-2017” includes the construction of several PV parks with a total capacity of 20 MW [77]. According to most recent information of the New and Renewable Energy Authority, two grid-connected PV parks with 20 MW capacity each are currently under preparation [78].

3 Introduction into Wind and PV Energy Technologies

3.1 Wind Energy

Wind turbines, or more accurately wind energy converters, turn kinetic energy contained in the movement of air masses into a usable form of energy. The basic layout is a rotor with a varying number of blades transmitting the captured energy towards an energy converter, e.g. a milling machine [79]. Harnessing wind energy for stationary industrial machines had its first golden era in the 18th and 19th century, with Europe having an estimated 200,000 windmills reaching the highest density in the Netherlands where more than 9000 windmills deployed in diverse industrial processes enabled an industrial boom leading to e.g. a monopoly in the export of sawed wood. It was not steam engines but the rural electrification of the 20th century which made traditional wind mills an obsolete technology with comparatively high maintenance costs. The reduction in operation and maintenance effort for rural water pumping stations was the goal of the subsequent invention of Halladay’s wind turbine with automated pitch and azimuth adjustment. Wheeler’s additional reduction in complexity promoted this technology to become the standard for rural areas of North America and Australia with 6 million produced turbines by 1930 [80] [81]. The leap to electricity generation was achieved in 1881 with the researcher Pour La Cour in Denmark, whose wind turbines reached an output of 10-35 kW at high reliability rates. He already envisioned an energy storage system through the transformation to hydrogen via electrolysis. With his publications in 1920 and 1925 the physicist Albert Betz laid the scientific foundations of today’s wind turbine technology [82]. According to his Elementary Momentum Theory, the mechanical power output of the converter is calculated by Equation 1.

illustration not visible in this excerpt

Equation 1: Mechanical power of the wind converter [83]

illustration not visible in this excerpt

Equation 2: Power of undisturbed wind at rotor area [84]

Divided by the power of the wind at the same rotor area without any power drain (Equation 2), Betz derived the optimal power coefficient cp or Betz value to be 0.593, achievable only through rotors utilising ascending force. Modern rotors are on the verge of reaching a power coefficient of 0.5 [85].

illustration not visible in this excerpt

Equation 3: The power coefficient [86]

Technologies were refined over the following decades with a renaissance in the late 1970s [87]. By the end of the 90s grid-connected turbine costs had fallen below 800 € per kWp. Additional feed-in tariffs supported wind energy in reaching economic viability [88].

3.1.1 Introduction into Wind Energy Systems Technology

According to Hau (2008) wind turbines are classified by their aerodynamic designs leading to three groups: horizontal rotor axis, vertical rotor axis and concentrators. Nevertheless the subgroup of lee running horizontal axis wind turbines is by far the most widely used concept in the electrical energy generation through wind power [89]. The subsequent technical explanation therefore refers to this subgroup focusing on the turbine sizes and characteristics relevant for the further analysis.

illustration not visible in this excerpt

Figure 6: Components of a horizontal-axis wind turbine [90]

Figure 6 depicts the components of a wind turbine. The blades comprising the rotor are connected to the rotating shaft either fixed or flexible. The rotating shaft, an intermediate and in some models not obligatory gearbox and the generator form the drive train which is housed inside the nacelle mounted on top of the tower. The drive train’s challenge is to transform the low frequency mechanical power of 10-20 rpm[91], or 0.16-0.33 Hz, to an electric power output meeting the grid frequency of 50 Hz or 60 Hz, depending on country. In the megawatt range the horizontal or azimuth angle of the nacelle is adjusted actively with the azimuth drive, also called yaw system. The position of the nacelle and rotor thereby can be horizontally adjusted following the wind flow [92].

Being the element to absorb the air mass’ energy, consequently the aerodynamic characteristics of the rotor and the area size swept by it is of essential value. The ascending force sets the blade into motion and thereby the rotor into its rotary movement. Depicted by the strip theory the ascending force on the blade is the result of both wind speed and pitch angle. With the number of blades the rotor power coefficient increases. Rotors composed of three blades have a four percent higher coefficient than two-bladed rotors, also reaching a higher start-up momentum. As the addition of a fourth blade provides just 1-2 percent improvement, the 3-bladed design remains the most common [93] [94]. The composite material used - epoxy resin with carbon fibres and wood, enforced by a steel structure in exceptionally large blades [95] - has an additional safety advantage as the material first tends to decompose into fibres instead of failing abruptly under critical loads. Sudden ruptures are more of a concern for metallic muffs and structures connecting the blade with the rotor shaft [96].

A passive stall with a fixed blade-shaft connection is only used in small to medium wind turbine systems. With systems in the megawatt range the blade is connected to the shaft via a hub, containing bearings and the blade pitch control, to the shaft. This enables the axial turning of the blades over their entire length during operation. Thereby the aerodynamic performance and the stall angle to limit the received power and protect the turbine are actively controlled. The adjustment of only sections of the blades is no longer common practice, as mechanical loads at the tips of the rotating blades are too high with rotor sizes in the megawatt range [97].

Contemporary drive train concepts are classified into fixed-speed and variable-speed concepts. In the first, also known as the Danish concept, the stall-regulated wind turbine rotor drives via a multistage gearbox the Squirrel Cage Induction Generator (SCIG). The SCIG has to operate in a narrow speed band around the synchronous speed with a maximum of 1 to 2 % torque variation [98]. Its pros of being a robust and relatively cheap concept are outweighed by its cons: A lower performance due to a relatively high slip and a direct translation of speed fluctuations to electro-mechanic torque, causing high mechanical and fatigue stresses on the WT system. As a consequence to theses stresses, a costly and heavy three-stage gearbox is required [99]. Furthermore the SCIG draws excitation current from the grid [100].

As a result, today’s market is governed by variable-speed concepts. This group is further divided into the partial-scale power converter concept, known as DFIG (Doubly Fed Induction Generator) depicted in Figure 7, the direct-drive concept with a full-scale power converter, known as Electrically Excited Synchronous Generator (EESG) depicted in Figure 8 and the emerging Permanent Magnet Synchronous Generator (PMSG) concept [101].

illustration not visible in this excerpt

Figure 7: Scheme of a variable speed with partial power conversion (DFIG) [102]

With the DFIG concept the whole drive train is on a horizontal axis inside the nacelle. The WT rotor is coupled through a multi-stage gearbox to the generator’s rotor. Only the generator’s stator is directly connected to the grid. The generator’s rotor is connected to a power converter transferring 25-30 % of the generator’s total capacity and enabling a variation in speed of ± 30 % around synchronous speed. The partial-scale power conversion enables a smooth grid connection with an improved efficiency compared to SCIG. Wear is reduced due to short transmission distances resulting in an easier control of dynamic forces. Further advantages are a good balance with the rotor weight, the absorption of the high rotor torque at the gearbox, and the opportunity to use standard gearbox and generator components [103]. Disadvantages are its sensitivity to wind gusts which lead to misalignments of the drive train, and deterioration of gearbox components. Consequently the DFIG concept requires a heavy gearbox and therefore a more rigid tower design [104]. Furthermore the inherent generator slip ring causes electrical losses and requires regular maintenance. This concept is used by most manufacturers: Vestas, Gamesa, Repower and Nordex [105].

The EESG concept with full-scale power conversion is typically applied by manufacturer Enercon. An annular multi-pole generator is directly connected to the turbine’s rotor shaft omitting the gearbox. The following converter translates all power output to grid voltage and frequency. Thereby the generator rotation does not have to be fixed at grid frequency, allowing for a fully controllable generator speed even at very low rotor speeds. Further advantages are less wear due to fewer moving parts and less stress due to a high level of speed variability and a smooth grid connection over the entire speed range.[106] Generally the generator’s rotor torque increases with the turbine’s rotor size. Slow machine rotation enables a better management of the generator’s torque, thus enhancing the system’s suitability for very large turbines. As a result, the ESSG concept lately attracts more manufacturers (e.g Siemens [107], GE, Alstom [108] ) to enter the direct drive sector especially for the upper end turbine sizes used off-shore, where reliability and low maintenance rates are critical for turbine economics [109] [110]. The major drawback is the costly and heavy generator, which has to be excited with slip rings causing inevitable field losses.

illustration not visible in this excerpt

Figure 8: Scheme of a variable speed with full power conversion (EESG) [111]

Having its origins in compact ship engines the direct drive Permanent Magnet Synchronous Generator (PMSG) concept is a relatively new drive-train concept. Removing the need for excitation power - as compared to EESG and DFIG - enhances the efficiency and energy yield. Wear is reduced as slip rings are obsolete, thereby improving the reliability. Yet higher cost for materials, a performance challenge for the permanent magnet - which demagnetises at high temperatures - as well as control difficulties offset the better power-to-weight ratio [112].

Li and Chen (2007) surveyed various studies comparing system costs and annual energy yields of the most used concepts. Direct-drive systems (EESG) are superior in energy yield, still, from the energy yield over cost perspective, partial-converter systems (DFIG) with multi- stage gearboxes are performing better. Globally variable-speed concepts with partial or full power conversion clearly dominate the market, with fixed-speed concepts being phased out by 2013, a situation which ten years ago was quite the opposite. This domination is consistent to an increasing demand for smooth grid connection and manifests itself in a market share of 63.8 % for DFIG and 20 % for EESG systems in 2009, both with strong growth tendencies [113].

With extensive wind energy penetration in a power network, requirements for wind turbines regarding grid code, e.g. power quality and assistance in grid stability, are getting more severe as well. Modern wind turbines cope with the most stringent grid code requirements, enabling active power output control of the energy production unit [114].

The analysis for North Africa will be discussed along with two reference turbines: a DFIG turbine by Gamesa and an EESG turbine by Enercon.

In regard to energy generated over required space for plant equipment, a 500 kW wind turbine (11.7 MWh per m² and year) is comparable to a 750 MW black-coal fired thermal power block with 15-20MWh per m² and year [115]. With rapid technological advancement this power output to required space ratio might be even favourable for wind energy compared to a coal-fuelled energy, especially when incorporating required space for resource extraction into this equation.

3.1.2 Expected Challenges by the North African Environment

Considering the environmental conditions of North Africa with high dust loads, sandstorms and hot, dry climates, a drive train with gearbox holds additional risks in terms of reliability. To minimise the ingress of contaminants during operation the journal “Machinery Lubrications” suggests a completely sealed gearbox with a flexible expansion chamber to allow gearbox ventilation as well as V-rings as external seals instead of standard labyrinth seals [116].

Furthermore, winds carrying a high density of abrasive particles, like sand grains, reduce the aerodynamic performance of the blades by causing progressively growing fine cracks in the surface. Particle impact velocities are highest at the outer tip edges of the blade. Normally, the prevention/reduction of such abrasion damage is achieved through the use of extremely tough materials. This is not so simple in WT systems which require flexibility for the absorption of kinetic loads. Current preventive methods include coating with aircraft-quality polyurethane paint and improving the ductility of the surfaces at the leading edges of an airfoil by applying ductile elastomeric tape. The tape absorbs the impact energy of particles but has to be replaced frequently in order to maintain high performance. Present developments include nanocomposite layers at leading edges, which absorb the impact energy over a larger area, and a new blade material consisting of thermoplastic composite layers [117].

3.2 Photovoltaic Energy

The first functional PV module dates back to 1883, however modern PV development commenced within the US Bell Labs in 1954 with the discovery of pn-junctions in diodes generating electricity through light. An efficiency of 6% was already reached within a year. By 1958 PV-cells were in use for backup powering of satellites with many following developments in PV technology originating in power supply for satellites. The first oil crisis in 1973 triggered pressure on developing alternatives to fossil fuelled power generation, which manifested itself in rising governmental funding for PV research. During industrialisation of PV in the 1980s trusts like Siemens and traditional oil companies, e.g. Shell and BP, added PV technology into their portfolio. Interest also grew in improving all components used in PV systems with a particular focus on DC/AC converters, leading to cost improvements for the whole system [118].

3.2.1 Introduction into Photovoltaic Systems Technology

Photons, either in diffuse or direct light, excite weakly bonded electrons within the valence band of semiconductors by passing an amount of energy, measured in electron volts (eV), to them. When the photon’s energy exceeds a certain threshold of energy - the band gap, which is typically at 1.1eV with silicon - bonds between electron and semiconductor break up, allowing the electron to move within the conduction band. Any energy absorbed, which does not surpass the band gap threshold is converted to heat. The photon’s energy is characterised by Equation 4.

illustration not visible in this excerpt

Equation 4: Energy of a photon[119]

Contacts collect the free moving electrons, transferring them to an outer circuit where the electric current - containing electrons of a slightly lower potential than the initial band gap they surpassed - shed energy by doing work. Afterwards electrons are restored to the semiconductor.

illustration not visible in this excerpt

Figure 9: Functionality of a solar cell [120]

The width of the band gap is adjusted by doping of the pn-junction. P is the positive side doped with e.g. boron, with fewer electrons than the surrounding silicon and consequently accepting one electron from the conduction band. N is the negative side doped with e.g. phosphorous, with more electrons than the surrounding silicon and consequently donating one electron to the conduction band [121]. The functionality of a one-junction solar-cell is illustrated in Figure 9. In order to exploit more of the photon’s energy, which - depending on wavelength - carries between 0 and 4 eV, multi-junction solar-cells are deployed with several pn-junctions containing best matching band gaps arranged in a transparent, stacked configuration to boost efficiency. Nevertheless there are physical limits, with 1, 2, 3 and 36 pn-junctions having maximum efficiencies of 37, 50, 56 and 72 % respectively [122].

Large numbers of solar cells are interconnected and encapsulated to form PV modules producing DC current. This, once transformed and harmonised via an inverter to AC current, can be fed into the electricity grid. To optimise the cell’s orientation in relation to solar irradiation, PV modules are mounted on a rig which can be either fixed or movable in order to track the sun’s movement over day/year. Aside from the modules, all additional PV system components, including energy storage system e.g. batteries, are referred to as the balance of system (BOS) [123].

Figure 10: Interconnected solar-cells building up PV modules which connect to inverters form PV systems [124]

An advantage in PV systems is their scalability, with system sizes ranging from just one module (50-250 Wp) to multi megawatt systems such as the world’s biggest PV installation to date: 97 MW installed capacity by 1.3 million thin-film PV modules, ground-mounted in Ontario, Canada [125].

Differing PV Technologies

In 2008 crystalline silicon technologies - mono and multi - encompassed 85-90 % of the PV market with cell efficiencies of around 17.5 %, followed by 10-15 % market share for thin-film technologies with cell efficiencies of around 11 %. Organic solar cells and concentrating photovoltaic (CPV) technologies were rather niche markets, each counting for less than 1% market share [126]. 18.23 GW of new modules were installed in 2010, with 21 GW expected for 2011 [127]. By now most manufacturers are incorporating thin-film into their portfolio, with some solely building on thin-film technology. As a result an accelerated expansion of thin-film market share is expected. GBI Research estimated 20 % thin-film share by 2011 and a 2020 market with c-Si at 50%, thin-film at 40% and emerging technologies e.g. CPV at 10 % [128].

The dominance of crystalline silicon (c-Si) technologies is rooted in the abundance of silicon as base element and the application of knowledge developed by the micro-electronics industry with initial use of scrap material from the same. Both industries require silicon with highest purity, since impurities in c-Si cells would lead to recombination of valence electrons in the conduction band, thereby generating heat instead of electricity [129].

Within the manufacturing process of c-Si cells the p-side of the pn-junction is produced by enriching the molten silicon at 1400°C with boron before it solidifies as a slowly growing perfect mono-crystal, which afterwards is cut into wafers. Phosphor atoms diffuse into the wafer’s surface creating the n-type layer. These mono-crystalline cells have highest efficiencies at highest costs. Multi-crystalline cells with one wafer containing hundreds of mono-crystalline grains are of lower efficiency, which is offset by its lower cost due to simpler wafer-growing equipment and processing. However, to absorb and process a given fraction of sunlight the mechanical, physical and optical characteristics of silicon require a Si-layer ten times thicker than thin-film cells. Additionally thin film technologies, using semiconductors like CdTe, Cu(InGa)Se2 and amorphous Silicon (a-Si), have a higher impurity tolerance than c-Si and can be deposited on a wide range of substrates (e.g. glass and plastic) in subsequent processes at much lower process temperatures between 200°C and 500°C. Multi-layers varying in thickness have different purposes such as reducing resistance and reflection losses, contacting and interconnection, besides, of course, forming one or more pn-junctions. Therefore the share of thin-film technologies is rapidly growing due to its cost competitiveness, despite a lower efficiency than c-Si-cells [130].

The application of PV technologies in the field will be further described in Chapter 5, with Section 5.3.1 putting a focus on technological challenges that are to be expected with the North African environment.

Concentrating PV (CPV)

Solar cells are the most cost intensive part in PV modules and complete PV systems. The idea of concentrating solar irradiation on more efficient solar cells - in some cases beyond 40% - but with a higher price tag, is not far-fetched. At concentration factors of 200 to 300 for concentrators using Si-cells, and 1000 to 2000 for those using GaAs-cells, the cell efficiency is of higher importance than its cost. Compared to other PV technologies CPV can only harness direct solar irradiation, which however is predominant in North Africa [131]. The CPV module’s orientation is adjusted on two axes to optimally track the sun’s direct irradiation.

Compared to fixed PV modules the duration of high power output each day is effectively extended.

The technology has undergone much development but still remains at a niche technology. However, CPV is accounted with a high potential for future utility scale grid connected power supply with 30 % AC-system efficiency attainable in the medium term. Installed CPV capacity is forecasted to increase from 10 MW in 2010 to 1500 MW in 2015 [132] [133] [134].

Current CPV modules by manufacturer “Soitec” are operating with a concentration of 500, accomplished through Fresnel lenses. Solar energy is converted with triple junction GaInP/GaInAs/Ge cells, thereby reaching 26 % AC-system efficiency [135]. Manufacturer “Amonix” is equipping the largest CPV park to date, with a capacity of 30 MW using similar technology. It is due to go online by the second quarter of 2012 [136]. Amonix has simultaneously increased its manufacturing capacity from 100 MW in 2010 to 240 MW by the end of 2011 [137]. Other technological possibilities include light-guiding solar optics, which capture direct solar irradiation guiding it by internal reflection to a high efficiency solar cell which is directly bonded to the centre of the optic. Concentration factors of 1000 are attainable in a module of reduced thickness and without the risk of optics-cell misalignment [138].

4 Cost-Supply Assessment of Wind Energy in North Africa

This chapter discusses the potential, as well as the restrictions encountered by WT applications in North Africa.

Figure 11 illustrates the general methodological outlay depicted in Section 1.3, adapted to the requirements of the wind energy assessment.

illustration not visible in this excerpt

Figure 11: Scheme of LCOE (Levelised Cost of Electricity) calculation methodology adapted to wind energy

While wind resource data quality is essential to obtain reliable Cost-Supply results, data sources are evaluated in deep prior to the geographic assessment. Subsequently a detailed depiction of the geographic restriction criteria as well as technical characteristics provides information on the analytic steps. Results are accrued in country-specific Cost-Supply- Curves and maps depicting the geographic distribution of the Levelised Cost of Electricity (LCOE) of wind energy. Finally results will be set into the context of current and future energy demand.

4.1 Theoretical Potential of Wind Energy

The earth's surface is heated by varying levels of solar irradiation across its surface, creating areas of high and low air density. These areas constantly move towards equilibrium creating wind. The earth’s rotation is a second, important source of energy powering constant winds, e.g. the west drift at northern latitudes and the trade wind at the tropics. Since the movement of air masses close to the earth’s surface is very much influenced by the ground characteristics, e.g. topology, reliably mapping wind conditions over extended areas is a complex matter.

This section shall reveal the source of energy driving wind turbines: magnitude, timely distribution and location of wind speeds in North Africa. The procedure for evaluating the databases and further processing will be explained as well.

4.1.1 Wind Resources in North Africa

For countries with an established wind energy industry, wind sites with highest wind speeds are already in use, as it is the case in Denmark and Germany. According to the Danish engineering consultancy BTM Consult cited in Hau (2010), the realisation of off-shore wind parks will be the main contribution to the growth in installed wind turbine capacity for such countries [139]. While the wind energy sector in the North African region is yet in a rather early stage of development, with promising on-shore sites available, only on-shore potentials and applications are considered in this thesis.

Wind atlases, which provide a raster of data points with monthly and annual averages of wind speeds, describe the overall theoretically available wind energy in a certain territory. To reduce the effort in data collection in a vast area, only a sample of points is included for data recording. The data can either be gathered through specially designed observational wind masts - at well selected locations, measuring at various heights - or by general weather stations at e.g. airports, measuring commonly at 10 m height. With this climate data on hand, numerical wind atlases are calculated through complex models.

Mesoscale models take into account topographic features and the surface roughness, thereby transforming input data from scattered measuring points into a grid of virtual stations. For reliable forecasts of the wind turbine’s energy output further improvements in the simulation of local wind phenomena are calculated using regional to local wind simulation models, e.g. WAsP by RISØ Institute. Figure 12 illustrates the methodology in developing a wind atlas.

Figure 12: The wind atlas methodology [140]

illustration not visible in this excerpt

4.1.2 Height Transformation and Roughness Length

The provided data is linked to a height above ground which is between 10 and 50 m. Roughness lengths describe the reduction of wind speed due to friction at the surface. A higher roughness length value translates to a greater reduction of wind speeds at ground level. For comparison purposes, wind speeds are recalculated to a common height of 50 m via local roughness lengths (ݖ଴ሻ and the application of Equation 5.

illustration not visible in this excerpt

Equation 5: The logarithmic wind profile and height transformation [141]

The above formula can also be used in calculation of wind speeds at wind turbine hub height.

Information on local land characteristics is derived from data on vegetation included within the VectorMap0, a vectorised version of the Digital Chart of the World [142]. The vegetation data has been collected through satellite imagery by the Advanced Very High Resolution Radiometer up until 1992 and is provided at a resolution of one kilometre [143]. Types of vegetation are further associated to roughness lengths according to Hoogwijk [144] and Hau [145]. Agricultural land is defined with a ݖ଴ value of 0.25, grassland & grazing areas with 0.03 and forest with 0.75. All other areas are considered bare land, scrubland, savannah or desert and are defined with a ݖ଴ value of 0.01.

The methodology described above is only applied when spatial resolution of wind atlases is of similar degree to the resolution in land characteristics. However, when the spatial resolution in land characteristics, which consequently defines roughness lengths, is much higher than the wind atlas resolution, distortions would occur when calculating winds to greater heights using the methodology from above. By these means, although in the same raster-tile, the wind speed would be much greater over forest areas than over grazing areas. This is contradictory, since wind speeds at high altitudes in one region with uniform topography would be levelling out to a value not greater than winds over the point in the respective region with smallest roughness lengths at ground level, thereby having lowest influence on the wind’s flow characteristics.

Figure 13 compares the influence of varying roughness lengths in the calculation of wind speeds at 10 m base height to atlas comparison height of 50 m and WT hub height of 120 m. Assuming a uniform wind speed of 8 m/s at 120 m altitude in one raster tile, the wind speed at base height of 10 m would be calculatively distributed according to roughness length from 5.9 m/s with z0 = 0.01 and 4.78 m/s with z0 = 0.25 to 4.08 m/s with z0 = 0.75 (burgundy curve). However, since just one average wind speed is known for each raster tile assuming wind speeds of 4 m/s at base height, the wind speeds at hub height of 120 m would be conversely sky rocking to 7.84 m/s over forest with z0 = 0.75 (light red curve) compared to 5.5 m/s over bare land associated with z0 = 0.01.

illustration not visible in this excerpt

Figure 13: Logarithmic wind profile with varying roughness lengths z 0 and hub heights

Consequently, for databases of much lower resolution than the resolution in land characteristics (1 km), a uniform roughness length of 0.01 - equalling land characterised by savannah, bare land and desert - is used. Section 4.2.4 will reveal this category as being predominant in the observed region.

illustration not visible in this excerpt

Equation 6: Calculation of inherent roughness length [146]

When a wind atlas provides wind speeds at several heights, inherent roughness lengths for the conversion of wind speeds to further altitudes are obtained by Equation 6.

4.1.3 Description of Wind Data

A number of wind atlases are providing data on North African wind conditions, however with varying resolutions and reliability of data. Available national and global wind atlases will be presented in the following with a subsequent empiric comparison.

Global Wind Atlases

CRU

The Climate Research Unit (CRU) at the University of East Anglia provides a global climatic atlas, CRU CL 2.0, including a wind atlas interpolated from data recorded between 1961 and 1990 at 3952 globally distributed weather stations and provided at a resolution of 0°10’, equalling approximately to tiles of 18 km edge length [147].

illustration not visible in this excerpt

Figure 14: Wind speed data at 10 m height by CRU CL 2.0 [148]

Evaluating the CRU database, Hoogwijk found that 80 % of the global area holds an annual average wind speed lower than 4 m/s at 10 m measurement height. When comparing the CRU average regional wind speeds to those of regional wind atlases, Hoogwijk (2004) found similar patterns but lower values - some even below minimal regional wind speeds. Furthermore Hoogwijk (2004) states that unit conversion errors and non-uniform measurement heights contributed to errors in data acquisition for the CRU database [149]. The density of weather stations included in the CRU data input for its wind analysis is low, particularly in North Africa, where most 5° tiles (ca. 550x550 km) encompass between zero and 5 stations [150]. According to Figure 14 showing CRU CL 2.0 data over North Africa, wind speeds are especially low in south-central Algeria. This is contradictive to the Algerian wind map[151] depicted in Figure 19, which - by processing data obtained at 75 meteorological stations situated all over Algeria - describes the same area as holding the highest wind speeds in Algeria.

Consequently the CRU database is considered inaccurate and not employed in this assessment.

MERRA

The interpolated raster dataset MERRA [152] by NASA uses data recorded in three streams since 1979 up until today. Meteorological observation from various satellites, surface stations, aircrafts, ships and buoys are included. However, no specification of the measurement point density is given for the region of interest [153]. Data depicted in Figure 15 is provided at a spatial resolution of 0°40’ - equal to raster tiles of ca. 73km edge length - and at heights of 10m and 50m above ground, thereby enabling the calculation of inherent roughness lengths applying Equation 6.

illustration not visible in this excerpt

Figure 15: Wind speed data at 50 m height by MERRA [154]

In so doing, roughness lengths vary between 0.000 and 2.829 with over 85% having a z0 = (0; 0.035] and 14 % with z0 = (0.35 and 0.275]. This is consistent with the general assumptions for roughness lengths given in Section 4.1.2.

Since the spatial resolution of MERRA is considerably different to the resolution of land characteristics, these inherent roughness lengths will be used in calculating MERRA’s wind speeds to different heights. MERRA data is furthermore available in various timely resolutions - between monthly averages and hourly averaged data - and enables the recalculation of wind speed direction. For average wind speed calculation in North Africa monthly data is used.

SWERA - NASA/SSE

The global wind atlas SWERA [155], distributed through the United Nations Environmental Programme, provides a resolution of 1° equating to raster tiles of 111 km edge length. The data for North Africa is depicted in Figure 16. It was developed by the NASA Surface Meteorology and Solar Energy through satellite data recorded from 1983 to 1993. Verification via two sets of data recorded at globally distributed airports found biases of -0.0 and -0.02 and an RMS error of 1.3 for both. The low resolution as a drawback averages out high potential areas giving a generalised view of a country’s wind energy potential. Moreover the detailed roughness lengths derived from high resolution land characteristics (1x1 km) is unsuitable with wind raster tiles of 111x111 km. A generalised roughness length of 0.01 is therefore assumed in the further processing of SWERA.

illustration not visible in this excerpt

Figure 16: Wind speed data at 50 m height by SWERA [156]

National Wind Atlases

Egypt

Wind data measured in the period of 1991 to 2005 at 30 meteorological stations is incorporated in the “Wind Atlas of Egypt” developed conjointly by the Egyptian NREA and the Danish RISØ Institute. Of these stations, 22 are specially erected wind masts at characteristic sites measuring wind speeds at heights of 27 and 47 m. The wind map, interpolated over topography and land use with RISØ Institute’s modelling software “WAsP” using the “Karlsruhe Atmospheric Mesoscale Model” (more information on www.mesoscale.dk), depicts annual mean wind speeds at a height of 50 m and a resolution of 7.5 km [157]. Taking into account the good resolution, wind speeds at required heights are adjusted according to land characteristics.

illustration not visible in this excerpt

Figure 17: Egyptian wind atlas with annual mean wind speeds at 50 m above ground [158] Tunisia

The “Atlas Éolien de Tunisie” [159] was developed by the Spanish National Centre for Renewable Energies and the Tunisian National Agency for Energy Conservation. It is published at a resolution of 0.1° equating to a grid cell sizes of approx. 15km². Wind masts were erected at 17 characteristic locations measuring winds at heights of 20 and 40 m. The Mesoscale model “SKIRON” was used for interpolation. The wind map depicts annual mean wind speeds at 10 m height. Considering the adequate resolution, the wind speeds at greater heights are adjusted, using the specific land characteristics methodology, as described above.

illustration not visible in this excerpt

Figure 18: Tunisian wind atlas with annual mean wind speeds at 10 m above ground [160] Algeria

The Algerian wind resource map [161], illustrating annual mean wind speeds at 10 m and 50 m height, was originally developed for the estimation of hydrogen production potential through wind energy deployment. Though sourced data, obtained at 75 meteorological stations scattered over entire Algeria, is described as very accurate, the available data is rather schematic and lacking in detail. As observed in Figure 19 average wind speeds between 8 and 9 m/s at 50 m height over a conjunct region bigger than Tunisia are puzzling and seem exaggerated.

Aîche-Hamane and Khellaf (2003) further show, that in regard to wind speed distribution, coastal Algeria holds generally lower potentials than southern Algeria, which is characterised by steady trade wind [162].

illustration not visible in this excerpt

Figure 19: Map of the annual mean wind speeds in Algeria at 50 m above ground [163] Libya

The wind atlas for Libya, developed since 2000 and administered by C.U.B.E. Engineering [164], could not be made available due to license uncertainties based on the current political changes in Libya.

Morocco and Western Sahara

Wind conditions for the Atlas “Les resources éoliens du Maroc” were recorded at a total of 40 meteorological stations and especially erected wind masts from 1990 to 2005, providing data for the Moroccan and Western Saharan territory [165]. Each class of 1 m/s width is transferred with its arithmetic mean value to GIS. The potential pattern of the map depicted in Figure 20 is consistent with data derived from existing wind park locations. Moroccan wind parks are situated at the Strait of Gibraltar and its hinterland, as well as the Atlantic coast of southern Morocco. At these locations wind parks experience very high yields, observable in Table (Anex.) 1. While the resolution is not specified, due to the schematic outlay of the map, a generalising roughness length of 0.01 is assumed.

illustration not visible in this excerpt

Figure 20: Moroccan wind atlas with annual mean wind speeds at 10 m in MO & EH [166]

4.1.4 Wind Data Set Comparison

Among the available data sources the Tunisian and the Egyptian wind atlases are of best observational input and processing, coupled with highest output resolution. For the further empirical comparison process these atlases are used as reference, since they are assumed to be the closest to reality. The Moroccan atlas is of schematic outlay and therefore not considered as reference but included in the comparison as well. Since there is no suitable national data source available for Algeria or Libya, the two global atlases (MERRA & SWERA: Group N) are compared against the Tunisian & Egyptian reference atlases (Group N) to determine which is most representative. Prior to the comparison, the various databases (DB) are merged to one set of polygons encompassing all data jointly. The size of a polygon is equal to the highest resolution of all databases with each polygon holding location specific wind speed data of all databases.

Comparison by Average

Because of data processing in GIS, polygons are of differing sizes requiring the weighting of wind speeds by the polygon’s area. The absolute difference in average is obtained by Equation 7.

illustration not visible in this excerpt

Equation 7: Calculation of average wind speeds weighted by polygon sizes

Given the results depicted in Table 4, averages obtained by MERRA significantly underestimate both reference atlases. SWERA, on the other hand is diverging less, with slightly lower average values than the Egyptian reference and higher values than the Tunisian reference atlas. The table furthermore shows averages for all observed countries from all analysed databases.

Table 4: Results of comparison by average at 50 m comparison height

illustration not visible in this excerpt

By the results depicted in Table 4, the average wind speeds in Morocco and Western Sahara - both obtained through the Moroccan atlas - are close to the average wind speeds calculated by SWERA. Values from MERRA are much lower for both countries. However, as depicted in Figure 20, the Moroccan atlas provides wind speed potentials in a more detailed areal distribution than the compared global atlases. The spread between maximum and minimum wind speeds is consequently much greater in the Moroccan atlas than with its global counterparts, which level maxima and minima out, due to their bigger raster tiles (Figure 15 and Figure 16).

Comparison by Pattern

The comparison of pattern aims to evaluate if a global atlas is deviating uniformly, thereby a constant factor, from a reference atlas or if the differences are heterogeneous.

By this comparison, the characteristic distribution of low and high wind speed areas within a country is evaluated. The Egyptian wind atlas, depicted in Figure 17, is characteristic with highest wind speeds in the Gulf of Suez area and two zones of medium high wind speeds east and west of the river Nile. The Nile delta holds low wind speeds. The wind speed pattern of Tunisia, Figure 18, depicts a line of high wind speeds starting at the north-eastern “Cap Bon” and extending in direction south-east.

First the average wind speed levels of MERRA and SWERA are country-wise elevated to the reference atlases’ averages. The country specific difference in average, depicted in Table 4, is added to each value of the global atlas. The delta between reference and elevated Group G value is calculated by Equation 8.

illustration not visible in this excerpt

Equation 8: Elevating global atlas ’ average to reference atlas ’ average

The smaller the variance of these values, the better the patterns match. Equation 9 consequently calculates the area weighted variance.

illustration not visible in this excerpt

Equation 9: Variance calculated weighted by polygon sizes [167]

Regarding the pattern comparison, the ıp² variance values, depicted in Table 5, show that SWERA correlates better to both reference databases than MERRA. The pattern similarity of Western Sahara as of the atlas “Les resources éoliens du Maroc” matches better to MERRA than with SWERA.

Table 5: Pattern and RMS comparison between reference DB and global DB

illustration not visible in this excerpt

Comparison by RMSe

The Root Mean Square error (RMSe) or deviation describes the difference between two data sets. As described in the validation of the SWERA database, the RMS is used to measure the accuracy of a data source compared to a reference data base. The area weighted RMSe is obtained by Equation 10.

illustration not visible in this excerpt

Equation 10: Root Mean Square error (RMSe) weighted by polygon sizes [168]

Results of the RMSe comparison in Table 5 reveal a better accuracy of SWERA in both cases, with MERRA showing strong dissimilarity in Tunisia. Again for Morocco MERRA strongly differs to the national atlas; SWERA diverges considerably as well.

Result of Databases Comparison

In conclusion, SWERA outperforms MERRA in all comparisons to the reference atlases of Egypt and Tunisia. As a result the SWERA database is utilised in the further assessment for Libya and Algeria. Since the Egyptian and Tunisian wind atlas are of high quality, they are utilised in the further assessment of these countries. Average values by SWERA are similar to averages of the Moroccan wind atlases, covering both Western Sahara and Morocco. The distribution of wind speeds is however of much greater detail with the Moroccan wind atlas, which as a result is employed in the following assessment.

NASA’s MERRA platform provides several databases on varying meteorological topics. For the above comparison, monthly values of the MERRA database have been analysed. When processing MERRA’s hourly time series values, resulting annual averages are higher and appear more accurate than averages calculated by MERRA’s monthly values. Reasons for this discrepancy could not be found. However, due to complexity constraints, these hourly values could not be included into the entire analysis, but will be used in the following Section 4.1.5.

4.1.5 The Form Parameter k

When wind speed data is not available as time series, the form parameter k - also called shape parameter - and the annual average in wind speeds - here called wind regime - are employed to describe the frequency distribution of wind speeds. Favourably large k-values indicate relatively constant wind speeds [169]. Form parameter k is part of the Weibull function, developed by Swedish engineer Weibull, and has its origin in the reliability analysis of metallurgical failures [170].

Without the inclusion of a site specific form parameter k, the Rayleigh distribution is assumed for the frequency distribution of wind speeds. This special case of the Weibull distribution function has the form parameter set to k = 2, thereby representing the most common distribution with moderately gusty winds.

illustration not visible in this excerpt

Equation 11: Weibull distribution adapted to wind [171]

For accurate wind speed distribution predictions it is essential to know the k-values of wind regimes. The database MERRA provides hourly values of wind speeds starting in 1979 until 2011. As mentioned above, these hourly values are very much closer to reference atlases

illustration not visible in this excerpt

than averages by MERRA’s monthly values and are consequently suitable for further processing. K-values for 50 reference points are obtained by Equation 12.

illustration not visible in this excerpt

Equation 12: Regrouped least square method linear regression [172] Where xj and vj are calculated by Equation 13.

illustration not visible in this excerpt

Equation 13: Transferring index and magnitude by median rank to logarithmic scales [173]

First, magnitude and index value of measured wind speeds are processed by the median rank method, thus arranged on logarithmic scales. Furthermore variable A is represented by the median of the value range. In a second step the results obtained above are inserted into the regrouped linear regression, which uses the least-squares-method, to obtain form parameter k.

illustration not visible in this excerpt

Figure 21: Distribution of form parameter k in North Africa

Surprisingly, no k ” 2 is obtained for North Africa, with 70% of the territory characterised by k-values above 2.5. This indicates more constant wind speeds than the generally assumed value of k = 2. Figure 21 illustrates the distribution of k-values over North Africa by these classes and further depicts the distribution of analysed measurement points. According to this map, wind speeds are particularly constant within the southern region of Western Sahara and the Egyptian region encompassing Alexandria, Cairo and the Gulf of Suez. Gustiness is strongest at the North Moroccan and Algerian coastline. For the ease of further processing, the obtained k-values are categorised into three classes by the intervals [2.00; 2.25), [2.25; 2.75) and [2.75; ’).

4.2 Geographical Potential of Wind Energy

This chapter will explain the steps towards the calculation of the geographically available wind potential, thus the suitable and available surface multiplied with the applicable power density factor.

4.2.1 Area Restrictions for Wind Energy

Topology, Geomorphology and Hydrology

According to Held (2010) and Hoogwijk (2004) areas higher than 2000 m above sea level and beyond 15° inclination are considered as not approachable for wind turbine erection and PV ground installation. Additionally air density, a major component with the energy contained in moving air masses, is reduced to a value of 0.945 kg/m³ at 2000 m altitude, compared to 1.29 kg/m³ at sea level at 20°C. This translates to 25% less power output and a generally uneconomic condition for wind turbine deployment at these heights [174]. For the calculation of slope, high resolution data on elevation [175] is first transferred to the Africa Equidistant Conic projection, which projects horizontally distributed points true to their original distance in meters, followed by a spatial analysis for slope calculation.

Shifting sands and dunes are present all over the North African regions. They bear risks regarding stable foundations and abrasion. According to Trieb et al. (2009) sand dunes can move up to 200 m per year. Consequently sandy surfaces are not suitable and are excluded with a buffer of 6 km thereby taking into consideration the movement over 30 years which covers the 20 year lifetime with a ten year bonus for the planning and erection of new projects. Salt pans are present all over the North African region and are excluded due to their intense corrosive characteristics [176].

Lakes and streams as well as canals, perennial and non-perennial stream, e.g. wadis, are excluded from the analysis area. VMap0 by NIMA [177] provides the latter group only as line features. Consequently an average width of 100 m including a security distance is assumed for these water ways.

Settlements, Industrial Buildings, Airports

Urban agglomerations and settlements are excluded from the analysis. A further buffer around respective areas is defined due to the wind turbine’s noise emissions. Requirements are varying by country. Since 2010 Spain obliges a lump-sum 1000 m buffer around existing or planned urban areas [178]. Germany requires a 1000 m distance to residential neighbourhoods[179] with citizen initiatives demanding 1500 m [180]. In the UK the minimal distance from residential premises is claimed according to turbine hub height (hub height of 25-50m: 1000 m buffer; hub height of 50-100m: 1500 m buffer; hub height of 100-150m: 2000 m buffer) [181]. Within projects realised in the North African region Morocco required a 200 m buffer to the nearest village [182] and Egypt considered a buffer distance but realised that the nearest residential area at 6.5 km distance to the Zafarana wind park is well beyond any noise emission zone [183]. VMap0 provides information on built-up areas as polygons and points, which are considered with a 0.5 km radius, thereby assuming a surface of 1.55 km². A 1000 m distance buffer around built-up areas is adopted in this analysis.

Industrial premises are excluded with a 1000 m buffer. VMap0 provides information on location of oil and gas extraction points, power plants and transformer stations, which are considered with a 2 km radius assuming an average area of 12.56 km².

Wind turbines, regionally significant in Germany from a height of 100 m [184], entail risks for aviation and thus have to be situated taking into account airports and its respective security zones. Though in some German law cases a distance of 1900 m was accepted, keeping in mind that a pilot has to navigate at least 150 m above disclosed aviation obstacles, a distance of 4 to 5 km was generally approved [185] [186]. VMap0 again just provides airport locations, therefore a 1500 m radius, assuming 7 km² airport ground with an additional 5000 m buffer zone, is adopted.

Roads and Rails

Requirements regarding the distance between wind turbines and road or rail infrastructure vary with most German federal states asking for 100 m distance due to safety reasons [187]. Vmap0 [188] provides data on rails and roads solely as lines. With assumed widths of 20 m for railways and 40 m for roadways, total widths of 220 m and 240 m are associated to the lines respectively. The resulting area is excluded.

Protected Areas and Bird Protection

Data on protected areas, classified in IUCN categories (see annexed table in chapter 9.1 for criteria) is provided by the World Database on Protected Areas [189], which incorporates the UN List of Protected Areas, as well as nationally registered or planned protected areas. The study on concentrated solar power potential by Trieb [190] excluded all protected areas, while Held [191], analysing wind potentials in Europe, excluded protected areas of the IUCN categories I to III, which are biosphere, vulnerable environments and natural monuments. Within this study protected areas of the IUCN category I to III are excluded. For all other areas - including IUCN category IV to VI as well as areas not classified by the IUCN but nationally protected - a land use factor of 20 % is adopted in this study.

The influence of wind turbines on ornithology, especially in relation to bird mortality and dislocation, is disputed. However, Morocco, Tunisia and Egypt are on major bird migration routes between Africa and Eurasia. Therefore a closer look at suggestions, legislations and examples regarding wind turbine restrictions in relation to ornithology is necessary. A map on North African bird migration flyways is provided by the Egyptian Environmental Affairs Agency [192].

According to German law cases wind turbines are not to be erected in zones of above- average bird migration, known as bird migration corridors. Bird Life International advises “precautionary avoidance of locating wind farms in statutorily designated or qualifying international [...] or national sites for nature conservation, or other areas with large concentrations of birds, such as migration crossing points, or species identified as being of conservation concern” [193]. German State Bird Conservancies recommend a minimal distance of 1200 m to areas important for birds [194]. However, the reality is different with wind parks being built in bird protection areas, e.g. Enercon’s “Altes Lager” wind park in Brandenburg, Germany [195].

For the Moroccan wind parks at the Strait of Gibraltar, a hot spot in bird migration, wind turbine siting was done in order to not obstruct migration flyways, leaving corridors for birds to pass through. The height at tip of blades was limited to 100 m [196]. For the Egyptian Zafarana wind park counting 110 wind turbines, the height at tip of blades was limited to 110 m with 1120 m wide corridors parallel to birds’ flight direction. Coloured blades, stroboscopic lights and deflectors at overhead transmission cables warn animals and sum up to an additional cost of 1.55 million US$ [197]. According to statistical surveys incorporated in the above cited environmental studies, the environmental impact of the analysed wind parks on ornithology was negligible. The environmental study for the Spanish section of the wind park area at Jabal El-Zayt, Egypt, suggests the installation of rather large 2MW wind turbines, thereby meeting energy generation targets with fewer turbines and bigger “escape corridors” [198]. The study on the German section revealed a low local impact on avifauna with bird migration bottlenecks being confined in the southern sections of the designated wind park area and along the coastline, thereby outside of the wind park [199].

Important Bird Areas (IBAs) are distributed all over North Africa, many times overlapping generally protected areas. IBAs of category “A4.iv” only (Congregations: Site known or thought to exceed thresholds set for migratory species at bottleneck sites; see annexed table in chapter 9.1 for more criteria) are considered in this study. Location and size of IBAs is provided by BirdLife International [200]. A circular area with the respective IBA’s size and an additional 2000 m buffer for deviation between real and circular shape, is built around the IBA’s location coordinate. Some part of the IBA area is already incorporated into protected area. For the remaining area of these 7 IBAs, comprising a total surface of 2920 km² - scattered over Morocco, Tunisia and Egypt with 95 % of the total IBA area situated in the Gulf of Suez region in Egypt - a conservative land use factor of 20 % is adopted. Since the observed IBAs represent only 0.066 % of the whole North African area included in the further economic assessment, the considered IBAs are of very low impact in the total assessment and of limited impact in the high potential Gulf of Suez zone.

Parallel Land Use

With just small required ground space, mainly determined by service tracks, wind turbines are suited for parallel use of cultivated areas and forests. What level of parallel land use is feasible in order to prevent land use conflicts? In relation to standard wind turbine density, Hoogwijk (2004) adopted land use factors of 10 % for forests, 50 % for bush lands, 70 % for agricultural lands, 80 % for grasslands, 90 % for savannahs and 100 % for deserts. Held (2010) assessed the potential of wind energy in Europe adopting more conservative factors with 10 % for forests, 50 % for agriculture, grass and grazing lands. Cultivated lands in the North African regions are precious and are thereby considered within this analysis with a more conservative land use factor of 10 % for agricultural and 35 % for grazing areas. Use of forests areas is reduced to 10 % of standard density only. A land use of 80 % is applied for areas which are neither classified nor excluded and considered bare lands, scrublands, savannah or desert with no or low human usage.

4.2.2 Further Restrictions

Natural Hazards

The North African region is characterised by low volcanic activity, with no currently active volcanoes and just few volcanoes active in postglacial time, all of unknown exact eruption dates [201]. Tsunami runups in Northern Africa historically have not been frequent, though high death tolls occurred around Alexandria-Egypt in 1303 and Agadir-Morocco in 1755. Tunisia experienced some damage in 1894 and Egypt in 1810 [202].

North Africa has experienced seismic hazards up to a magnitude of 6.5 on the moment- magnitude scale [203]. The structural characteristics of wind turbines are different to many building structures, with large fixed and rotating masses on a long, stiff pole, which amplifies the ground seismic movements. After field observations at the Tahachapi Pass, California, the University of California in San Diego performed real size tests of a 900 kW wind turbine with earthquake simulation accelerations up to 1G, comparable to the Kobe-Japan earthquake of 1995 with magnitude of 6.9 on the moment-magnitude scale. Turbines with rotating blades are withstanding earthquake movements better than parked rotors as the blades flapping in the wind enhance energy dissipation. On the whole, existing wind turbine tower designs are well suited to withstand earthquake loads of global historic magnitudes [204] [205]. Thus no areas are excluded or reduced in land use due to threats of natural hazards.

Other Areas: Military, Tourism, Overhead Electricity Lines

Hau (2008) additionally considers overhead electricity lines, areas with particular tourism, directional radio corridors and transmitters [206]. However, as no data is available on these topics for the observed area these restrictions are not included in the analysis. Military Areas are not excluded as data on location and size was not available. Though the North African electricity grid is analysed and incorporated in this study, the necessary spatial accuracy for area exclusion is not accomplished.

4.2.3 Wind Power Installation Density

The standard power density for installed wind turbines is obtained by Equation 14.

The adopted value for standard power density is distributed over a wide range in previous studies. Archer et al.[207] assume 9 MW per km² , Hoogwijk (2004) assumes 4 MW per km² in suitable global areas which Held (2010) reduced to 3 MW per km² for suitable areas in

Europe due to social acceptance criteria. Czisch (2005) assumed a lump sum power density of 2.4 MW per km² on all areas, already taking into account reduced power densities of 0.16 MW per km² on agricultural lands and 0 MW per km² on nature reserves.

illustration not visible in this excerpt

Equation 14: Wind power installation density [208]

When comparing these theoretical values to actual power densities in regions with highest WT concentrations the picture is quite different. Denmark (43,094 km² [209] ), currently the country with highest wind turbine density, reached 3802 MW [210] of installed capacity by the end of 2010, including off- and onshore, thus having a WT density of 0.088 MW per km². By the end of 2010 the German federal state of Schleswig-Holstein has globally the highest density of installed wind turbines with 0.191 MW per km² and a capacity of 3003 MW. With 33.3 % of the potential capacity already installed, a projected standard density of 0.572 MW per km² is calculated [211]. Nonetheless, these values are taking into account all areas. Bofinger et al. (2011) furthermore evaluate 22 % of Schleswig-Holstein’s area as available for WT installation, thus equating to a projected standard density of 2.4 MW per km² on included areas in Schleswig-Holstein.

Regarding previous studies, social acceptance and actual density figures a standard power density of 4 MW per km² on not excluded areas is adopted for this study. This standard power density is further reduced according to usability factors depicted in above Section 4.2.1. A summary of these factors, respective data sources and resulting usability assumptions is provided by Table (Anex.) 4.

4.2.4 Results and Discussion of Geographical Potential

Resulting average densities by country are depicted in Table 6.

Table 6: Areas distribution according to land use categories and usability; average power densities on included areas (standard density of 4 MW/km ² )

illustration not visible in this excerpt

The listed results and the geographical distribution of usability and excluded areas in Figure 22 reveal Tunisia and Morocco as being most densely covered with vegetation. Western Sahara consists mainly of rocky desert, which is suitable for WT installation, with very few settlements, which are part of the red-marked exclusion area. At the same time it comprises of very little vegetation but sizeable protected areas in the southern region. Thus Western Sahara achieves high WT density coupled with very high area inclusion of 94.98 %. Libya encounters the highest average WT density, which is due to its little vegetation and land use and no protected areas. However, besides few settlements, Libya has a substantial portion of its country area covered by dunes, which are marked as red and excluded for any WT installation since they do not provide stable grounds.

The application of exclusion and restriction criteria results overall in the inclusion of 74 % of North Africa’s total surface of 6,004,390 km² at a WT density of almost 3 MW per km².

illustration not visible in this excerpt

Figure 22: The geographical distribution of land categories and excluded areas

4.3 Technical Potential of Wind Energy

As Albert Betz revealed in 1920 (please refer to Section 3.1.1 for more information) it is physically not possible to convert all of the wind’s kinetic energy into mechanical energy. While today’s wind turbines are already very efficient, the wind turbine system as a whole certainly encounters further losses, due to its mechanical and electrical components. The total potential AC power output of wind turbines in North Africa is assessed in the following section.

4.3.1 Hub Height

Bofinger et al. employ hub heights of 100 m and 150 m in a recent study on German wind potentials [212]. On the other hand, wind parks in Egypt and Morocco affected by environmental protection restrictions are limited to a height at tip of blades of 100 m. Considering these restriction and today’s technical standards, a general turbine hub height of 80 m is set. This hub height is consistent with studies by Held (2010) and Schermeyer (2011). Thus, with a rotor diameter of 80 m, height at tip of blades is equated to 120 m. Taking into account the historic and projected development in rotor sizes and hub heights, a hub height of 120 m is assumed to estimate the wind energy potential in 2030 and 2050.

Table 7: Average, minimum and maximum wind speeds at hub height in incl. areas

illustration not visible in this excerpt

Though country-wise averages, illustrated in Table 7, are similar with national atlases (EG, EH, MO, TN) and the global atlas (DZ, LY), considerable difference occurs regarding the spread between minimum and maximum wind speeds. The global atlas tends to average all values out leaving a very small spread of just 2 m/s. Tunisia exhibits the biggest difference between maxima and minima with 9 m/s.

illustration not visible in this excerpt

Figure 23: Wind speed at 80 m hub height; excl. areas by geographical assessment

For analysis of the geographical distribution of wind regimes, annual mean wind speeds are mapped at 80 m hub height in Figure 23. Wind regimes are exceptionally good at the West Saharan, southern Moroccan Atlantic coast at the Strait of Gibraltar and its hinterland as well as south of the Atlas Mountains. South west, central south and far south-east Algeria hold reasonably good wind regimes, as does northern Libya around Benghazi and south-eastern Libya. Starting at the north-western Tunisian “Cap Bon” good wind regimes stretch inland into Tunisia on an axis towards south-east. Outstanding wind regimes are again found in Egypt west of the Gulf of Suez, just opposite of the Sinai Peninsula. At this wind surfer’s prime spot strong afternoon breezes are funnelled south from the Mediterranean towards the Red Sea through mountainous ridges situated parallel to the Gulf [213]. Along the river Nile basin and at its western Mediterranean coastline Egypt holds high wind regimes as well.

4.3.2 Losses

External and internal factors lead to losses in the energy yield of wind turbines

Wind Park Effects

Aerodynamic interference between two neighbouring WTs occurs until a distance of 20 times the rotor diameter. Hau suggests a turbine spacing of 8-10 times the rotor diameter on the axis of the primary air flow direction and a distance of 3-4 times the rotor diameters in other directions, thereby reaching an array efficiency of ca. 90 % [214]. Suggested distances equate to ca. 10.6 MW/km² with the reference turbines of this assessment. More recent studies advise a greater distance of 15 times the rotor diameter resulting in more cost-efficient power generation [215]. Employing Hau’s value for secondary air flow direction spacing this equates to ca. 4.8 MW/km². Both density values furthermore demonstrate that the assumed standard density of 4 MW/km² is of rather conservative nature including both social acceptance criteria and technical possibilities. With this lower density factor the resulting average array efficiency is considerably higher than 90 %.

Availability

Due to maintenance, repair, service and operational faults wind turbines incur downtime. Hau suggest limiting expectations in availability to 98 % when planning wind turbine installations. This is consistent with the Tunisian wind park Sidi Daoud, operating since 2000, and having an availability of over 95 % [216].

Electrical losses

Electrical losses occur at the generator, the transformer and the transmission within the wind turbine and to the wind park’s substation. Hau further states that while mechanical losses are included in the WTs power curve electrical losses should not exceed 2 % [217].

Conclusion

Considering all losses between potential power generated and AC electrical output fed into the grid a lump sum loss of 5 % is suggested [218]. This value is adopted in this study as well equalling to an overall performance ratio of the wind turbines of 95 %.

4.3.3 Reference Wind Turbines

The power output of a wind turbine is defined by its power curve, thereby relating a specific wind speed to the wind turbine’s AC output. Within this section the selected reference turbines are presented.

illustration not visible in this excerpt

Figure 24: Distribution [%] of turbines in NA Figure 25: Share [%] of installed turbines by according to their nominal power output manufacturer in North Africa

Sources: Please refer to Table (Anex.) 3 for a list of wind parks in North Africa and respective sources.

Figure 24 and Figure 25 depict the current range of wind turbines installed in wind parks - existing or currently under constructed - in the observed countries (please refer to Table (Anex.) 3 for details and sources). The lion’s share of the turbines are provided by Spanish manufacturer Gamesa, followed far behind by globally leading Danish manufacturer Vestas.

Gamesa is especially strong with North Africa’s prevalent turbine classes 660 kW and 850 kWh.

The situation in Europe however is quite different with the average capacity of newly commissioned wind turbines in Germany at 2.0136 MW in 2009 and 2.057 MW in 2010 [219]. For Denmark, Morthorst et al. of Risø Institute interestingly picture a sharp decline in average WT sizes of new installations, which is in complete contrast to global trends, where India transcends 1 MW and UK, US and Spain approach 2 MW average new wind turbine size.[220] Scheduled projects in North Africa are showing a tendency towards larger wind turbines as well (Table (Anex.) 3). Consequently for forecasting potentials up until 2030 the wind turbine G-80 with a capacity of 2,000 kW, rotor diameter of 80 m and hub height of 80 m, is used as a reference turbine. Figure 26 depicts the performance of the G-80 as very similar to its direct competitive product, V-80 by Vestas. Considering small variations in rotor size and turbine capacity, the G-80 is on average when comparing it to the performance of competing products of the same class. Gamesa ranks on third place according to the global market for large commercial wind turbines in 2007. The fourth biggest WT manufacturer, Enercon, [221] currently has no installations in North Africa.

illustration not visible in this excerpt

Figure 26: Power curve, rotor diameter and capacity of wind turbines [222]

This wind turbine capacity and rotor size is consistent with studies by Held (2010) and Schermeyer (2011) though actual reference products are different in some cases. For LCOE forecasts for 2030 and 2050 this analysis will use the 2,005 kW wind turbine by manufacturer Enercon, namely the E-82, at a hub height of 120 m. The E-82 is a direct-drive wind turbine (for EESG concept description refer to Section 3.1.1), thereby omitting the gearbox. Since manufacturers and markets for newly installed turbines are moving towards this concept, especially for environments where high reliability is crucial, a turbine of this concept is used in the long term assessment. Information on specifications and power curves for both turbines is summarised in Table (Anex.) 7. Figure 26 further shows Enercon’s E-82 to reach full load capacity at lower wind speeds than competing products. The E-82 is thereby more suitable for sites with just moderate wind speed potentials than competing products. Reasons may be found in the streamlined hub shape and tip winglets reducing air flow losses [223].

The threshold wind speed, required to surpass for stable electricity production, is called cut- in wind speed and is at 3.5 m/s with Gamesa’s G-80 and 1.5 m/s with Enercon’s E-82. Nominal capacities are reached at winds of 17 m/s and 13 m/s respectively. Both wind turbines stop power production at wind of 25 m/s, called cut-out wind speed, by actively turning blades into stall and activating brakes for a complete stand still. During operation blade-pitch and azimuth angle are automatically adjusted for optimal power production.[224]

4.3.4 Capacity Utilisation

While the distribution of actual wind speeds within a wind regime at a specific site is defined by form parameter k, described in Section 4.1.5, the annual power production of a turbine is calculated by relating this distribution of wind speeds at the point of interest to the WT’s power curve. The relation between annual power output under these wind conditions and the available nominal capacity of the WT over a year, i.e. the capacity utilisation, is defined as Full Load Hours (FLH) with an upper limit of 8760 FLH. Since the annual performance of wind turbines is analysed, losses described above are included in the calculation of this capacity utilisation. Thereby the FLH value, obtained by Equation 15, describes the calculative time a system would run at nominal capacity in one year.

illustration not visible in this excerpt

Equation 15: Capacity utilisation (FLH) of wind turbines [225]

First the spectrum in wind speeds is calculated for varying wind regimes - accurate to a tenth - under form parameters k = {2.0; 2.5; 3.0}. In a second step FLH values for each reference turbine under varying wind regimes and with varying k parameters are calculated. For interpolation and the ease of further processing, polynomial regression curves for each reference WT and form parameter k are calculated with MS Excel putting wind regime and resulting FLH into direct relation. Next to the diagrams illustrating the WT system’s FLH under various wind conditions, these 6 polynomial regression functions are depicted in Table (Anex.) 7 as well.

The visualisation in Table (Anex.) 7 reveals the regression curves of not being accurate at wind regimes below 4 m/s. Minima of the 6 curves are between 2 and 3.5 m/s with a sharp increase in FLH at wind speeds lower than the minima. Since this is a contradiction to the actual WT power curves and generally illogical, for the further assessment no wind regimes lower than 4 m/s are observed. Consequently the wind regime at 80 and 120 m height is set to zero at affected polygons. By this methodology low speed wind regimes are excluded of the analysis resulting in an overall reduced technical potential. Thereby the pure technical potential cannot be provided through this methodology.

In a last step the regression curves are multiplied with the lump sum performance ratio (PR) adopted by the evaluation of Section 4.3.2.

In further adjusting the technical potential to the requirements for the subsequent economic potential analysis, sites considered of economically not viable yields are excluded as well. Bofinger et al. applied a threshold of 1600 FLH in site selection [226]. Following the approach by Held (2010), sites not surpassing the threshold of 1300 FLH are excluded from the further analysis.

Including the PR of 95%, thresholds for Gamesa’s G-80 the rounded wind regimes vG = {5.3; 5.5; 5.7} with form parameter k = {2.0; 2.5, 3.0} respectively. Thresholds for

Enercon’s E-82 are at the rounded wind regimes vE = {5.0; 5.2; 5.4} with form parameter k = {2.0; 2.5, 3.0} respectively also including the PR with 95 %.

By Equation 16 the installed capacity in each polygon i is obtained for each of the three scenarios s (G-80 fixed k = 2, G-80 with variable k, E-82 with variable k).

illustration not visible in this excerpt

Equation 16: Installed capacity considering capacity utilisation threshold [227]

The annual AC power output of WT systems in each polygon under scenario s (aPis) is calculated by Equation 17.

illustration not visible in this excerpt

Equation 17: Annual power generation of wind turbine under various scenarios [228]

4.3.5 Results for Technical Potential of Wind Energy

Table 8 illustrates cumulated results for each country and North Africa with and without the variation in form parameter k. As discussed above, this is not the pure technically available potential, but it forms the input factor for the following economic analysis.

Analysing the results of Table 8, one has to bear in mind, that due to a higher spatial resolution, the employed national atlases generally depict a wider spread in average wind speeds. This spread is illustrated in Table 7. The noticeably small power generation potential in Tunisia is due to a large spread with a minimum wind speed of 1.79 m/s and the maximum of 11.02 m/s at 80 m hub height. Producing an average wind regime of just 4.99 m/s at 80 m hub height, only between 38 and 50 % remain as part of the analysed area depending on hub height and wind regime defined by form parameter k. All other areas in Tunisia are excluded since they yield <1300 FLH. Table (Anex.) 8 further illustrates the remaining area when excluding sites below the capacity utilisation threshold of 1300 FLH.

Table 8: Installed capacity, annual AC power output and average annual FLH by country and installation type

illustration not visible in this excerpt

Effects of Variable k and Capacity Utilisation Threshold

By including form parameter k with varying values depending on location into this analysis, the calculative power output of North Africa is lowered by 3580 TWh/a in comparison to a general assumption of k = 2 as used in studies by Hoogwijk (2004), Held (2010) and Czisch (2005).

FLH over wind regime diagrams for both reference turbines, as depicted in Table (Anex.) 7, show lower performance for k = 2.5 and k = 3 at wind regimes between 4 and 8 m/s than the curves employing k = 2. However, at sites providing a wind regime above 8 m/s the situation is reversed and k > 2.0 outperforms k = 2.0.

K parameters of 2.5 to 3.0 together with wind speeds around 6 m/s - the North African average - thereby translate to a predominant WT operation in partial load, which is the case with Tunisia, Algeria and Libya with an average capacity usage of just 1600 to 2000 FLH. In contrast, Morocco and Western Sahara exhibit highest capacity usages of around 3000 FLH at 80 m hub height and 3500 to 3700 FLH at 120 m hub height. In the Western Saharan case this is due to highest average wind speeds coupled with highest k parameter, resulting in frequent operation at nominal capacity. In the Moroccan case only 24 % of the areas included after geographical assessment are yielding above 1300 FLH. Consequently much of the Moroccan territory is characterised by low to moderate wind speeds. However, the areas which surpass the threshold are very high yielding for WT installations.

When k = 2 is applied, Libya is not affected by the capacity utilisation threshold of 1300 FLH. However, since Libya’s minimum wind speed of 5.15 m/s at 80 m hub height is just at the threshold wind speed, Libya’s suitable area drops by 24.2 % when including parameter k above 2 in the analysis. Table (Anex.) 8, gives further information on the remaining area when applying k fixed and k variable and a threshold of 1300 FLH.

Since the inclusion of form parameter k provides a more accurate picture of the frequency distribution of available wind speeds, the further analysis will produce results calculated with varying k parameters. As pointed out above, for general evaluation purposes values of the Gamesa-G-80 will be considered, for forecasting future potentials beyond 2030 potentials of the Enercon-E-82 will be utilised.

Discussion and Comparison with Results of Similar Studies

In regard North African wind turbine power output potential, forecasted demands in 2030 for the EU (4560 TWh/a with a 41% share in RES [229] ) and North Africa (647 TWh/a [230] ) could be fully covered by 28 % of the North African wind potential at 80 m hub height .

In Germany about two percent of the total surface is geographically available for wind energy exploitation and at the same time holding a potential greater than 1600 FLH [231]. While this value is proven and viable for Germany and might also be for neighbouring countries, North Africa is very different in its geographical structure and settlement density. Consequently more area is available for WT installation. Thus 40 % of the North African area is suitable and achieves more than 1600 FLH at 80 m hub height. Values range from Tunisia (15 %) to Western Sahara (77 %).

For North Africa Hoogwijk (2004) assesses a technical potential of 3000 TWh/a and a power density of 0.42 MW/km² on all surfaces regardless of usability. However, the CRU global wind atlas, which Hoogwijk employed, represents North Africa with an average wind speed of 2.9 m/s only. By restricting usage to sites holding wind speeds above 4 m/s, 90% of the total surface is thus excluded [232]. Consequently, the technical potential in this study is six times higher although similar land use restrictions are applied. Overall power density for North Africa equates to 1.74 MW/km² with values varying between low densities in Tunisia (0.33 MW/km²) and Morocco (0.4 MW/km²) and very high density in Western Sahara (2.53 MW/km²).

Czisch (2004) computes a power output of 1552 TWh/a for sites of a capacity usage greater than 3040 FLH in Algeria, Western Sahara and Morocco. Applying the same restrictions with the Gamesa-G-80, Morocco’s potentials add up to 386 TWh/a and Western Sahara’s to 470 TWh/a. Algerian potentials do not surpass the threshold of 3040 FLH.

Czisch (2004) further calculates that Libya, Tunisia and Egypt hold a combined generation potential of 1622 TWh/a at capacity usages greater than ca. 2160 FLH. With the same restriction this study determined a value of 2195 TWh/a, mainly provided by Egypt contributing ca. 1817 TWh/a. Czisch uses the ERA-15 global atlas with raster tiles of 1° (equal to 111x111 km like SWERA) and a density factor of 8 MW/km² compared to 4 MW/km² in this study.

Enzili [233] evaluates Morocco with a potential of 4896 TWh/a and a installed capacity of 2448 GW. These figures are very close to the values calculated here: 5,149 TWh/a generated at 80 m hub height with varying k values and a installed capacity of 3,083 GWp.

While Scholz (2010) analyses the wind energy potential of North Africa and Europe jointly, the potentials of wind energy in North Africa cannot be differentiated.

4.4 Economic Potential of Wind Energy

In contrast to conventional fossil fuel fired power plants, where as much as 40-60 % of electricity generation costs are related to expenditures for fuel and O&M, WT energy generation is capital intensive with low operational cost [234]. The following section grasps the factors driving economic viability of wind energy.

4.4.1 Financial Parameters

Different financial incentive schemes support renewable energy deployment in North Africa. Egypt, for example, underwent a model with strong state-participation: The Zafarana wind parks are financed by international governmental agencies providing grants and soft loans, with the proprietor, the National Renewable Energy Authority (NREA) of Egypt, exempt from customs duties. The cost of debt is low - interest rates from 0.75 % to 1.3 % - with long debt payment periods - 30 years, thereof 10 years grace period. At Egyptian projects in cooperation with Denmark, the Danish International Development Agency (DANIDA) paid the interest rate for 9.5-year credit loans [235]. Completed projects in Tunisia were highly subsidised by government loans with an interest rate of 0.1 %, 30 years debt payment period thereof 24 years grace period [236].

However, these incentivising financial parameters are no more than an initial push for the RES industry. They are not realistic for the total investment required to accomplish the share of RES envisioned in future energy mixes of the countries. Future projects in the Gulf of Suez area are scheduled by the NREA to be private, specifically of the Build-Own-Operate (BOO) scheme (for further information see Section 2.1.5). Consequently, financial parameters have to be oriented at open market levels. Altering Kost and Schlegl (2010), the Weighted Average of Capital Costs (WACC) is rounded up to 8 % with a capital distribution of 30 % equity, with 10 % return on equity, and 70 % debt, with an interest rate of 7%.

The WT system’s lifetime is set to 20 years, which is the generally adopted value in comparison studies [237] [238] [239] [240] [241].

Investment and O&M

In estimating the initial investment and annual operation and maintenance costs (O&M), several comparisons studies are considered. Hau (2008) determined a rule of thumb of 130 % to equate the cost increase from ex-works turbine costs to turnkey wind energy investment. Within the economics of various projects, stated by Hau (2008), grid connection accounted for 6-12 % of total project costs [242]. This leads to the assumption that turnkey WT prices without grid connection are 20 % higher than ex-works turbine costs. Investment and O&M costs of quoted studies are adapted by these means and provided for comparison purposes in Table (Anex.) 9. The evaluation concludes with the financial parameters stated in Table 9. These values are further utilised in this study. Initial investment includes costs related to equipment, installation, commissioning and average transport costs, but is exclusive of grid connection. O&M costs, including parts replacement, maintenance, insurance and management overhead, can be assumed either for each kWh generated or as a percentage of the initial investment. For comparison purposes, O&M costs of the quoted studies are converted to percentage of the initial investment and depicted in Table (Anex.) 9 as well. The analysed total system sizes vary greatly: Hoogwijk (2004) and Hau (2008) consider single wind turbines below 1 MW capacity amongst others. The capacity of 200 MW of the Egyptian wind park “Jebel El-Zayd 2” is financed with 340 million € [243]. By 2015 the installed wind power capacity within all projects in the “Jebel El-Zayt” area is expected to reach 2530 MW [244].

Table 9: Assumptions for the LCOE calculation of wind energy valid for the year 2012

illustration not visible in this excerpt

In contrast to values adopted by Kost and Schlegl (2010), annual O&M expenditures are not increasing over lifetime.

4.4.2 Cost-Supply Curve of Wind Energy in 2012

The Levelised Cost of Electricity (LCOE) is subsequently obtained by Equation 18.

illustration not visible in this excerpt

Equation 18: General of LCOE calculation [245]

To calculate the LCOE, one can also directly process the sites capacity usage in full load hours employing Equation 29 located within the economic assessment of PV systems - Section 5.4.1.

4.4.3 Results for Economic Potential of Wind Energy

Resulting Cost-Supply curves for 2012 are depicted country-wise in Figure 27. Morocco and Tunisia are represented on a smaller scale within the right-side diagram. The Moroccan atlas provides four classes for the wind speeds specifications of Western Sahara (purple line); with the lowest one not surpassing the threshold of 1300 FLH. The resulting Cost-Supply curve for Western Sahara appears in a step-like fashion compared to a smooth Tunisian Cost- Supply curve (green line), with a very detailed differentiation of potentials. The spread between minimum and maximum LCOE is considerably larger in countries where country specific atlases were employed (EG, EH, MO TN) than in countries where a global atlas was used (DZ, LY).

Figure 27: Cost-Supply curves depicting marginal LCOE [ct € /kWh] over cumulative, annual power generation [TWh/a] for 2012; G-80, 80 m hub height, variable k

Best North African sites hold LCOE potentials of just 3 ct€ per kWh. Steep curves for Egypt and Tunisia indicate little generation potential at lowest LCOE relative to the total country specific generation potential.

Discussion of Results

Elsobki et al. (2009) state staggered Egyptian consumer prices as of 2008, starting for residential users at 0.72 ct€ per kWh for the first “lifeline” of 50 kWh monthly. Residential demand over 1000 kW per month is priced at 6.47 ct€/kWh. Prices for industrial and commercial users vary similarly between 3.16 and 7.84 ct€ per kWh [246]. According to this analysis, Egypt holds an annual wind energy generation potential of 1858 TWh/a below the threshold of 8 ct€ per kWh. A generation potential of 334 TWh/a is situated below the threshold of 6 ct€/kWh, that is 10.7 % of the total Egyptian generation potential and almost twice the Egyptian electricity demand forecasted for 2020 [247].

An LCOE of 8 ct€ per kWh is not surpassed by 5379 TWh/a North African generation potential, the marginal cost of 6 ct€/kWh is not reached by 1267 TWh/a, that is 7.1 % of the total North African potential and a 3.5 fold of the North African electricity demand forecasted for 2020 [248].

The cost of electricity generation through wind turbines is calculated for 2012 using Gamesa’s G-80 at 80 m hub height as a reference with wind regimes variable in k. Grid connection is not included. Transport costs - marine and land - are included for average WT installation sites, with locations not significantly far-off suitable transport infrastructure. In a vast country, such as Algeria and Libya, these constraints can have a considerable impact on siting economics and will therefore be analysed in Chapter 6.

The geographical distribution of highest yielding sites, thereby providing lowest LCOE is illustrated in Figure (Anex.) 4. The potential distribution is consistent with wind regime distribution as described for Figure 23. Morocco, Western Sahara, Tunisia and Egypt appear to be much less balanced in the distribution than Algeria and Libya. This again is due to better resolution of the national atlases in site specific wind speeds compared to the global atlas.

Comparison with Similar Studies

The costs of wind energy, as stated above, are comparable to values provided by Kost and Schlegl (5.4 ct€/kWh - 11.9 ct€/kWh) at sites with capacity usages between 1300 and 2700 FLH. However, the exceptionally high yielding sites at the Gulf of Suez, southern Morocco and Strait of Gibraltar are closer to capacity usages of offshore wind turbines. The LCOE for offshore wind power is still much higher with 10-14 ct€/kWh at sites achieving 3600 FLH [249].

According to Held, highest yielding onshore wind sites in the EU achieve LCOE as low as 4 ct€/kWh at capacity usage above 3000 FLH [250]. Values obtained in the study in hand are consistent with Held considering higher wind speeds at best North African sites and improvement in equipment costs since 2010.

Schermeyer (2011) analyses site specific potentials by an hourly resolution in wind speed measured at the same sites coupled with an hourly resolution in air pressure derived from ambient temperature values and air density by elevation. Sites are selected according to the projected wind parks approved by the Moroccan government. Employing Gamesa’s G-80 at 80 m hub height as reference, an initial investment of 1176 ct€/kWh and a WACC of 8.5 %, Schermeyer equates LCOE of 4.2-5.4 ct€/kWh at Moroccan sites yielding between 2951 and 4289 FLH. Though employing similar values, the same sites equate to 1 ct€/kWh lower LCOE of 3-4 ct€/kWh in the analysis on hand.

While economic and technical inputs into the equation are very close, the wind input data and its processing are of greater detail with Schermeyer (2011): Comparing data from meteorological stations in Morocco he states that a height difference of 500 m results in 5 % less air density. However, power curves of reference turbines are rated by default at 1.225 kg/m³, which is - according to Schermeyer (2011) - the upper end of observed air densities at Moroccan at sites. Consequently, he utilises air density specific power curves provided by the manufacturer. Elevation in combination with temperature fluctuations thereby translate to a 5 % reduction from the nominal capacity of 2 MW [251]. Thus, the study in hand possibly overestimates the generation potentials by 5 % taking into account the overall elevation distribution and temperature levels of North Africa. For higher accuracy the inclusion of elevation and temperature data coupled with high resolution wind atlases is suggested.

Sensitivity Analysis

By the argument of possible parallel land use when deploying wind turbines, the geographical assessment of this analysis includes cultivated lands, forests and to some degree also protected areas (IUCN classes > IV, IBA A4.iv). According to Hau (2008), only the required basement area of 200-400 m² cannot be used parallel [252].

Table 10: Land use sensitivity analysis for wind energy

illustration not visible in this excerpt

Since cultivatable lands are scarce in Northern Africa and may not be suficciently protected by this assumption, a sensitivity analysis is conducted shedding light on two scenarios. Scenario 1 analyses the North African wind energy generation potential when excluding agricultural and grazing lands, forests, important bird areas and protected areas. Scenario 2 additionally reduces the default usability value of 80 % adopted for the remaining areas (referred to as “other”), thus savannah, desert, scrubland and bare land to 50 %.

By the results given in Table 10, more conservative land use factors of scenario 1 have only a marginal effect (1.3 %) on the North African generation potential.

From a country-specific perspective, the Tunisian wind energy potential is reduced by 27.8 %; Moroccan potential is reduce by 15.5 %. A significant change in wind power LCOE can be observed in Tunisia: The minimal wind energy LCOE is elevated by 1.5 ct€/kWh, thus highest yielding Tuinisian wind sites are situated in areas of parallel land use. Moroccan minimum wind energy LCOE remains at 2.99 ct€/kWh.

Regarding both generation cost and potential, Egypt, Libya, Algeria and Western Sahara are not sensitive to parallel land use.

In scenario 2, the reduction of the default usability factor for savannah, desert, scrubland and bare land (“other”) drastically narrows the available North African potential from 17,865 TWh/a, as stated in Table 8, to 6615 TWh/a.

4.4.4 Cost Development of Wind Energy until 2030 and 2050

Prices for wind turbines have been fluctuating drastically: 2008 was characterised by shortages in steel and copper, resulting in prices for ex-works wind turbines peaking at 1260 € per kW. The oversupply in 2009, due to clients financial constraints and extended production capacities, meant a sharp drop in WT prices of 978 € per kW [253].

Learning curves describe the reductions in unit price due to growing production volumes, which lead to learn effects related to manufacturing and R&D. For each doubling of cumulative production the unit price is reduced by the learning rate.

Forecasted LCOE for wind energy are obtained by Equation 19.

illustration not visible in this excerpt

Morthorst et al. (2009) suggest applying the learning curve developed by Neji et al. (2007) directly on current LCOE. Thereby improvements regarding wind turbine efficiency, operation and maintenance and initial investment cost are embodied. Morthorst (2009) further states learning rates of 9 to 17 % [255]. The efficiency improvements expected by the U.S. Department of Energy vary depending on WT system component, e.g. 4 % for drive trains and 25 % for rotors in relation to capacity usage [256]. As a result, estimating total system efficiency improvements is regarded as not within the complexity limits of this analysis, and the approach suggested by Morthorst will be adopted. Influences on prices due to demand or supply driven market fluctuations are not considered with this assumption.

For forecasting purposes, this study employs Enercon’s wind turbine E-82 at 120 m hub height. This altered reference provides an enhanced power curve, compared to Gamesa’s G-80, and a 40 m increase in hub height. To some degree, expected efficiency improvements are thereby already taken into account. The learning rate applied on today’s LCOE is thereby assumed with 10 %.

Basing its analysis on the competition of various technologies, the IEA’s resulting energy source portfolio estimates a globally installed WT capacity of 1000 GW in 2030 and 2000 GW in 2050 [257] up from a 2011 global capacity of 240.5 GW [258]. As a result, LCOE for 2030 are estimated to be at 80.5 % of today’s value. Forecasts for 2050 indicate the LCOE of 72.5 % of today’s LCOE. The employed forecasts in cumulative installed capacity are rather conservative compared to further studies, which forecast 1500 GW to 2500 GW for the year 2030. Applying these higher forecasts and a WT system learning rate of only 3 %, Kost and Schlegl estimate initial investment in 2030 to be at 89.4 % of today’s investment [259].

The E-82’s lower cut-in wind speeds and generally higher wind speeds at 40 m additional height, result in more sites surpassing the described threshold of 1300 full load hours. Total North African annual WT generation potentials are thereby 45 % higher than with Gamesa’s G-80 at 80 m hub height, reaching 25893 TWh/a (see Table 8). Consequently, the choice of turbine and hub height is essential for optimal site utilisation.

Previous to LCOE forecasting a baseline scenario is calculated employing prices of 2012 and generation potentials by the E-82, see Figure (Anex.) 1. This baseline scenario reveals generation cost improvements not being uniform over the observed countries. For Egypt, the lowest LCOE drops only by 0.12 ct€/kWh when in Algeria and Libya the drop for minimal LCOE stands at 1.8 and 1.36 ct€/kWh respectively. With this baseline scenario applied for Egypt, 2040 TWh/a are generated below 6 ct€/kWh and 3606 TWh/a are generated below the threshold of 8 ct€/kWh. Egypt’s total potential adds up to 4465 TWh/a.

Results and Discussion of Forecasted Wind Energy LCOE

illustration not visible in this excerpt

Figure 28: Cost-Supply curve depicting marginal LCOE for 2030; E-82, 120 m hub height, variable k

Cost-Supply curves for 2030 are calculated by decreasing this baseline LCOE by 19.5 %, with results illustrated in Figure 28. In the case of Tunisia, we see a much less sloped curve depicting a doubling in generation potential, due to more sites surpassing the 1300 FLH threshold. In the Egyptian case 4211 TWh per year could be generated at costs lower than 8 ct€ per kWh in 2030. A potential of 3350 TWh/a, that is 75 % of the total potential, stands at a marginal LCOE of 6 ct€/kWh.

Following the declining LCOE until 2050, we find that in Egypt a wind power generation potential of 4395 TWh/a is situated below the threshold 8 ct€/kWh, 3733 TWh/a, equal to 83.6 % of the total Egyptian potential, are below the 6 ct€/kWh threshold.

The plotted energy costs of 2030, Figure (Anex.) 5, show less sites being excluded due to the 1300-FLH-treshold compared the cost map for 2012 (Figure (Anex.) 4). In particular in northern Algeria and northern Libya as well as in Tunisia, the turbine’s better efficiency at lower wind speeds allows more sites to remain included in the Cost-Supply assessment.

Analysing the energy supply for North Africa and Europe in 2050, Scholz (2010) computes wind LCOE between 2.5 and ca. 10 ct€ per kWh. According to Scholz (2010), the electricity demand of North Africa and the European continent will add up to 9,560 TWh/a in 2050.

By this assessment, 9560 TWh/a could be generated at a marginal LCOE of 4.94 ct€/kWh. For 2050, a generation potential of 20015 TWh/a is forecasted with marginal costs of 6 ct€/kWh. Cost-Supply curves for 2050 are illustrated in Figure (Anex.) 2.

4.5 Discussion of the Results

The Cost-Supply assessment of wind energy in North Africa points out that today a generation potential of 1267 TWh/a is below the grid parity LCOE of 6 ct€/kWh, the value adopted by Kost and Schlegel (2010) for the year 2010. This equals 7.1 % of the total generation potential and an area of ca. 234,600 km², between the country size of Tunisia and Western Sahara. The analysis for 2030 reveals all potentials available at lower costs than the expected power mix LCOE of 10 ct€/kWh[260]. These results are consistent with Kost and Schlegel (2010) and demonstrate the cost effectiveness of wind power.

The European Wind Energy Association (EWEA) [261] states wind onshore LCOE of 6.5 ct€/kWh in 2010; similar to the values equated in the study in hand. When comparing the cost of wind onshore energy to other technologie - incorporating moderately volatile fuel prices, gas-fuelled power plants are more economical by 2.5 ct€/kWh, the LCOE of coal-fired power plants is similar to onshore wind power LCOE. LCOE of nuclear energy is higher by 3 ct€/kWh. Already by 2020 onshore wind power will hold lower LCOE than all other energy technologies observed by the EWEA and will further improve its competitiveness to LCOE of 5.5 ct€/kWh in 2030. Offshore LCOE is ought to decrease from LCOE of 9 ct€/kWh in 2010 to LCOE of 6.5 ct€/kWh in 2030, and will consequently remain at a higher price than onshore wind energy.

Although onshore wind energy is already balanced in cost with fossil-fuelled energy technologies, it has its drawbacks as well. Wind energy cannot simply be switched on when required. Possible cost increases for the power network due to the volatile nature of wind resource are not included in this assessment. Especially in Egypt and Morocco, scheduled projects will boost the cumulative installed wind capacity (please refer to Section 2.1 for further information). For one sample year, Schermeyer (2011) simulated the hourly variation in generated power of the total Moroccan wind power capacity scheduled at 2000 MW in the year 2020. As depicted in Figure 29, the minimal capacity usage is 21 MW; the maximal usage is 1713 MW. How may one be able to decrease this spread in capacity usage and the underlying fluctuation? What are the implications of this volatile power source for power networks regarding secure electricity supply? Section 7.2 will dwell deeper into this topic.

illustration not visible in this excerpt

Figure 29: Simulated electric energy generation by Morocco wind parks [262]

Country sizes in North Africa are enormous. The highest yielding sites might be too remote to make energy exploitation viable. So far, this analysis did not take into consideration the distance of potential sites to existing infrastructure. Chapter 6 will extend the Cost-Supply assessment of wind energy in North Africa by the inclusion of distance to infrastructure. Furthermore, the potentials in supplying the European electricity demand at competitive prices including the transmission costs to European load centres will be evaluated.

5 Cost-Supply Assessment of PV Energy in North Africa

Applying the general methodological approach in LCOE calculation illustrated in Section 1.3, this chapter gives detailed information on the calculation of today’s LCOE for open area ground installed, roof installed and facade integrated PV in North Africa with an additional LCOE estimation for 2030 and 2050. The general methodology is therefore adapted according to the scheme depicted in Figure 30.

illustration not visible in this excerpt

Figure 30: Scheme of LCOE calculation methodology adapted to PV energy

5.1 Theoretical Potential of Photovoltaic Energy

5.1.1 Solar Irradiation

Just before entering the earth’s atmosphere the sun’s irradiation is 1.353 kW/m², with the spectrum - defined as Air Mass 0 (AM 0) - similar to the radiation of a black body at the sun’s temperature of 5762°K. Passing the atmosphere, the spectral distribution of light is altered resulting in a reduction in energy. According to the position on earth, the atmosphere’s thickness varies. AM 1 is the resulting spectrum at a zenith angle of 0°, equal to sun light passing through the atmosphere at the shortest, direct way. Generally an AM value of 1.5 is assumed. Local aerosol, ozone and water vapour concentrations, as well as air pressure further reduce solar radiation. Increasing ground albedo enhances irradiation by diffuse reflection [263]. Spectra at AM 0 and AM 1.5 are depicted in Figure 31. Solar cells utilise diffuse and direct sun light, both are included in global irradiation.

illustration not visible in this excerpt

Figure 31: Irradiance spectrum with AM 1.5, AM 0 and black body 5762 ° K [264]

As illustrated in Section 3.2.1, the characteristics of the pn-junction within the solar cell are adjusted according to the wavelength spectrum to reach an optimal power outcome.

5.1.2 Acquisition and Transformation of Solar Irradiation Data

The Climatologically Solar Radiation Model by the U.S. National Renewable Energy Laboratory uses information on cloud cover, water vapour, trace gases and aerosol concentrations to calculate irradiation on horizontal surfaces. The data is provided in monthly and annual averages through the SWERA platform by the United Nations Environment Programme. Validation of the data has shown that data accuracy diverges by about 10 % to recorded ground values. This is due to differing local weather conditions, e.g. cloud cover, within the grid tiles of 40x40 km. Complex topologies further reduce reliability in data accuracy. The gathered data for irradiation on horizontal surfaces is recalculated to global latitude-tilt irradiance (LTI), the data type used in the further analysis [265].

For PV ground installation this analyses will only take flat to medium inclined surfaces into account. In regard to rooftop installation, it is assumed that North African roof tops are generally horizontal as well. Consequently, for optimal irradiation utilisation PV modules are mounted on a southerly oriented rig, equalling to an azimuth angle Į of 0°, with a tilt angle ȕ equal to the latitude position of the installation site which in North Africa is between 19° and 37° north. Data for this type of insolation, called Latitude Tilt Irradiation (LTI) is directly provided by the SWERA database.

Although mounting of the PV module on a rig moveable on one or two axes would allow an optimal sun tracking and thereby a better insolation capture, it also comprises much higher investment and O&M costs and is consequently uneconomic for genuine c-Si and thin-film modules. Consistent with Held[266] only fixed installation of the PV modules at the above explained angles is considered in this study.

Following the approach by Held, PV facade installations are assumed to be in a vertical orientation (ȕ = 90°) towards south.[267] Since no data for this type of insolation is available, irradiation on these surfaces is obtained by first calculating the above mentioned LTI data to irradiation on horizontal surfaces and further calculating it to irradiation on vertical southerly orientated surfaces; jointly by Equation 20.

illustration not visible in this excerpt

Equation 20: Converting LTI to irradiation on vertical surfaces [268]

A small part of the observed area is located below the Tropic of Capricorn at 23°26’ north. This tropic is defined as the northernmost latitude with a zenith of 0°. Consequently close to summer solstice southerly oriented facades located south of this Tropic are receiving insolation from the north resulting in a negative insolation in June and July. These negative values are readjusted to zero for annual irradiation average calculation purposes. The phenomenon is depicted in Figure 32.

illustration not visible in this excerpt

Figure 32: Irradiance on latitude-tilted and vertical surfaces in the observed region at most northern and most southern locations with similar longitudes

PV modules can be integrated into facades at different angles, e.g. parallel to structural walls or as a canopy providing shade in summer [269]. While solely vertically oriented surfaces are included, the resulting power generation is prone to underestimation. Average annual irradiation values by country are depicted in Table 11, Section 4.2.4.

5.2 Geographical Potential of Photovoltaic Energy

5.2.1 Suitable Area for Ground Installation

Following a bottom-up approach this study analyses the total area of North Africa in regard to various factors, thereby excluding areas not suitable or not available for PV ground installation.

The databases and the methodology to ascertain the suitable area for PV ground installations are very much similar to the methods applied with the analyses for wind energy depicted in Section 4.2.1. A summary providing data sources and analysed factors for PV and WT technology is made available through Table (Anex.) 4.

Topology, Geomorphology and Hydrology

For accessibility reasons, especially during construction phase, areas higher than 2000 m and of greater inclination than 15° are excluded. Since aspects can be calculated with GIS, areas oriented towards northwest, north and northeast with an inclination greater than 5° could be identified. While these areas do not receive satisfactory irradiation their exclusion would enhance the accuracy of the overall potential assessment.

Within the analysis for CSP deployment in North Africa, Trieb excludes a 10 km buffer around shifting sands and dunes thereby including dune motion over 50 years. Their moving nature, unsuitability for construction and presumably downgrading effect on system availability give reasons for the exclusion with a security buffer. Salt pans are excluded as being of high corrosive risk for structural components [270].

Again for CSP rollout Trieb does not exclude small waterways arguing, that the exact location of large scale CSP plants is at a tolerance of ± 500m leaving room to omit waterways [271]. Conversely while PV systems are very variable in scale ranging from system below 1kW to multi megawatt parks all kinds of perennial and non-perennial waterways are excluded with the same methodology as described in Section 4.2.1.

Settlements, Industrial Buildings, Airports

Settlements, industrial buildings and airports are excluded from the area available for ground installations but will be further analysed for the potentials of rooftop and facade mounted PV systems. Since Vmap0 [272] provides only large, urban agglomerations in polygons, smaller settlements, available only as points, are taken into account by creating a circular buffer assuming an average size of 1.55 km². Points classified as industrial buildings and airports are assumed of an area of 12.56 km² and 7 km² respectively.

Roads and Rails

Solar Roadways incorporating solar cells into glass driving surfaces are an exciting idea yet still in initial development [273]. Covering roads with PV roofs is a further idea. However, with the abundance of available area in North Africa only genuine, ground installed PV is incorporated here. Vmap0 provides information on road and rail networks solely as lines, consequently these features are excluded with an assumed width of 40 m and 20 m respectively.

Protected Areas

Following Trieb[274] all areas categorised as protected by the WDPA [275] are excluded from the available area for PV ground installation.

Parallel Land Use

Held (2010), according to Sørensen (1999) [276], considers 0.5% of agricultural land, 1% of grassland and 5% of marginal land as available for PV ground installation [277]. These values are slightly lower than Hoogwijk’s, who assumed 1 %, 1 % and 5 % respectively. While agricultural and grazing lands are scarce and thereby precious in North Africa, areas classified by Vmap0 as crop and grass [278] are excluded from the further analyses. Additionally, forests are not suitable and are excluded as well.

Further Restrictions

Further possible restrictions, e.g. military areas and tourism areas, are not included in this analysis. For a detailed explanation see Section 4.2.2. Hilgers (2010) evaluated the impact of natural hazards on the potential of CSP in North Africa

5.2.2 Coverage Factor and Results for Ground Installed PV

Recapping the restrictions of this section, only bare land, scrubland, desert and savannah are considered for PV ground installation within this study. The next step analyses the usability of the included areas.

While Scholz (2010) excludes crop and grass land as well, she assumes 100% availability on bare land and deserts with a distribution of ѿ Wind, ѿ PV and ѿ CSP plants. Within included areas Hilgers (2010) evaluates density factors of 5 and 10 % for CSP plants in North Africa, finding only generation potential but not generation cost as being affected by the variation in density factor.

Following assumptions for marginal land by Held (2010) and Hoogwijk (2004), PV ground systems are assumed to be situated on 5 % of the available surfaces. Yet, in order to prevent shading by neighbouring modules, accommodate system components and provide space for service tracks in PV parks, only 50 % of this surface is actually assumed to be covered by PV modules, resulting in a coverage factor CF of 2.5 % on areas not excluded by the geographic assessment of above.

Table 11: Areas for PV ground installation, module area and received irradiation

illustration not visible in this excerpt

Results for Geographical Potential of Ground Installed PV

Results on area exclusion diverge greatly within the North African country spectrum. Only 9.7% of the area is excluded with Western Sahara, since it is predominantly characterised by uninhabited and uncultivated rocky desert and thereby highly suitable for PV ground installations. This is in contrast to Tunisia, where 52.1 % of the area is not available. Here, a more favourable climate allows agricultural use of land, and population density is much higher than in Western Sahara with many settlements and urban agglomerations. Still a sizeable proportion of Tunisia is of sandy deserts, in particular in the south western region. These figures as well as the estimated PV ground installation module are listed by country in Table 11 with a geographical illustration of insolation and excluded area in Figure 33.

In regard to PV no comparison study depicting figures on total included area for PV ground systems were found. Though analysing CSP potentials in North Africa - without Western Sahara, Hilgers applies similar exclusion criteria with a difference regarding inclination: slopes greater than 2 % - equal to 0.9° - are excluded. As a result on average 56 % are included [279]. Not considering West Sahara, this study includes 66 % of North Africa's surface, thereby still 10 % more than Hilgers.

illustration not visible in this excerpt

Figure 33: Diffuse and direct solar irradiation on surfaces tilted according to latitude Annual global irradiation is thereby reaching best values between 2500 and 2700 kWh per m² in south-eastern Algeria, southern Libya and most of Egypt. Hilgers compared the SWERA database regarding direct normal irradiance (DNI), used for energy generation by CSP and CPV technologies, with a database developed by the German Aerospace Centre (DLR) and utilised by Trieb et al. [280]. In southern and south-western Algeria, Hilgers found SWERA DNI being 300 kWh per m² and year lower than the DLR database. Verified by Hilgers (2010) through additional studies, aerosols, which convert direct irradiation to diffuse irradiation, have been measured inaccurately in this region by the satellites sourcing data for the SWERA database. Nonetheless, since solar-cells convert both, direct and diffuse irradiance, the undervaluation in this region is regarded of minor influence.

5.2.3 Area for Building Integrated Installation

The assessment of available area for facade and rooftop installation of PV is developed according to the top-down approach utilised by Czisch (2005) and Hoogwijk (2004) and picked up by Held (2010) as well. It is assumed that the number of buildings and the surface provided by them increases linearly with population. Hoogwijk (2004) further analysed the effect of increasing GDP on rooftop areas. Nonetheless she adopted a lump sum factor of 0.11 % of the global terrestrial surface. In this study the approach from Held is adopted.

Calculation of Suitable Area

Estimation factors applied by Held (2010) are based on empirical values developed by the International Energy Agency assuming 18 m² roof area and 6.5 m² facade area per capita. Usability factors of 40 % and 15 %, for rooftop and facade areas respectively, and the incorporation of population values for 2012, 2030 and 2050 (diagram given in Figure 3; Chapter 7, values listed in Table (Anex.) 2) lead to the suitable rooftop and facade area which is assumed to be fully covered with PV modules. Results are illustrated in Table 12.

Compared to Czisch (2005), who assumes roof area per capita as in Germany and calculates a total roof area of 1344 km² for North Africa, this study equates similar results of 1213 km² roof area today, 1513 km² in 2030 and 1717 km² in 2050. According to Hoogwijk (2004), North Africa has 570 km² of rooftop area suitable for PV installations. This value is estimated according to country-specific GDP.

Table 12: Suitable and fully covered rooftop and facade area

illustration not visible in this excerpt

Resulting Irradiation on Rooftop and Facade Installations

Buildings and settlements are not uniformly distributed in North Africa, with a much higher density in coastal, northern areas. As explained in Section 5.2.1 every local polygon pi classified as “Settlements, Industrial Buildings or Airports” is associated with a certain size [Abbildung in dieser Leseprobe nicht enthalten] The local irradiation within this polygon on modules installed on rooftops [Abbildung in dieser Leseprobe nicht enthalten] or [Abbildung in dieser Leseprobe nicht enthalten] facades [Abbildung in dieser Leseprobe nicht enthalten] is respectively obtained by Equation 21.

illustration not visible in this excerpt

Equation 21: Irradiation on rooftops and facades

illustration not visible in this excerpt

This method takes into account irradiation on PV modules according to building distribution. The inclusion of airports and industrial buildings is consistent with IEA’s empirical values, which furthermore include agricultural and “other” buildings [281].

Table 13: Annual average irradiance on suitable rooftops and facade areas

illustration not visible in this excerpt

As expected, results depicted in Table 13 reveal vertically installed PV modules on facades to receive much less solar energy. Due to its northerly location Tunisia encounters least irradiance on tilted roof top modules. At the same time Tunisia’s latitude position qualifies for better irradiance on facades than Western Sahara, of which the territory is situated at the more southern Tropic of Capricorn experiencing the phenomena described in Section 5.1.2.

5.3 Technical Potential of Photovoltaic Energy

Section 3.2.1 pointed out, that PV technology is physically limited in the conversion of light to electrical energy. Furthermore, solar cells, modules and PV systems as a whole encounter losses. The gap between received insolation and AC system output is analysed in the following section.

5.3.1 Performance of PV system

Besides the PV system components efficiency, the overall performance ratio is influenced by environmental factors, which can be especially harsh in North Africa, due to its desert climate.

Module Efficiency

For average module efficiency calculation, the portfolio of 2009’s five biggest PV manufacturers as of actual yearly production [282] is analysed with today’s best serial products by technology taken into consideration. Furthermore, mono and multi c-Si as well as thin film technology are considered at the 2011 market share of 40 %, 40 % and 20 % respectively (see Section 3.2.1). With average best module efficiencies of 15.475 % (mono c-Si), 15.025 % (multi c-Si) and 12.1 % (thin-film) the average best market efficiency is set at 14.62 %. Product details by company are depicted in Table (Anex.) 10. Average module efficiencies by technology are consistent with values published within IEA’s PV Technology Roadmap 2010 [283]. By just considering today’s best products, this approach takes into consideration the short-term efficiency improvements of the three technologies, as depicted in Figure 34. Due to better economics and improving efficiencies of thin-film products, the forecasted boost in market share of thin-film PV modules will offset module efficiency gains made by c- Si technologies until the midterm future [284]. Compared to a status-quo in market distribution by technology, with increasing thin-film market share the average module costs will encounter greater reductions and average module efficiencies will not alter greatly in mid- term future.

According to Hoogwijk (2004), module efficiencies for c-Si PV modules stood at 12 to 16 % in 2004. While in 2004 the PV market was governed almost entirely by c-Si technology, she further assumes module efficiencies to rise to 15-20 % in the midterm future and 30 % on the long term [285]. These much higher efficiencies are not adopted within the analysis in hand, since market data as of today depicts a much lower increase in average module efficiency, with forecasted tendencies as stated above.

illustration not visible in this excerpt

Figure 34: Development in module efficiency of PV technologies starting in 2010 [286] Temperature Sensitivity

Temperature sensitivity “is usually the most important performance loss.”[287] Since North Africa is characterised by the hottest climates on earth, a thorough analysis of temperature influence on solar cell performance is crucial. However, prior to the evaluation some technical explanation is needed.

The performance of PV modules is measured at Standard Operating Condition (STC) with 1000 W/m2 irradiance, 25°C module temperature and AM 1.5 spectrum. Solar cells are inversely and linearly temperature sensitive with the semiconductor’s performance being impaired by a decreasing band gap and increasing recombination. While the short-circuit current ISC is relatively unaffected, the open-circuit potential VOC declines linearly with rising temperature. Thermal insulation by the encapsulation of the module boosts cell temperature much over ambient temperature. In order to compare outdoor performance, the Nominal Operating Cell Temperature (NOCT) is specified as the cell temperature during open-circuit module operation at ambient temperature of 20°C, 1 m/s wind speed and 800 W/m² irradiance. Resulting cell temperature under varying irradiance and ambient temperature is obtained by Equation 22.

illustration not visible in this excerpt

Equation 22: Calculating the operation temperature of the solar-cell [288]

PV modules typically have a power output temperature coefficient (Ȗ) of -0.5 %/°K translating to a reduction in 0.5 % of efficiency for each degree °C the cell temperature rises above STC temperature of 25°C. The temperature effect on module efficiency is calculated by Equation 23.

illustration not visible in this excerpt

Equation 23: Influence of temperature sensitivity on module efficiency [289]

Due to the characteristics of deployed solar cell materials, PV technologies perform differently in hot climates:

The mono-crystalline product “QPeak 245-265Wp” by PV manufacturer Q-Cells [290] - currently the best module efficiency in the Q-Cells portfolio with 15.9 % module efficiency measured at STC (Ȗ = -0.46 %/°C; NOCT = 47±3°C ) - achieves at ambient temperature Tambient = 40°C and irradiation G = 1kW/m² an actual efficiency of Ș(Tcell,G) = 12.3 % at an actual cell temperature Tcell = 73.73°C. With this hot climate only 77.6% of the nameplate efficiency Ș(STC) is reached. Incrementing Tambient by 5°C to 45°C would change Tcell to 83.75°C with Ș(Tcell,G) lowered to just 11.6 % module efficiency and 73 % of nominal power output. The assumed ambient temperature and irradiation represent typical summer values in the Sahara [291].

Q-Cells’ premium thin-film product “QSmart UF L 95-115” holds STC module efficiency of 12.2 %, Ȗ = -0.38±0.04 %/°K and NOCT = 51±2°C. Under the same environmental conditions as above, the efficiency of this thin-film module drops to 80 % of Ș(STC) at Tambient = 40°C and to 78 % of Ș(STC) thus a resulting module efficiency of 9.5 % at Tambient = 45°C.

Figure 35 discloses the tmperature sensitivity of the evaluated thin-film module compared to the mono-c-Si module by illustrating the decreasing percentage of Ș(Tambient,G=1) / Ș(STC) against an increasing ambient temperature with values of the above studied PV modules.

In relation to their nominal module efficiency the thin-film module operates significantly better at hot climatic conditions while the multi-c-Si module operates marginally better in cold climates.

illustration not visible in this excerpt

Figure 35: Temperature sensitivity of mono-cSi and thin-film PV modules

These calculations outline the counteracting impact of increasing ambient temperature and irradiance on module efficiency. For a complete assessment, a range of climatic factors like wind speed, ambient humidity and type of module installation are necessary to further evaluate the performance [292] [293].

According to the CPV Consortium, CPV technology operates best at hot climates with 96 % of nominal power output compared to 80 % for multi-Si, 84 % for mono-Si and 89 % for thinfilm solar cells. All technologies are evaluated at 40°C ambient temperature [294].

Lifetime Reduction in Desert Environments

While industrial scale PV industry is relatively new and PV modules are built for 20+ years of operation, little is known about the end of life of PV modules. Most common causes for failures are moisture penetration, which is of little importance in North Africa, and temperature fluctuation, which is very high in desert environments.

Suleske (2010) investigated on 1865 grid-connected mono and multi c-Si PV modules aged between 10 and 17 years and installed in the Arizona desert. IR scanning revealed hotspots at 2.2 % of the modules. These modules were having cells with a temperature difference greater than 5°C compared to the remaining cells in the rest of the module. Hot spot formation occurs when local shading or failure reduces the local cell short-circuit current. In series string, the affected cell dissipates the backed up current as heat energy instead of producing power, resulting in drastic reduction of module power output. Large sized modules (250-300 W compared to 50-75 W) with glass front and back are more affected by thermal stresses and are showing a higher concentration of hotspots due to higher stresses on cell components [295].

Browning of the module centre was most common with 89.1 % followed by delamination of the encapsulant with 3.4 % of the PV modules. Modules with glass backsheets were less affected of the latter than those with polymer backsheets, which are used in modern modules. Consequently the power output of modules not showing any hot spots decreases at rates between -0.93 and -1.92 % per year, with 17 year old modules experiencing a 33 % reduction in power output. Modules connected in series, as common in large utility scale arrays, are more prone to high degradation rates. Here, high voltage and mismatch of module outputs in the module string result in current leakage causing cell corrosion.

Cells with danger for safety due to fire hazards, burns, broken glass or broken interconnects where not observed confirming a good safety record of PV products.[296]

Further Losses and Conclusion on Performance Evaluation

When in early days the inverters accounted for much trouble regarding reliability and ACoutput quality [297], today in that sense they are a negligible issue. For 2011, the US Department of energy estimated a conversion efficiency of 96 % [298]. Technological leader SMA reaches 99 % maximum efficiency with standard products [299].

In 2005, comparing global regions, Czisch (2005) calculated a worldwide performance ratio of 82 % with maxima in high mountain Antarctica (99 %), minima in the southern Sahara (73 %) and a close-to-average value of 81 % in Germany. The availability was set at 95 % in this study taking into account technical improvements but also increasing shutdown frequency at the end of the system’s lifetime [300]. Both, Held (2010) and Hoogwijk (2004), applied an overall efficiency ratio of 75 % comprising losses due to the inverter, mismatch, shading and cable conductivity.

A degression by -0.30 % per year in PV system AC power output is used in the profitability calculation by Kost and Schlegl (2010). In this study for simplicity reasons a calculated performance degression of system lifetime is not included.

The assumption of the performance ratio takes into consideration a market shifting towards thin-film technologies. Thin-film technologies appear as better suited regarding temperature sensitivity impact under North African environment stated by the CPV Consortium [301]. System performance reductions over lifetime are considered within the lump sum performance ratio of 80 % set in this study. Together with the calculated module efficiency of 14.62 % an overall system efficiency of 11.7 % is obtained.

5.3.2 Results for Technical Potential of PV

While the module efficiency is rated at STC condition, the installed capacity of a module is directly obtained by multiplying its efficiency by 1kW/m² resulting in 146.2 Watt per m² nominal capacity for the reference module used in this study.

Full Load Hours

The amount of hours per year a system operates at its installed capacity, called Full Load Hours (FLH), is calculated with Equation 24.

illustration not visible in this excerpt

Equation 24: Annual capacity usage (full load hours) of a PV system

Installed Capacity

Furthermore, the installed capacity of PV ground systems is obtained by Equation 25.

illustration not visible in this excerpt

Equation 25: Installed capacity of ground installed PV systems

The installed capacity of roof and facade systems is calculated with Equation 26.

illustration not visible in this excerpt

Equation 26: Installed capacity of rooftop and facade installed PV systems

Annual Energy Generation Potential - the Technical Potential

Thus, the annual AC power output, thereby the technical potential of the various installation types t is obtained by Equation 27.

illustration not visible in this excerpt

Equation 27: Annual energy output of the PV system - technical potential

By country-wise cumulating the annual power output and cumulative installed power, the values depicted in Table 14 are obtained.

Table 14: Installed capacity, annual AC power output and average annual FLH by country and installation type in the year 2012

illustration not visible in this excerpt

Taking up forecasted electricity demands of Chapter 2, Table 14 reveals that solely the technical potential of ground installed PV in Egypt (4,691 TWh per year) more than covers the demand of Europe in 2030 (4,560 TWh per year). Egypt furthermore possesses by far the highest potential in rooftop installed PV. This is reasoned in Egypt’s largest population of all North African countries and at the same time the highest yearly insolation on latitude tilted surfaces of all North African countries. Conjointly, both factors translate to best average Full Loads Hours and by far the highest electricity generation potential of PV roof systems in North Africa. Algeria, with a vast potential of 10,926 TWh per year generated by ground installed PV could cover the demand of Europe and North Africa in 2030 (5207 TWh/a) twice.

Continuing the analysis with the assessment of the economic potential of PV, the costs associated to the potentials depicted above will be evaluated.

Comparison with Similar Studies

Hoogwijk (2004) states the technical PV generation potential of centralised, ground installed systems with 49000 TWh/a in North Africa. Rooftop installations bear a technical potential of 100 TWh/a in her analysis [302].

Scholz (2010) analyses the technical potentials of PV in a much more extended area encompassing Europe and North Africa. Since she assumes a usability of 100% on bare land and desert compared to much lower usability factors for other land use types, most of the technical potential is situated in North Africa. Furthermore, one third of this suitable area is covered by PV generating about 30,000 TWh annually.

Czisch (2005) finds a potential of 288 TWh/a for rooftop PV systems in North Africa. Consistent with the study in hand, Egypt holds the highest potential of 151 TWh/a in his analysis.

Kost and Schlegl state maximum capacity usage of PV ground systems in North Africa with 2000 FLH [303] ; very close to the value of 2050 FLH calculated in the analysis in hand.

Within her analysis on potentials for rooftop and facade mounted PV in the EU, Held equates capacity usage up to 900 FLH for facade-mounted systems and 1500 FLH for roof-mounted PV systems in southern Spain and France [304]. While irradiation on tilted surfaces in North Africa is higher, vertical systems receive less irradiation compared to more northerly latitude positions in Europe.

5.4 Economic Potential of Photovoltaic Energy

As Section 3.2 points out, the PV market can be considered as very dynamic including rapid technological development, boosting sales volume as well as expansion in production capacities, and a consolidation to form global players out of a still very fragmented market with many PV manufacturers. The purpose of this section is to evaluate the factors influencing the economies of PV systems over their lifetime. Combined with PV electricity generation potentials, a full picture of the LCOE of PV installation types by geographical distribution shall be developed.

5.4.1 Financial Parameters

Following the approach of various other studies the lifetime of the PV system is set to 20 years. While this assumption might be antiquated and only true to old module types as analysed in Section 5.3.1, it is still common practice [305] [306] [307] [308]. Furthermore, since a shift towards thin-film technologies, having shorter product lifecycles than c-Si modules [309], is observed, the average life-time expectancy is assumed to be stagnating at this value.

The Weighted Average of Capital Costs (WACC) is rounded up to 8 % with a capital distribution of 30 % equity, with 10 % return on equity, and 70 % debt, with an interest rate of 7 %.

Investment and O&M

Subsequent to the evaluation of current market prices and various studies for PV systems of varying sizes and types - with higher priorities on more recent studies, the initial investment and values associated to operation and maintenance (O&M) for the PV systems are set according to values illustrated in Table 15. The annuity is calculated by multiplying the Net Present Value NPV with the annuity factor a obtained by Equation 28.

illustration not visible in this excerpt

Equation 28: The annuity factor [310]

The initial investment (Net Present Value - NPV) per kWp of installed capacity of the PV system includes the PV modules and the cost associated to Balance of System (BOS), which comprises inverters, mounting rigs, cables, installation and commissioning of the PV system. The annual cost of O&M is expressed as the percentage CO&M of the initial investment and includes parts replacement and maintenance costs, e.g. surface cleaning and insurance, but is exclusive of any land or roof rental costs. While PV systems are regarded as of high reliability and low susceptance to failure [311], low costs for operation personnel are assumed and included in CO&M as well. All costs are exclusive of transport, grid connection and tax. The influence of the PV installation site’s distance to the electricity grid and transport infrastructure is evaluated in Chapter 6.

Table 15: Financial input parameters for the LCOE calculation of PV installation types

Abbildung in dieser Leseprobe nicht enthalten [312]

Analysed system sizes [kWp] 50-10000 2-500 50

For a detailed overview of the values found in previous studies and at current market prices please refer to Table (Anex.) 11. Analysed system sizes vary greatly. This is due to the inclusion of both residentially and utility scale roof mounted systems on large buildings. The same occurs with ground installed PV, where values for relatively small systems of just 50 kWp up until 10 MWp PV parks are included.

Calculating the Levelised Cost of Electricity (LCOE) of PV Systems

The specific leveraged cost of energy (LCOE) for PV energy generation according to installation type t is calculated by Equation 29.

illustration not visible in this excerpt

Equation 29: Cost-supply calculation for PV energy depending on installation type [313]

5.4.2 Cost Development of Photovoltaic Energy

As depicted in Section 5.3.1, though average module efficiencies are assumed not to be changing considerably over the mid-term future, PV system costs as of € per kWp are ought to decline drastically. Experience curves forecast unit prices in relation to an increasing accumulated production [314]. This assumes that by a doubling in production volumes, learn effects, related to manufacturing and R&D, will lead to unit price reductions. In the context of PV module price and investment forecasting, the cumulative installed global PV capacity figures as the cumulative production. The forecasted initial investment is obtained by Equation 30.

illustration not visible in this excerpt

Equation 30: Forecasted PV system costs employing the learning curve [315]

Nemet analyses 156 learning curves for PV modules and finds learning rates of 17 to 24 % between the 5th and the 95th percentile peaking at 20 % [316]. Consequently, the study in hand adopts a learning rate of 20 % to forecast the initial investment of the three types of installation in 2030 and 2050. Kost and Schlegl (2010) assume a learning rate of 20 % until 2015 followed by a learning rate of 15 % until 2030.

illustration not visible in this excerpt

Figure 36: Forecasted initial investments for PV systems, LR=20%

Forecasts on globally installed PV capacity are obtained through data provided by IEA’s PV “Technology Roadmap” and are refined with recent values [317]. Compared to studies by Greenpeace/EPIA and the “ETP Blue Map” scenario, the forecast of installed capacity stated in IEA’s “Technology Roadmap” is balanced between both extremes [318]. Increasing cumulative installed capacity and decreasing PV system’s initial investment respective to PV installation type are depicted in Figure 36. By these input values, initial investment in 2030 will be reduced to 38.9 % of today’s investment and dropping to 26.4 % by 2050. In comparison, Kost and Schlegel forecast initial investments in 2030 to stand at a level of 33 % compared to today’s initial investment.

The cost of annual O&M, as the percentage of initial investment for new PV installations, is assumed to be declining at a learning rate of 10 %. This is consistent with decreasing O&M percentages forecasted by the U.S. Department of Energy. However, this study does not take into consideration increasing annual O&M costs over the PV system’s lifetime, as do Kost and Schlegl (2010)

Resulting financial parameters for 2030 and 2050 in comparison to today’s values are illustrated in Table 16. Average module efficiencies are stagnating at today’s values in this scenario.

Table 16: Financial parameters to calculate LCOE of 2012, 2030 and 2050

illustration not visible in this excerpt

5.4.3 Resulting Cost-Supply Curves for Ground Mounted PV

The mapped potentials in Figure (Anex.) 12 reveal highest yielding sites located in a broad band following the Tropic of Cancer. The band extends from south-eastern Algeria close to the town of Tamanrasset, across the entire south of Libya until south-western Egypt. With just a few exceptions, the entire territory between the river Nile and the Gulf of Suez appears to hold low LCOE for PV ground systems. Best sites in Morocco are situated south of the Atlas Mountains extending towards nearby Western Sahara. Compared to other North African countries, Tunisia has only few high potential sites located at its southern tip.

Figure 37: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2012

illustration not visible in this excerpt

Country-specific Cost-Supply curves for the 2012 scenario, regarding PV ground systems, are illustrated in Figure 37. The illustration of maxima and minima LCOE reveals Libya and Egypt achieving lowest LCOE. Mapping LCOE, as illustrated in Figure (Anex.) 12, furthermore puts into context the abrupt upswing of the Cost-Supply curves just before reaching the maximum country-specific generation potential: Sites with highest costs of electricity generation are located along the Mediterranean coast and at the southern Moroccan Atlantic coast. Reasons for this are most likely to be found in increased cloud coverage, though a more detailed climatic evaluation would be required here. For 2012, North Africa holds a potential of 6749 TWh/a at marginal generation costs of 14 ct€/kWh. For today, in Egypt a generation potential of 2013 TWh/a - about 16 times the Egyptian electricity demand in 2009 [319] and 12 times the demand of 2020 - is achievable by marginal LCOE of 14 ct€/kWh.

The forecast of LCOE for PV ground systems, as depicted for North Africa in Figure 38, is completed with linearly interpolated LCOE levels. Colours indicate the range of LCOE a specific generation potential is situated in at a specific time. Starting in 2012, the cost of energy slopes steeply in the midterm future and decreases at smaller rates in the long term forecast between 2030 and 2050. The minimum LCOE, at highest yielding sites in Egypt and Libya, decreases from 13.4 ct€/kWh to 4.9 ct€/kWh in 2030 and 3.2 ct€/kWh in 2050. When in 2012 the majority of potentials are situated at LCOE between 14 and 16 ct€/kWh, LCOE for almost all sites are reduced to 2-4 ct€/kWh by 2050.

Figure 38: Cost-Supply development for energy generated by PV ground installed systems in North Africa until 2050

The forecasts for 2030 states a North African generation potential of 27271 TWh/a, almost the whole potential of 28380 TWh/a, at marginal costs of 6 ct€/kWh. In the 2050 scenario the entire North African PV ground system generation potential is situated below 6 ct€/kWh.

Country specific ground-PV Cost-Supply curves for 2030 and 2050 are provided in Figure (Anex.) 9 and Figure (Anex.) 10 respectively. In the 2030 scenario, almost the total Egyptian generation potential is already situated below the cost threshold of 6ct€/kWh.

Figure (Anex.) 13 and Figure (Anex.) 14 map the geographical distribution of Levelised Cost of Electricity for PV ground systems in North Africa for 2030 and 2050 respectively.

Comparison with Similar Studies

With financial parameters of 2010 and for sites receiving an annual irradiance of 2000 kWh/(a m²), Kost and Schlegl (2010) equate LCOE between 15 and 17 ct€/kWh, depending on module price. Locations with identical insolation values generate power at an LCOE of 17.5 ct€/kWh in the scenario for 2012 within the analysis in hand. Best sites receive a much higher insolation of 2610 kWh/(a m²).

Consistent with forecasted LCOE for ground-PV in the assessment in hand, where best sites in 2050 accomplish LCOE of only 3.3 ct€/kWh, Scholz (2010) calculates lowest LCOE of 3.3 ct€/kWh in 2050 for most competitive North African sites.

Cost Competitiveness of PV Ground Systems versus WT systems

The assessment for 2012 reveals LCOE of PV ground systems at most favourable sites being 10.5 ct€/kWh higher than wind LCOE at highest yielding sites. This gap will be reduced to 2.8 ct€/kWh by 2030. Minimum LCOE of PV ground systems and wind power will even out in 2050.

Comparing the Cost-Supply curve’s inclination of ground-PV with Cost Supply curves of wind power, the PV curve’s slope is found to be minor. For the entire assessment of ground-PV, the global atlas SWERA, with a resolution of 1°, is employed. Levelling out areas of high and low radiation within the same raster tile, the spread between minimum and maximum LCOE is reduced. However, solar irradiation is considered less site-specific than wind speeds since topology and land use have lower effects on the potential’s magnitude.

In conclusion, the assessment of 2012 reveals that electricity produced by PV ground systems is not competitive to wind power as of today. Nonetheless, when keeping up learning rates of 80 %, PV ground systems will reach competitiveness in the long term future. High yielding PV sites are not as confined as high yielding WT sites. Consequently, by 2030 the cumulative generation potential achievable at low LCOE is much higher with PV ground systems than with WT systems. PV’s lower relative costs in O&M play in favourably as well.

Considering thresholds of 6 ct€/kWh with both technologies in the 2030 scenario, the generation potential of PV is twice as high as the generation potential of wind power.

Applying a more stringent LCOE threshold of 4 ct€/kWh for the 2050 scenario, ground-PV potentials (27097 TWh/a) outperform WT potentials (4405 TWh/a) by a factor of 6.

So far this competitiveness analysis is solely performed from an economic perspective. Capacity usage of PV systems at best sites is of 2088 FLH; considerably lower than the capacity usage of 5000 FLH for WT systems at highest yielding wind power sites. Consequences regarding electricity supply security as well as possibilities for capacity credit enhancement are further described in Section 7.2.

According to Hau (2008), PV parks require a factor 100 to 1000 more land to reach a specific power output than win turbines [320].

5.4.4 Cost-Supply Curve for PV Systems Installed on Buildings

Due to a higher initial investment, minimal LCOE for 2012 are 3.4 ct€/kWh higher for roof- installed PV compared to ground-mounted PV. Steep Cost-Supply curves, country-wise illustrated for 2012 in Figure 39, indicate few buildings being situated in zones of exceptionally high irradiation. Egypt is an exception since it generally receives high levels of irradiation in almost the entire country. Consequently, Egypt could generate a potential of 68 TWh/a at marginal costs of 17 ct€/kWh, this equals to 40 % of the Egyptian electricity demand forecast for 2020, thereby double of the targeted 20 % in RES supply.

Figure 39: Cost-Supply curve depicting marginal LCOE for PV rooftop installations with financial parameters for 2012

illustration not visible in this excerpt

In Figure 40, Cost-Supply curves for North Africa depict the forecasted steep decline[321] in electricity generation costs for roof-mounted PV systems. While the potential in cumulative power output is directly linked to rising population figures, the generation potential increases simultaneously.

In the Egyptian case, 96 TWh/a - thus 46 % of the entire Egyptian generation potential of 2030 - will be at marginal LCOE of 6 ct€/kWh in 2030. For 2050, the entire Egyptian potential of 249 TWh/a will have generation costs lower than 5 ct€/kWh, equal to 39 % of the forecasted Egyptian electricity demand in 2050.

Figure 40: Cost-Supply curves for PV rooftop installations from today to 2050

From a cost perspective, PV systems integrated or mounted on facades are not competitive with the power mix LCOE of today and of 2030 [322]. Due to less irradiation on vertical surfaces, the capacity usage of PV systems installed on facades is much lower than roof- mounted systems. Coupled with 80 % higher initial investment for facade systems, the resulting minimal LCOE of facade systems is almost four times the minimal LCOE of roof systems as of today. While learning rates are assumed to be equal for both installation types, this relative difference in LCOE remains present throughout the forecasted scenarios. By the LCOE and generation potential forecasts illustrated in Figure 41, it can hardly be expected, that facade installed PV systems contribute with a significant proportion to the future energy supply. The North African electricity demand is forecasted to be increasing from 208 TWh/a in 2010 to 647 TWh/a in 2030 and 1225 TWh/a in 2050 [323]. Facade-PV-electricity could supply 1.4 % of the North African demand in 2050 at LCOE between 13.9 ct€/kWh and 23.6 ct€/kWh.

illustration not visible in this excerpt

Figure 41: Cost-Supply curves for PV facade installations from today to 2050

In contrast to the size of the countries, the scattered settlement areas are rather small. Due to resolution constraints, maps depicting the distribution of economic potentials of PV systems installed on buildings are not included in this composition of results.

According to the phenomena described in 5.1.2, the cost of energy produced by PV systems mounted on facades increases with proximity to the Tropic of Cancer and rises further south of the Tropic. In this area, due to the zenith of 0° in summer months, vertical surfaces receive low levels of annual irradiation.

Comparison with Similar Studies

Kost and Schlegl (2010) state LCOE of small and large roof mounted systems at best North Africa sites with 18.4-20.4 ct€/kWh and 17-19 ct€/kWh respectively. This is at least 1ct€/kWh higher than in the assessment in hand. Though initial investment in the evaluation by Kost and Schlegl (2010) is on average 200 €/kWp more expensive, the discrepancy in resulting LCOE is reduced by lower WACC of 6.5 %, compared to 8 % WACC in the analysis in hand. Czisch (2005) adopts roof-mounted PV system costs of 5500 €/kWp and an interest rate of 5 %. The globally lowest LCOE of 27 ct€/kWh in his analysis is achieved by sites in Saudi Arabia. No default module efficiencies are stated in both studies.

Assuming module costs of 5034 €/kWp, 10 % interest and a default module efficiency of 14 %, Hoogwijk (2004) evaluates lowest LCOE for roof installed PV in North Africa with 46 ct€/kWh. These results are much higher than LCOE stated by Czisch (2005), suggesting that Czisch adopted a better default module efficiency than Hoogwijk (2004).

In 2007 the U.S. Department of Energy stated targets of 9.3-12.9 ct€/kWh for residentially roof mounted PV systems in 2011 and a further reduction in LCOE to 5.7-7.2 ct€/kWh by 2020 [324]. These values are about 5 ct€/kWh lower than LCOE of today and in 2020 respectively.

LCOE calculated by Held (2010) for generation potentials in the EU range from 60.9 ct€/kWh to 121.8 ct€/kWh for facade mounted and 21-47.4 ct€/kWh for roof-mounted PV systems. Taking into consideration the higher irradiation levels in North Africa and decreased PV system costs since 2010, Held’s minimal LCOE values are consistent with the LCOE obtained in the study on hand [325].

5.5 Discussion of the Results

Considering the cost of electricity generated by PV of the various installation types, photovoltaic energy is neither competitive to the power mix LCOE, stated by Kost and Schlegl (2010) with 6 ct€/kWh, nor to the current upper end electricity price in Egypt of 9 ct€/kWh [326]. With the expected growth of the power mix to LCOE to 10 ct€/kWh by 2030, best sited PV ground systems will reach grid parity between 2020 and 2025. While module efficiencies are assumed to be stagnating, increasing average module efficiencies would speed up LCOE degression. On the other hand, increasing commodity prices would slow down cost improvements. Similar to wind power, which experiences learning rates of 90 % [327] or even 97 % [328], learning rates for PV technologies will most likely not remain stable but worsen as well.

The competitive edge of wind power over PV energy will diminish in the midterm future and even out in 2030 with lowest LCOE of both technologies achieving parity by 2050. Since irradiation levels are similar over extended areas, siting is less complex with PV systems regarding optimal irradiation exploitation.

Though rooftop installed PV systems are of higher LCOE than ground mounted systems “on a societal level, this arrangement may still be more efficient than having utilities install PV plants, because homeowners generally have much lower revenue expectations” [329]. Within this assessment, valuable effects in regard to building integration of PV systems are not taken into account. However, smart usage of PV systems in architecture can shade rooftops, facades and windows, thereby reducing the heat intake of buildings [330]. Furthermore, construction material can be reduced by the incorporation of PV systems into roof or facade structures. With module efficiencies of 4.6 to 5.6 %, the multifunctional glass produced by Japanese Taiyo Kogyo Corporation generates electricity, provides a heat shield, blocks UV radiation and enables light transmittance by 10 % [331]. Nonetheless, including these benefits into the analysis on a monetary basis is not within the assessment’s scope.

Stand-alone systems with adjacent energy storage for self-sufficient energy supply in remote areas are not considered in this analysis, although by information depicted in 2.1, these islanded systems are the most common application of PV systems in North Africa today.

Electrification rates of households in North Africa are already high [332]. Consequently, distance to grid is of low concern for building integrated grid-connected PV, since the generated electricity is assumed to be fed into the grid by genuine building connection line. For ground mounted PV, geographic constraints for remotely located sites will be analysed in the following chapter.

6 Inclusion of Infrastructure into the Assessment

6.1 Background and Objective

The detailed economic assessments of Chapter 4 and 5 reveal North Africa holding enormous generation potentials in both wind and photovoltaic energy: Generation potentials for PV add up to 28,379 TWh/a in ground installed, 314 TWh/a in roof installed and 21 TWh/a in facade installed PV as of today. Depending on hub height and turbine type, between 18,363 and 25,973 TWh/a could be generated by wind power. However, these potentials are geographically scattered and may not be easy to tap. The effect of remoteness on LCOE of wind and PV energy will be evaluated in this chapter.

While electrification rates for households are already high in all countries besides Morocco [333], the current slow paced rollout of building integrated PV is assumed to be in line with the existing electricity grid infrastructure. Grid enforcement due to building integrated PV may be required in the future.

PV ground systems and WT installations require grid connection from sites of energy generation towards a feed-in point. When realising very large scale wind or PV parks at high yielding sites, grid connection over long distances is reduced to a reasonable cost per kWp of installed capacity and kWh of generated energy. Regardless, the current electricity grid would have to be greatly up-scaled in capacity and extent in case of an increase in PV or wind energy installation with feed-ins in the multi TWh/a range [334].

Generally speaking, a wind or PV park only requires one transport access and one grid connection, regardless of the installed capacity of the park. The trade-off between costs of infrastructure development over each unit of energy generated and benefits due to better yielding, remote sites becomes more favourable for the latter with increasing park size. However, for the slow paced rollout, as observed in most North African countries, existing infrastructure for systems transport and electricity transmission turns into an interesting aspect in siting new projects of small or medium sized PV or wind parks.

This chapter will in a first step analyse the investment for grid connection by distance. Subsequently, spots which are currently available at low LCOE including costs for grid connection will be highlighted. Information on distance to existing and operational transport infrastructure, thereby roads and railways, will not be processed monetarily but included in the resulting maps.

For PV and Wind energy generation in the North African region, existing infrastructure has not been considered in previous studies and is consequently a novelty in the Cost-Supply assessment for PV and wind energy in this region.

6.2 Electricity Grid Infrastructure

6.2.1 The North African Power Grid

Interconnectors between all North African countries are well established with harmonisations intentions between national power networks underway. Since 2010, Morocco is connected and synchronised with the European electricity grid through an HVAC link via the strait of Gibraltar. Further Maghreb-Europe interconnectors are planned for 2020 with a HVDC link between Italy and Tunisia and further links between Algeria and Spain in 2020 and Algeria and Corsica in 2025 [335].

Figure 42: Map of the current electricity grid in North Africa; lines ≥ 225 kV only [336]

illustration not visible in this excerpt

Figure 42 depicts the North African electricity grid as of 2012. The base map provided by the 2010 grid map of the Arab Union of Producers, Transporters and Distributors of Electricity [337] (AUPTDE) is completed by AUPTDE’s map of 2001 [338] and augmented by national grid maps for Tunisia [339] and Algeria [340], including stretches currently under construction and to be completed in 2012. Sourced maps do not provide data in high spatial resolution, consequently the minimal classes for distance to electricity lines is set with [0; 50 km], followed by (50; 100 km], (100; 250 km], (250; 500 km] and distance to grid above 500 km.

Solely electricity lines of the high to very high voltage range with • 225 kV are included, thereby enhancing the probability of the grid being able to cope with additional power fed-in by potential WT or PV parks. Considering the proximity to existing substations would further refine the results but is not included in the assessment in hand. The depicted substations are of all voltage levels and are merely for illustration purposes.

Grids of southern neighbours are considered as well but do not alter the map, because they are a direct continuation of the considered grid or are of lower voltage levels than the considered minimal level of 225 kV.

6.2.2 Cost of Grid Connection

To cover the costs of the grid infrastructure and operation, the transmission grid owner receives regulated yearly revenues defined by tariffs (UoS - Use of System), which are charged for each unit of electricity either generated or demanded, depending on the prevalent scheme [341]. Moroccan grid tariffs as of 2012 are 8cDH/kWh [342], equal to 0.71 ct€/kWh [343], Egyptian wheeling charges for high voltage UoS are about 1 ct€/kWh [344] [345].

The schematic outlay for the connection of a 50 MW wind park to an existing grid with 100 MW capacity provided by a coal fired power plant and 50 MW provided by a Natural Gas Combined Cycle power plant is shown in Figure 43. Considering large scale wind parks, Elsobki states that a 220kV grid can handle up to 1500 MW of connected WT capacity [346].

Figure 43: Sketch of electricity transmission and generation infrastructure

With conditions corresponding to reality, transmission grids are not operating at their full capacity. The costs associated with new installations of grid connections by distance and wind parks size are analysed by the following case studies.

First Case Study

Marigal and Stoft calculate costs of grid connection per kWh (prices of 2006, 5% discount rate, 20 year life time) based on a 50 MW wind farm with 30% capacity usage requiring a 115 kV line. No other generators than the wind park are connected to the line. Figure 43 depicts a sketch of this situation. Considering the cost for transmission line of 220,000 USD/km an LCOE of 9.66 ct€ per kWh and 1000 km distance is calculated [347].

Including all transmission system cost elements, e.g. substations, line, construction, the total LCOE of the grid connected wind park is 9.6 ctUSD per kWh, thereof 1.6 ctUSD per kWh accounting for the complete transmission system [348] bridging a distance of 100 km. Converted to Euro the value of 11.51 ct€ per kWh [349] transmitted over 1000 km is recalculated by the financial parameters of this project to grid infrastructure investment of 188,480 € per km distance and an annuity of 15,124 € per km.

Second Case Study

The wind parks around Jabal el-Zayat at the Egyptian Gulf of Suez are planned to be online by 2026 or 2027 with a targeted overall capacity of 2000-3000MW on an area of 625 km².

Additional to the existing 220 kV line, a 500 kV line has to be constructed by 2015 to evacuate electricity generated by 2290 MW [350] installed WT capacity. The overall cost of transmission infrastructure and technical assistance is 344.9 million USD, not including O&M costs [351]. The 280 km transmission system will be connecting the augmented substation at Samallout, west of the Nile, with the new substation at the Gulf of Suez. Besides small deviation to bypass settlements, 89 % of the stretch is passing uninhabited and uncultivated desert land [352], thus very similar to conditions prevailing throughout North Africa.

The annuity equates to 71,078 € per km [353], assuming 20 year life time and 5 % discount rate. The adopted WT capacity ratio of 30 % leads to 2628 FLH and a yearly wind park output of 6.018 TWh/a. Conjointly, 1.18 ct€/(kWh 1000 km) is calculated.

Conclusion and Results

The resulting surcharges per kWh in both examples differ greatly. Economies of scale greatly reduce the cost per MWp connected. Furthermore, with rising number of WTs in a wind park, fluctuations in energy fed into the grid decrease, in particular when generators are spread over a wider area [354]. As a result, required capacity of the grid connection can be reduced thereby yielding higher FLH of the transmission system adjacent to large wind parks compared to small wind parks. By statements depicted in Section 7.2, storage capacities at the energy production site to increase the capacity usage of the adjacent grid are far from economical compared to improving the transmission grid infrastructure and spreading generation sites geographically.

PV rollout in North Africa is still in a very early stage (for further information, please refer to Section 2.1). Only few data on existing projects in North Africa is available. While PV is a fluctuating source of energy, in the further analysis, premiums for PV ground installations are assumed alike to premiums for WT installations.

As a consequence of the economies of scale described above different values are assumed for grid connection of medium scale and very large scale WT or PV parks. Grid connections are assumed to be solely built to connect the respective wind or PV park, thus no additional generators being connected to the transmission line. Consistent with the examples mentioned above, the capacity usage of the transmission grid adjacent to the park is assumed to be equal to the capacity usage of the same.

Setting the LCOE obtained in the first case study as base value, for midsized parks of a capacity between 20 MW and 90 MW the grid distance premium is calculated by Equation 31.

illustration not visible in this excerpt

Equation 31: Grid premium for midsize PV or wind parks

Setting the LCOE obtained in the second case study as base value, the grid distance premium for very large parks of a capacity between 1000 MW and 3000 MW is calculated by Equation 32.

illustration not visible in this excerpt

Equation 32: Grid premium for large or very large PV or wind parks

Assuming potentials to be evenly distributed in a distance class, the arithmetic mean for the grid distance classes mentioned in Section 6.2.1 is employed. Resulting grid premiums are given in Table 17 at the example of 3000 FLH.

Table 17: Grid premium by distance classes for PV ground and WT installations at 3000 FLH no wheeling charges included.

illustration not visible in this excerpt

The grid cost behaviour over time is not evaluated within this Cost-Supply assessment. However, since transmission systems are already employed for a long time cost improvements due to learning effects are not o be expected. In contrast, increasing commodity prices, e.g. copper and steel, could elevate the cost for grid connection.

6.3 Transport

Without taking Western Sahara into account, the North African coast not only comprises highest population densities in the region but also best transport infrastructure. As illustrated in Figure 44, the more inland one moves on the map, the more scarce the road and rail network becomes.

illustration not visible in this excerpt

Figure 44: Map of North African transport infrastructure including roads and rails [355]

Aside from the construction phase, the analysed technologies are of low personnel requirements during operation. PV and WT parks thereby do not require closeness to transport networks during operation. Average transport costs of the energy generation equipment are included in the initial investment. Calculating transport costs to remote sites, which may not be connected to the existing infrastructure, is a project specific task. A monetary value related to distance to transport infrastructure is not set in this study.

However, for site evaluation purposes, distance classes of (100; 200 km] and (200; ’ km] are pictured by overlaying textures in the resulting potential maps.

6.4 Resulting Cost-Supply Curves Including Grid Connection

It is assumed that wind or PV parks would be connected at shortest distance to the existing grid. Apart from new grid installations to reach energy generation sites, the grid enforcement is assumed to be along the existing grid.

The cost of energy, stated in the following section, is calculated for the point of connection to the national grid. The LCOE is in respect of technology, park size and year of investment, but does not include taxes or country specific wheeling charges, which may cover costs for general grid enforcement. Wind turbines generally reach much higher capacity usages than PV systems; consequently grid connection of PV parks is generally higher in this calculation.

The LCOE of generation potentials is furthermore put into context with the current and projected electricity demand of North Africa, in particular Egypt, and Europe. Utilised electricity demand values are stated in Table 2 and illustrated in Figure 3 according to forecasts by Trieb et al. (2009).

6.4.1 Photovoltaic Energy

Analysis for 2012

The rollout of PV systems in North Africa is at an early stage with only very few existing PV parks and no significant contribution to the country-specific energy supply.

Grid Premium for Midsized PV Parks (20-90 MW)

Today, the largest PV park worldwide stands at 97 MW of installed capacity [356]. Consequently, this “medium sized” scenario encompassing park capacities of 20-90 MW is well suited to forecasted PV technology’s LCOE including grid connection; resulting countryspecific Cost-Supply curves are illustrated in Figure 45.

illustration not visible in this excerpt

Figure 45: Cost-Supply curve for ground installed PV with prices for 2012, including distance to grid premium for medium sized PV parks

Since fewer sites achieve LCOE close to lowest country-specific LCOE, curves with grid premium for midsize parks are generally less bell shaped than curves without the inclusion of a grid premium as depicted in Figure 37. Grid premiums do not increase steadily with distance to grid but are classified by intervals depicted in Table 17. As a result, sudden “jumps” in Cost-Supply curves can be observed in countries where similar irradiation levels extend over a wide area.

The sudden “jump” in LCOE from 17 ct€/kWh to 18 ct€/kWh of the dark blue Libyan line at 6000 TWh/a is governed by sites yielding 2000 FLH and at 25 km distance to the grid, as well as sites yielding 2600 FLH and being located at a distance of approximately 375 km from the grid. Thereafter only sites at a distance • 375 km are following. The same occurs for Algeria and Egypt and Western Sahara. Cost-Supply curves of Morocco and Tunisia are hardly affected by the grid premium, since its electricity grids have a maximal distance of ca. 175 km to potential PV ground installation sites.

Lowest LCOE are elevated by 0.6 ct€/kWh to 14 ct€/kWh in Egypt. To compare Egyptian PV- Cost-Supply results with and without grid inclusion in 2012, the comparison threshold is adjusted to 14.6 ct€/kWh. A potential of 2013 TWh/a is situated below the 14 ct€/kWh threshold in the 2012 scenario without grid inclusion. Generation potentials of 1277 TWh/a remain below the threshold of 14.6 ct€/kWh, when including grid costs in the assessment. This reduction is due to a steeper slope in the Cost-Supply curve, reasoned in surcharges for remoteness. Still 1277 TWh/a equals to over 10 times the Egyptian electricity demand of 2009 and over 7 times the Egyptian demand of 2020. The scheduled 20 % of the Egyptian power demand to be supplied by RES in 2020 could be achieved at marginal LCOE of 14 ct€/kWh. A PV share of 20 % in the North African electricity demand in 2020 equating to 72 TWh/a could be covered with a marginal generation cost of 14.1 ct€/kWh.

In conclusion, this analysis reveals that sites with highest yields are located within proximity of the existing electricity grid. As depicted in Figure 19, sites achieving lowest LCOE are located in particular in southern Algeria around Tamanrasset and southern Libya around Murzuq. The potentials in south-eastern Libya are of low LCOE but appear to be situated far off from transportation infrastructure, which could drive up the installation costs. Lowest LCOE in Egypt are located along the banks of the Nile extending towards west of the Nile as well as the Gulf of Suez region.

illustration not visible in this excerpt

Figure 46: Map - LCOE of PV ground systems in 2012, including grid premium for midsize PV parks (20-90 MW)

Grid Premium for Large PV Parks (1000-3000 MW)

Compared to country-specific Cost-Supply curves without grid inclusion, grid premiums for large PV parks do not alter the curves shape by much; see Figure 47 for an illustration. Apparently, Libyan high potential areas are situated at longer distances from the grid than Algerian and Egyptian high potential areas, resulting in a greater elevation of the Libyan Cost-Supply curves than the Algerian or Egyptian, one compared to the 2012 scenario without grid connection. Nonetheless, from the current perspective, PV parks sizing above 1000 MW in capacity appear unrealistic but may be a viable dimension considering the vast open spaces available in North Africa.

illustration not visible in this excerpt

Figure 47: Cost-Supply curve for ground installed PV with prices for 2012, including distance to grid premium for large PV parks

The geographical distribution of LCOE for large sized pars in 2012 can be observed in Figure (Anex.) 15.

Analysis for 2030

illustration not visible in this excerpt

Figure 48 illustrates the forecasted LCOE for 2030 for midsize and large PV parks from a perspective comprising North African potentials jointly. Regarding the Cost-Supply curve with the midsize grid premium, 6011 TWh/a can be generated at marginal LCOE of 6 ct€/kWh. Considering the grid premium for large PV parks, a potential of 25806 TWh/a, 90 % of the entire generation potential is situated below the LCOE threshold of 6 ct€/kWh.

By the midsized-PV-parks scenario, 100 % of the North African demand in 2030 (647 TWh/a) could be covered at marginal LCOE of 5.5 ct€/kWh. The electricity demand of Europe and North Africa jointly (52067 TWh/a) is achievable at maximal LCOE of 5.9 ct€/kWh.

illustration not visible in this excerpt

Figure 48: North African Cost-Supply curve for ground installed PV with prices for 2030, including distance to grid premium for midsized and large PV parks

Including a grid premium for large PV parks, the entire North African demand in 2030 could be covered at marginal LCOE of 5.5 ct€/kWh, and the European and North African demand jointly with marginal costs of 5.3 ct€/kWh. Egypt could cover 100 % of its forecasted demand of 311 TWh/a in 2030 at marginal costs of 5.4 ct€/kWh, including the premium for midsize PV parks, and at costs below 5.1 ct€/kWh considering large PV parks. The geographical distributions of generation costs both premiums in 2030 are mapped in Figure (Anex.) 16 and Figure (Anex.) 17.

Discussion of the Results

Considering the grid connection for current maximum PV parks sizes, areas enabling PV electricity generation at low LCOE are confined to proximity to the existing electricity grid. By drastically increasing the PV park’s capacity, more remote areas are exploitable at favourable LCOE of up to 15 ct€/kWh. However, since the electricity grid already reaches extended available areas with high generation potentials, there is no need to locate PV parks far off the existing grid infrastructure to cover sizeable shares of electricity demand.

Until 2030, the LCOE of PV electricity will decline to levels, where the entire electricity supply for Europe and North Africa is achievable at costs below grid parity. However, further surcharges for transmission towards European load centres would have to be included here.

6.4.2 Wind Energy

Analysis for 2012

With Egypt and Morocco being an exception, electricity generated by wind power is not yet contributing significantly to the country-specific energy demand. Algeria and Libya are just gaining first experiences, while Tunisia so far basically extended its pilot project at Cap Bon. Consequently, the midsize scenario pictures take into consideration wind park capacity ranges as observed in other countries. Average capacities of wind parks in Germany are 9 MW, while new installed wind parks in the United States of America were developed with an average of 90 MW capacity in 2009 [357]. Existing wind parks in Morocco and Egypt as well as plans for Western Sahara suggest wind parks being installed at much larger sizes. The grid premium for large wind parks will take this development into consideration.

Grid Premium for Midsized Wind Parks (20-90 MW)

The geographical distribution of LCOE for 2012, including the grid premium for midsize wind parks, is depicted in Figure 49. An overlaying texture maps the distance of potential sites to transport infrastructure. Though not being far off the transport infrastructure, south-western Algeria, southern Libya and south-western Egypt are too remote from the electricity grid for economically viable potential exploitation below a threshold of 14 ct€/kWh.

Highest yielding sites in Western Sahara and Morocco, Tunisia and Egypt are not much affected by this grid premium since they are located within the limiting distance. In Algerian, grid cost including LCOE concentrate around the town of Ain Salah, very close to the site of Algeria’s first industrial scale wind park at the town of Adrar [358].

illustration not visible in this excerpt

Figure 49: Map - LCOE of wind energy in 2012; including grid premium for midsize wind park (20-90 MW)

Including distance-to-grid in their wind energy potential evaluation of Egypt, Elsobki et al. (2009) describe the Gulf of Suez area holding best potentials, since it is nearby the electricity grid and close to the main demand centre, Cairo, located at a distance of 250 km. Along the River Nile, good grid infrastructure is coupled with better wind potentials west of the Nile than east of the Nile. The western deserts hold good wind potentials but are far off the grid.[359] Supporting the findings of Elsobki et al. (2009), this analysis further reveals that the Egyptian Mediterranean coastline west of Alexandria features grid-accessible sites yielding high annual power output, consequently holding low LCOE.

Country specific Cost-Supply curves, as depicted in Figure 50, are much steeper with grid inclusion than without grid inclusion. Minimum LCOE are just elevated by a grid premium of 0.1 to 1.4 ct€/kWh. Maximum LCOE are up to 14 ct€/kWh higher than in the 2012 scenario without grid consideration. These sites with cost of energy as high as 26 ct€/kWh are beyond economic viability. Conversely maximum wind LCOE for Morocco is just elevated by 1 ct€/kWh.

By this scenario with parameters as of 2012, Egypt could satisfy a demand of 224 TWh/a at LCOE lower than 6 ct€/kWh, that is 7.1 % of the total generation potential and 1.3 times the Egyptian demand forecasted for 2020. A potential of 1123 TWh/a could be generated at a marginal cost of 8 ct€/kWh. As depicted in Section 2.1.5, Egypt plans to supply 12 % of its electricity demand in 2020 by wind power, this demand is forecasted with a total of 171.5 TWh/a. Thereby, an installed capacity of 7000 MW is ought to provide 20.58 TWh annually, which would account for a marginal cost of 3.72 ct€/kWh within the scenario in hand. Consequently, the potential LCOE of the planned wind energy capacity is well below Egyptian electricity prices of 7.84 ct€/kWh for consumer demanding over 1000 kWh per month [360].

From a North African perspective, a generation potential of 1126 TWh per year is achievable at costs below 6 ct€/kWh within this scenario. A marginal LCOE of 8 ct€/kWh is not surpassed by 3226 TWh/a. The North African electricity demand of 360 TWh/a in 2020 could be covered at LCOE of 3.5 ct€/kWh. Joint European and North African electricity demand of 2030, forecasted with 5206 TWh/a, could be covered at marginal LCOE of 9.9 ct€/kWh within this midsize wind park scenario with financial parameters for 2012.

Grid Premium for Large Wind Parks (1000-3000 MW)

Figure (Anex.) 6 maps 2012’s wind LCOE with a grid premium for large wind parks with capacities of 1000 MW to 3000 MW. Thereby, sites are not as confined to the route of the existing grid as with a grid premium for midsize wind parks. Compared to the map of 2012’s potentials with midsize-park-premium, south-western Algeria - around the town of Tindouf - has become the second favourable WT installation zone in Algeria, the other one continuing to be the region around Ain Salah. The south-western region is far-off transport and grid infrastructure but holds wind regimes of such high levels that they outweigh the remoteness to the electricity grid. Country specific Cost-Supply curves for this scenario are illustrated in Figure (Anex.) 3.

Within the scenario for 2012 including grid premiums for large wind parks, Egypt holds a potential of 1643 TWh/a below LCOE of 8 ct€/kWh and 334 TWh/a, that is 10.7 % of the total Egyptian potential and almost twice the Egyptian electricity demand of 2020, below the threshold of 6 ct€/kWh. North Africa could provide 1266 TWh/a, that is 7 % of the total North African generation potential, below 6 ct€/kWh. A potential of 4243 TWh/a holds a marginal cost of 8 ct€/kWh. Since wind energy deployment in Egypt is of an advanced stage, wind parks are growing in scale and are connected to the grid jointly as we see with the Zafarana wind park connection ought to evacuate 2290 MW of installed wind power capacity [361]. The 12 % wind power share of the Egyptian power mix planned for 2020 could be achieved at marginal costs of 3.6 ct€/kWh. Compared to the scenario with a grid connection for midsize wind parks, as stated above, the savings in LCOE for this “small” share of the overall potential is just 0.12 ct€/kWh. Since the Egyptian electricity grid already reaches the highest yielding sites today, as depicted in Figure 43, the difference is negligible.

illustration not visible in this excerpt

Figure 50: Wind energy Cost-Supply curves for 2012, midsize wind parks Analysis for 2030

Consistent with the analysis for 2012, for large sized wind parks, between 1000 and 3000 MW installed capacity, the grid premium, as depicted in Figure 51 including site specific LCOE, is only slightly altering the Cost-Supply curve’s shape. Wind turbine sites are essentially regrouped from purely FLH merit order to FLH-distance merit order, resulting in smoothened curves as observable with Egypt. Compared to the 2030 scenario without grid premium, the minimum LCOE for Tunisia is elevated by just 0.01 ct€ per kWh, with Algeria the minimum LCOE increases by 0.27 ct€ per kWh, still an insignificant increase. The upper end LCOE for Algeria has risen by 1 ct€/kWh, the maximum LCOE for Tunisia increased by 1.36 ct€/kWh. This indicates slightly steeper cost curves with similar starting points compared to the 2030 scenario without a grid premium.

In the Egyptian case, the analysis shows that a generation potential of 4143 TWh/a lies below the 8 ct€/kWh threshold, 3184 TWh/a are situated below the 6 ct€/kWh threshold, that is 71.5 % of the Egyptian generation potential, thereby 12.1 % less than without the inclusion of a grid premium for large wind parks. Still, the forecasted demand of 311.7 TWh/a in 2030 for Egypt is covered tenfold.

illustration not visible in this excerpt

Figure 51: Wind energy Cost-Supply curves for 2030, large wind parks

When considering medium sized wind parks with a capacity between 20 and 90 MW, the distance premium has a significant impact on site economics, as illustrated in Figure 52. Premiums on minimum LCOE as per country stand at 0.1 ct€/kWh, maximum LCOE, however, are by just 0.8 ct€/kWh to over 12 ct€/kWh higher than in the 2030 scenario which does not consider a grid premium.

Spotlight on Egypt, the analysis reveals 3565 TWh/a at generation costs lower than 8 ct€/kWh and 2148 TWh/a, that is 46 % of the Egyptian generation potential, below the threshold of 6 ct€/kWh.

illustration not visible in this excerpt

Figure 52: Wind energy Cost-Supply curves for 2030, midsized wind parks

Discussion of the Results

Already today, wind power is achievable at costs below electricity retail prices in Egypt. Feed-in top-ups as existent in Algeria or the planned Egyptian feed-in tariffs are not required for competitiveness but enable stable revenue streams [362] and, consequently, allow efficient debt structures. The competitive edge of wind power will enhance further, providing incentives for accelerated installation of wind parks. In regard to industrial development and employment opportunities, Egypt is so far the only North African country capitalising on the large scale wind power rollout, by establishing a domestic wind power manufacture and assemble industry (please refer to Section 2.1.5 for detailed information).

6.4.3 LCOE of North African Wind Energy on the European Market

Since wind power generation potentials are of a magnitude where they can cover high shares of electricity demand in North Africa as well as in Europe, to compare North African wind energy LCOE with the power mix LCOE in Europe, the cost of electricity transmission to the European market has to be included.

As described above, Morocco, Algeria and Tunisia are already connected or are planned to be connected to the European grid by HVDC submarine-cables in the midterm future.

Within his assessment of North African CSP potentials, Hilgers (2010) assumes all generation units feeding into 600 kV HVDC overhead transmission lines with a capacity of 4800 MW, which traverse the country towards switching stations at the coast. Capacity usage is assumed as equal to the full load hours of the connected energy generation unit. Subsequently, HVDC submarine cables with a capacity of 700 MW transmit the energy to Europe’s coast passing the Mediterranean basin. Connection points are according to existing or planned transmission corridors traversing the Strait of Gibraltar, the seabed between Tega (Algeria) and Almeria (Spain) as well as Tunisia-Sicily and Libya-Sicily. In this scenario, Egypt and Europe are linked overland via the Middle East.

Hilgers (2010) calculates transmission costs for CSP with values of 2010. They incorporate transmission losses of 5 % per 1,000 km for HVDC overhead lines and 4.8 % per 1,000 km for HVDC submarine connection. Grid premiums are calculated according to site-specific capacity usages of CSP plants with adjacent storage. Values vary in the range of 3607 and 4361 FLH. Thus, CSP capacity usage is comparable to the capacity usage of wind turbines at highest yielding sites and considerably higher than the capacity usage of the analysed PV systems. The following approximation is thereby only valid for wind energy.

For transmission from sites in Morocco to Europe, Hilgers (2010) equates average premiums of 1.2 ct€/kWh; for Tunisia-Europe 2.1 ct€/kWh, for Algeria-Europe 3 ct€/kWh, for Libya- Europe 4.5 ct€/kWh and for Egypt-Europe 33.9 ct€/kWh. In comparison, Schermeyer (2011) states LCOE for the present HVAC link between Morocco and Spain with 0.5 ct€/kWh.

Within the following scenario, Hilgers’ transmission costs to Europe are considered as country specific wheeling charges. Western Sahara is considered at same Europe- transmission-premium as Morocco, since Moroccan wheeling charges, as stated in Section 6.2.2, are valid for both countries as well. The site specific grid premium for large parks calculated in this analysis works as a LCOE surcharge regarding remoteness of the WT site. Large scale wind energy exports to Europe require a large scale wind turbine deployment. Consequently, only grid premiums for wind parks of capacities between 1000 MW and 30000 MW are considered within this analysis. This is in line with the capacity of transmission infrastructure considered by Hilgers (2010).

Cost reduction effects regarding HVDC and HVAC transmission costs are not incorporated into this analysis.

The assessment’s results for 2012, depicted in Figure 53, show lowest LCOE for wind energy generated in Morocco and Western Saharan, followed suit by Tunisia. More expensive transmission costs, as well as less favourable wind sites, elevate the minimal LCOE for Algeria and Libya to about 11 ct€/kWh. Egypt’s Cost-Supply curve does not appear in the diagram, since its minimal LCOE for wind energy, as depicted in the adjacent legend, stands at 39.2 ct€/kWh in this scenario.

Evaluating the cost for North African wind energy on the European market, we assume that 20 % of the North African domestic demand, thus 72.2 TWh/a in the year 2020, is supplied by domestic wind energy at lowest LCOE. This does not elevate the minimal wind LCOE for Europe, which still stands at 4.2 ct€/kWh marginal cost.

For 2020 the EU set bindings targets of 20 % of the demanded electricity to be generated by RES. North African wind energy could supply 20 % of the EU’s demand in 2020, thus 800 TWh/a, at marginal LCOE of 6.9 ct€/kWh. European retail prices were between 10 and 11 ct€/kWh in 2007 (Spain: 10.8 ct€/kWh; France 9.8ct€/kWh; Germany: 11.1 ct€/kWh)[363]. Thus from an economic point of view, the assumed 20 % share of wind energy in the power mix of 2020 is achievable and beneficial to the North African wind power market.

A power output of 4361 TWh/a, thus 100 % of the total energy of North Africa and the EU in 2020, has generation costs below 12.4 ct€/kWh. Kost and Schlegl (2010) assumed an LCOE of 6 ct€/kWh for the German power mix in 2010. This threshold is not surpassed by a North African generation potential of 857 TWh/a.

In a further scenario, the forecasted Levelised Cost of Electricity (LCOE) values for 2030 and 2050, as provided in Section 4.4.4, are completed with the above mentioned grid premiums. As stated in this section, grid premiums for transmission to Europe and remote sites in North Africa are added at costs of 2010. Linearly interpolating the Cost-Supply levels between 2012, 2030 and 2050, we obtain an estimated development in marginal costs of cumulative generation potential. Forecasted electricity demands for North Africa (NA) as well as North Africa and the EU conjointly (NA & EU) are added the to the forecasted LCOE of wind energy, including both transmission premiums.

illustration not visible in this excerpt

Figure 53: Wind energy Cost-Supply curves for 2012, incl. transmission costs to the European market for large wind parks

Demand forecasts for North Africa are by Trieb et al. (2009). European demand forecasts are provided by the European Commission [364]. The combined demand of North Africa and the European Union in 2050 is derived from Scholz [365]. The depicted Cost-Supply surficial-curve in Figure 54 is not representing the total North African generation potential but focuses on the relevant portions of the curve. The lowest level in this surficial curve is governed by potentials generated in Morocco and Western Sahara at sites close to the existing grid. The first rise consists of best Tunisian potentials followed by second best Tunisian, Moroccan and Western Saharan potentials with the best Algerian potentials mixing in. The subsequent plateau is governed by low and medium yielding Tunisian sites as well as Algerian and Libyan potentials. When covering first the total North African and then the total European electricity demand (total 4361 TWh/a) in 2020, the EU would import 4000 TWh/a of North African wind power at marginal LCOE of 10.8 ct€/kWh. By the same assumption, the EU would import its total demand in 2030 (4560 TWh/a) at marginal LCOE of 8.9 ct€/kWh. The demand of the European continent in 2050 (8334 TWh/a = 9560 TWh/a[366] - 1226 TWh/a[367] ) would be supplied at costs below 8.8 ct€/kWh.

It is important to stress out that incorporated transmission premiums for Europe and the location-specific distance premiums are not obtained by identical methodologies. Furthermore, capacity usage of the transmission system to Europe is assumed equal to the capacity usage of CSP plants although wind parks are observed in this approximation. Consequently, the results are solely to illustrate the economic viability of wind energy export to Europe. Findings cannot be considered as absolute LCOE of North African wind energy when reaching European shores.

illustration not visible in this excerpt

Figure 54: Forecasted Cost-Supply curve for wind energy including transmission cost to the European shore

Comparison to LCOE of CSP exports

According to Hilgers, CSP electricity would reach European shores from Morocco, Libya and Algeria at most economical LCOE of 17 to 19 ct€/kWh. After covering 100 % of the North African electricity demand of 2020 by CSP energy, an amount of 80 TWh[368], that is 2 % of the EU’s demand of 2020, could be imported to the EU at LCOE of 19 ct€/kWh. A share of 15 % of the EU’s demand of 2020 could be covered at LCOE of 21.4 ct€/kWh according to Hilgers’ findings.

By this approximation, the LCOE of wind energy, when exported to the EU, is very competitive to the LCOE of CSP. Considering the capacity usage of CSP plants with adjacent storage, which is similar only to wind turbines at highest yielding sites, the LCOE of wind energy exported to Europe from medium yielding sites is likely to be to low by about 10 to 20 %.

7 PV and Wind Energy in a Broader Context

7.1 Comparison to CSP

In particular for the “sun kissed” regions of the world, Concentrated Solar Power (CSP) technologies are receiving highest recognition as the technology of choice to provide renewable energy at low emissions and high capacity credit, i.e. the avoidance of conventional power. CSP is using the direct normal irradiance (DNI) portion of the global irradiation on earth to convert it via concentrating mirrors, either by parabolic through or tower technologies into thermal energy. In most cases, this thermal energy drives a steam turbine connected to a generator thereby producing thermo-electric energy. The efficiency of the applied Clausius-Rankine-Cycle increases with the temperature and pressure delta between input and output at the steam turbine. Thus, efficient heating of the input streams and rapid cooling of the output streams is paramount when operating in a closed cycle. Most commonly, the latter is achieved by wet-cooling condensers. These absorb thermal energy with water as heat transfer medium which subsequently is evaporated into the atmosphere. Direct solar irradiation is highest at the Tropic of Cancer and Capricorn, which are, in particular in North Africa, very water scarce areas. To reduce the water usage of 3 m²/MWh required by wet-cooling CSP plants, dry-cooling is feasible with 0.3m²/MWh water. However, when comparing wet- and dry-cooling, the efficiency drops from 14.5 % to 11 % for parabolic through and 13.5 % to 12 % for tower technology CSP plants [369].

illustration not visible in this excerpt

Figure 55: Principle of the parabolic through CSP plant [370]

On the other hand, at sites in proximity to the sea, salt water can be used for wet-cooling and simultaneously be desalinated. A 200 MW turbine fed by a 250 MW of solar concentrators would allow parallel desalination of 100,000 m³ per day. Within the North African region, high coastal irradiation (DNI > 2400 kWh/(m² a)) is available in Tunisia and Egypt [371].

Water usage is zero for wind power, followed by 0.02 m³/MWh for CPV and 0.11 m³/MWh for PV required for surface cleaning [372] [373]. To put these numbers into context, a nuclear power plant with tower cooling requires 4.41 m³/MWh, similar to a generic coal fired power plant requiring 4.43 m³/MWh.

Figure 56: Operational profile of a reference CSP plant with heat storage and co-firing during three days in August [374]

illustration not visible in this excerpt

The operational profile of a reference parabolic-through CSP plant (100 MW nominal capacity, storage with 7.5 hours operation at full load capacity) located at Ouarzazate, Morocco is depicted in Figure 56 [375]. Under optimal irradiation conditions the CSP plant generates power, while simultaneously loading its storage for night time power generation. When weather conditions prevent high levels of power generation by solar insolation (DNI curve), co-firing enables the CSP plant to contribute reliably to the electricity demand during peak loads, which is between 5pm and 9pm in the case of Morocco. Comparing operational data of existing and planned CSP plants, Schermeyer (2010) finds 12-18 % of CSP plant’s power output being generated by fossil fuelled co-firing.

Reliable, controllable power output - achieved by adding thermal storage combined with co- firing to a CSP plant - is CSP’s competitive edge when comparing it to the volatile generation of PV and wind power, as depicted in Figure 29 for the cumulative installed Moroccan wind capacity of 2020.

To take environmental constraints of North Africa into consideration, Trieb et al. [376], Hilgers (2010) and Schermeyer (2011) only consider dry-cooled technologies to be deployed in North Africa. According to Hilgers (2010), dry-cooled CSP tower technology with 11 MW capacity and adjacent molten salt storage - providing full-load operation for additional 9 hours - is the option with lowest LCOE of 17-27 ct€/kWh in North Africa [377]. In contrast, Kost and Schlegl (2010) state parabolic through with adjacent thermal storage to have lowest LCOE of 13-17 ct€/kWh compared to other CSP technologies. Kost and Schlegl (2010) use parameters of existing CSP plants and CSP plants under construction. All currently operating CSP plants employ wet-cooling [378]. For the reference CSP plant at Ouarzazate, Schermeyer (2011) calculates LCOE of 13.3 ct€/kWh.

Consequently lowest LCOE of dry-cooled CSP systems are similar to LCOE of PV ground installed system, which achieve LCOE of 13.5 to 14 ct€/kWh at best sites in North Africa. The LCOE of wind power - between 3 and 4 ct€/kWh at sites with highest yields - is up to 10 ct€/kWh more economic.

Concentrated photovoltaic (CPV) is another technology competing with CSP. CPV and CSP both solely use direct normal irradiation. In comparison to genuine, fix-mounted photovoltaic, CPV systems also track the sun, providing a more stable electricity supply over the day. The LCOE of CPV is currently evened out with the LCOE of CSP [379]. By improving systemefficiency from today’s levels of 22-27% [380] to the forecasted 35 % [381] as well as the projected cost decrease; CPV will gain competitiveness in the near future.

7.2 Meeting Electricity Demand with PV and Wind Power Supply

PV and wind electricity generation are highly dependent on the volatile resources: solar irradiation and air mass movement. When matching this unstable electricity supply with a predictably varying demand curve, the disadvantage of adjacent utility scale storage systems not yet being available is a drawback compared to CSP technology. Possibilities to overcome this disadvantage are depicted in this section.

With increasing proportion of wind power fed into the grid, the power quality control and back-up power or storage level gain importance [382].

7.2.1 Passive Balancing by Siting and Costs for Active Balancing

The magnitude of generated wind and PV power is variable in time due to prevailing local weather conditions. Within the Southern Great Plains Network in the United States of America, Mills and Wiser [383] analysed the influence of distance on fluctuation in power production, when varying the distance between PV or wind power generation sites arranged in an array. While the site specific PV electricity output can drop by 80 % in just one minute, a distance of 20 km efficiently smoothens sudden changes in PV power output on the 1 to 5 minute time scale to near zero.

To actively enable a stable electricity supply, fluctuations in the same have to be balanced by holding “spinning” reserves - plants online and synchronized with the grid - and “nonspinning” reserves, which are available on short notice. Mills and Wiser further evaluate the cost of these reserves required to compensate supply fluctuations during operational time of PV (only daytime) and wind power. Cumulative balancing costs for 1-minute, 10-minute and 60-minute deltas are calculated. In comparison to the allocation of the entire PV capacity at just one site, the spreading over a 5x5 grid with 50 km distance between grid points reduces the cost for balancing of PV electricity to 7 %, thus from 2.8 ct€/kWh to 0.2 ct€/kWh (*). Wind power arranged by the same array incurs lower balancing costs of 0.1 ct€/kWh (*). A penetration of 10 % of either PV or wind power is assumed [384].

By further increasing the spread over a country, in this case Germany, 5-minute PV fluctuations are reduced to ± 5% (Wiemken et. al 2001)[385]. Fluctuations in aggregate wind output are reduced to 20 % by the spread of WTs in Germany[386]. By 2020, on- and offshore wind power are scheduled to supply over 20 % of the generated electricity in Germany. Sensfuss et al. found the assured percentage of cumulative installed wind capacity, i.e. available with a probability of at least 97 %, to be increasing from 1.5 % in 2010 to 2.5 % in 2020 [387].

According to GWEC, common power networks are estimated to be adequate for a wind energy penetration level of 20 %, the level in Denmark today. For wind power penetration levels above 20%, additional costs of 10% of the initial WT investment incur for balancing. Hydro power and gas fuelled power plants in the overall power mix improve the flexibility of a power network, which as a result, can cope with higher penetration levels of wind energy [388]. Operational flexibility of fossil fuelled power plants can be improved by storing heat when operating as reserve capacity [389] Accurate RES power output forecasts are necessary for efficient scheduling of “spinning” and “non-spinning” reserves [390]. Errors for regionally aggregated wind farms are 10 % for a day ahead and 5 % for 1 to 4 hours ahead [391].

illustration not visible in this excerpt

Figure 57: Ratio between average wind speeds in July and January in Europe and North Africa [392]

Further extending the distances between wind park sites to 1000 km, passively balances the variations in wind power output over an entire day [393]. With greater, continental extension of the wind power generation network, seasonal differences in wind speed magnitude play in favourably. A map of the wind speed ratio between summer (July) and winter (January) is obtained with data provided by SWERA [394] and illustrated in Figure 57. Europe experiences wind speed maxima in winter, while the Arabian Peninsula is characterised by stronger summer winds. Western Sahara as well as central Libya and Egypt hold slightly stronger winds in summer than in winter. Generally speaking, the North African region depicts low seasonal differences, which is favourable for a stable electricity supply of North Africa and wind power exports to Europe.

As depicted in Figure 58, the magnitude of irradiation on tilted surfaces (LTI) in Europe is seasonally dependent. The winter electricity generation of PV modules in Scandinavia is reduced to 5 % of the summer output. The North African region in contrast, experiences low seasonal variations [395], in particular when approaching the Tropic of Cancer.

Abbildung in dieser Leseprobe nicht enthalten[396]

Figure 58: Ratio between average irradiance in July and January in Europe and North Africa

The variations of PV electricity production due to the sun’s movement can be further reduced by sun-traceably mounted PV modules [397], varying fixed PV module orientation within an array to other than azimuth angle Į = 0° (southerly orientated) and by spreading PV parks over time zones.

7.2.2 Electricity Storage

Utility-scale storage of electricity is commonly achieved by pumped hydro accumulation storage (PAC) [398]. Existing PAC plants in Europe add up to an annual potential of 144 TWh with LCOE between 5.7 and 17.5 ct€ per kWh [399]. Morocco has currently an installed PAC capacity of 464 MW storing electricity with an efficiency of 76 % [400]. Morocco plans to further increase its PAC capacity to 2000 MW by 2020 next to its current hydroelectric capacity of 1286 MW [401]. In Egypt, hydroelectric power plants along the river Nile generated 12,863 TWh in 2009 [402] and could ensure grid stability with increasing wind and PV power penetration. As illustrated in Table 3, other North African countries have yet a negligible installed hydroelectric capacity. Wietschel and Genoese (2010) evaluate storage costs for PAC with 4 to 8 ct€/kWh for day-scale storage with 8 FLH, and 8 to 16 ct€/kWh for weekscale storage with 20 FLH capacity. By adopting a grid electricity price of 1.5 ct€/kWh, these costs consider losses [403].

Compressed air energy storage (CAES) is an emerging, multi-MW storage technology available at costs of 6-12 ct€/kWh for day-scale storage and 8-16 ct€/kWh for week-scale storage. Considering the electricity price peak of 10 ct€/kWh over 45 hours time-span in Germany in 2009 [404], further storage technologies are currently not competitive.

These further technologies include hydrogen storage and REDOX-Flow batteries with MWscale storage capacity. Short term storage is achievable with flywheels, super- and doublelayer capacitors [405].

In the long term, Hau (2008) suggests hydrogen as the storage medium for electricity generated by wind and PV [406]. Furthermore, in the opinion of the Global Wind Energy Council, electricity storage directly adjacent to the wind turbine is “neither necessary nor economic” [407] considering the above mentioned balancing effects accomplishable by a relatively economic transmission infrastructure enhancement.

7.2.3 Vision of the Trans-Mediterranean Grid

Pointed out by the evaluation of Section 7.2, improving the transmission infrastructure in extent and capacity is essential in coping with higher penetrations of electricity generated by fluctuating sources, e.g. PV and wind. By estimation of the IEA, until 2030 investment in transmission and distribution networks have to reach 1.8 trillion USD in the OECD [408].

Two European initiatives target the development of North African RES generation and its integration into the European power demand market:

Medgrid is backed by French electric utility EDF as well as technology partners mainly of the transmission industry and further utilities. Until 2020, Medgrid envisions to install 20 GW of solar and wind power complemented with 5 GW of HVDC transmission infrastructure linking North African countries and Europe. Investments are estimated with 38 to 46 billion Euro, thereof 6 billion for electricity links [409].

The Desertec Foundation was initially conceived by politicians, economists and scientists as North African institutions. Activities are funded by the German government and industrial partners comprising technology companies, utilities, banks and insurances. The main technological focus in energy generation is set on CSP [410] with utility scale reference projects to be installed in Morocco, Algeria and Tunisia until 2020 along with HVDC links connecting North Africa with Europe. Along with enabling electricity exports to Europe, the Dii involves domestic demand of North African countries to be covered with higher shares of electricity by RES [411].

Medgrid and Desert signed cooperation agreements in November 2011 [412].

7.2.4 Cost Effectiveness and Supply Security

The Moroccan plans to integrate 2000 MW of wind power capacity and 2000 MW of CSP plants into the power network by the year 2020 have been assessed by Schermeyer (2011) from a supply security and cost perspective. He considers the forecasted power plant fleet of 2020 to comprise fossil fuelled plants (5658 MW), hydropower (91 MW), hydro and pumped- hydro accumulation storage (max. 1899 MW) as well as electricity trade (max. 1125 MW). In order to provide a secure electricity supply, the available and controllable capacity has to be 10 % higher than the maximum demand, which will peak in 2020 at 8500 MW during evening hours.

Within the economic evaluation, the total annual costs of the power system comprise variable expenditures for all included generation types as well as fixed expenditures for capacities to be built to cover the higher electricity demand of 2020. By the integration of RES, these variable costs are reduced, mainly due to less need for fuel. However, since fossil capacities are still required, total fixed expenditures do not decline. In the scenarios for 2020, wind turbines, operating at a network capacity between 17 and 80 %, and CSP plants, with adjacent 7.5 hours of thermal storage, feed into the grid on priority. The remaining demand is covered in merit order by other available capacities. Schermeyer (2010) discloses that supply security is reached by adding 553 MW of CSP capacity to the forecasted fossil and hydro power plant park. Additional electricity sourced by low priced wind power substitutes expensive electricity imports at peak hours and reduces the operational time of higher priced oil-fired steam turbines and combined-cycle plants [413].

In short-term scenarios the rising RES integration reduces the total power network’s LCOE. In spite of incorporating the expected rises in fuel costs, Schermeyer (2011) finds a 2020 scenario with the extension of the existing fossil-hydro power plant fleet with additional gas and coal fired power plants still by 1.3 ct€/kWh lower in the power network’s LCOE than the most economic power network configuration under RES inclusion, which would comprise of 3000 MW in wind power and 1000 MW in CSP. With a 50 % rise in fuel costs, the power network’s LCOE in the fossil-hydro scenario would be equal to the LCOE of the power network with most economic RES inclusion. The ambitious scenario of the Moroccan government with 2000 MW CSP and 2000 MW wind power capacity holds a power network LCOE of 2.8 ct€/kWh above the fossil-hydro scenario.

Since the Moroccan power networks comprises high shares in flexible capacities, Schermeyer (2011) states that from an electricity supply security perspective, Morocco could cope with RES capacities above 2000 MW of wind and CSP power.

8 Summary, Conclusion and Outlook

8.1 Summary and Conclusion

This study aims at assessing the economic potential - magnitude and Levelised Cost of Electricity (LCOE) - of PV and wind energy in North Africa. Within the context of this assessment are the North African and European targets for renewable energy sources (RES) shares in electricity supply, the growing North African and European electricity demand coupled with the need for repowering of ageing power plants as well as the constraints to achieve electricity supply security. For the potential assessment a Geographical Information System is employed.

Ambitious targets are set throughout the North African region to elevate the share of RES in electricity supply: Morocco and Western Sahara aim at a 42 % RES share in installed capacity by 2020, with total contribution of 14 % by wind and solar energy each [414]. For 2030, Algeria targets to reach a 20 % share of RES in its electricity supply. Thereof 20 % are planned to be sourced by wind energy and 10 % by PV energy [415]. Until 2016, Tunisian fossil fuel consumption is meant to be reduced by 22 % from the level of 2009. The expansion of Tunisian wind and solar power capacity is an essential part of this reduction plan [416]. Libya envisions a share of 30 % of its electricity generation to be supplied by RES in 2030 [417]. With several extensive projects in the pipeline, Egypt is making strong effort to reach its 2020 target RES share of 20 % by 2020 with a total share of 12 % to be sourced from wind energy [418]. The target of the European Union is set at 20 % of its electricity demand to be supplied by RES in 2020 [419].

However, the stage of development in order to meet these ambitious targets is distinct in each country (please refer to Section 2.1 for further information). As this study discloses, vast solar and wind potentials remain yet untapped. On the principles of using reliable input data with highest available resolution wherever possible, the adopted bottom-up approach narrows down the available resources wind and irradiation - the theoretical potentials - through a complex set of geographic restrictions and technical constraints. The suitability of resource datasets has been evaluated. Remaining, inherent uncertainties need to be accepted due to complexity constraints.

Wind Energy

Wind energy projects are well established in Morocco, Tunisia and Egypt and are currently developed in Western Sahara and Algeria. After increasing interest and advanced project status of a pilot wind park, the exploitation of Libyan wind potentials is presently on hold (please refer to Section 2.1 for a detailed description).

For the theoretical assessment of wind energy, several datasets are compared empirically to identify suitable data inputs for the subsequent analysis steps. The wind atlases for Tunisia and Egypt are assumed to be of best quality and are consequently used as references. Moroccan and West Saharan potentials are analysed on the basis of a national atlas. Utilising higher resolution data grids - in particular in the case of Algeria and Libya, where a low resolution global atlas is employed - with time series of wind speeds and air density [420], would further enhance accuracy. Wind speed data is processed with site specific roughness lengths only when high resolution wind atlases are available, as in the case of Egypt and Tunisia. A common roughness length of 0.01 is adopted for other processed resource databases.

Compared to previous studies, the available area for wind turbines is assessed with greater variety in exclusion and restriction criteria as well as more conservative assumptions for parallel land use of agricultural and grazing lands, forests and protected areas. Aiming at obtaining most realistic outcomes, hydrologic and geomorphologic features as well as restrictions due to bird migration routes, airports, roads, railways and natural hazards are evaluated geographically, too. Thereby the suitable area is narrowed down to 73.5 % of the total North African surface, with further restrictions on parallel land use. Installation density of WT systems is set rather conservatively (4 MW/km²) compared to similar studies [421] [422].

Regarding parallel land use, a sensitivity analysis is conducted revealing a marginal effect for North Africa as a whole. However, in the case of Morocco and Tunisia, a considerable potential is situated in lands of parallel use. In the case of Tunisia, these areas are also the highest yielding wind energy sites. However, since wind turbines require only limited space for the foundation, Hau (2008) states parallel land use on cultivated lands as reasonable. Furthermore, wind turbines have a much better power output over required area ratio as e.g. PV installations [423].

The geographical assessment for wind energy reveals exceptionally good wind spots to be spatially confined to the south Moroccan and Western Saharan Atlantic coast, the Strait of Gibraltar, south-central and south-west Algeria, the Tunisian “Cap Bon”, south-eastern Libya as well as the Egyptian Gulf of Suez and the north-central Nile region towards Alexandria. Natural hazards in North Africa bear negligent risk for wind turbine systems. Studies [424] [425] show that wind turbines are additionally well suited to cope with natural hazards in North Africa, e.g. earthquakes [426].

For the assessment of the technical potential, two wind turbines - namely Gamesa’s G-80 (representing the current market leader in North Africa - Figure 25) at 80 m hub height and Enercon’s E-82 (representing technologic development) at 120 m hub height - are considered. Thus, ongoing development in turbine technology [427] and hub height [428] is incorporated into the assessment. A further novelty in this assessment is the direct conversion of wind turbine specific power curves to the wind turbine’s capacity usage in Full Load hours (FLH) via polynomial regression. External and internal losses are included in the overall performance by a lump sum performance ratio of 95 % [429]. As a novelty in the assessment of North African wind energy potentials site specific form parameter k is included. By this higher accuracy in frequency distribution of wind speeds, the total available potential is adjusted to a lower value than when assuming the default value of k = 2 adopted in previous studies [430] [431] [432]. This is based on the fact that fewer sites surpass the threshold of economic viability of 1300 FLH as adopted by Held (2010). Compared to Germany [433], still a much higher share of the total area is economically exploitable even when considering the more stringent threshold of 1600 FLH employed by Bofinger et al. (2009).

Employing the wind turbine G-80 by Gamesa at 80 m hub height, the North African technical potential of wind energy amounts to 17,865 TWh/a, with unproductive sites already being eliminated. The Enercon-E82 scenario with 120 m hub height scenario reveals a 44 % higher generation potential than when utilising Gamesa’s G-80 as reference. Manufacturers provide a wide range of wind turbine models to match differing wind characteristics. Consequently, a future assessment should include a variety of wind turbines to match the site specific wind regimes and hub height restrictions thus achieving higher accuracy of the results.

For Egypt, Western Sahara and Morocco a decisive share of its wind energy generation potential is below the adopted power mix LCOE of 6 ct€/kWh assumed by Kost and Schlegl (2010). Tunisian wind potentials appear very heterogeneous with few sites achieving LCOE below 6 ct€/kWh. Lacking high resolution input data, Algerian and Libyan potentials do not exhibit a large spread between minimum and maximum LCOE with lowest wind energy LCOE of 7.8 ct€/kWh and 6.36 ct€/kWh respectively. Overall, a North African wind potential of 1226 TWh/a, equal to 3.5 times the forecasted North African electricity demand in 2020 [434], is holding marginal LCOE of 6 ct€/kWh.

From a cost perspective, wind energy is already competitive to most other electricity sources and ought to improve its competitive edge in the future. Wind energy LCOE was forecasted for 2030 and 2050 employing the Learning Curve methodology directly on the LCOE, thereby incorporating improvements in all economic and technical aspects [435]. The 2030 forecast of North African wind power potentials depicts 3350 TWh/a at costs lower than 6 ct€/kWh, while by 2050 a wind energy generation potential of 20015 TWh/a is forecasted to hold marginal LCOE of 6 ct€/kWh, more than 2 times the forecasted joint electricity demand of North Africa and Europe [436].

PV Energy

In contrast to wind energy, market penetration of PV systems is yet negligent in all North African countries. PV rollout is governed by governmental development projects for rural electrification with stand-alone PV systems. As of today, existing grid-connected PV parks or rooftop installations are scarce in all observed countries. Please refer to Section 2.1 for country-specific information.

A novelty in this study is the inclusion of three PV installation types: ground-mounted PV systems bundled in parks, rooftop PV installations and PV systems installed vertically on facades. Latitude Tilt Irradiance (LTI) was employed for ground and roof mounted PV systems since both types are assumed to be installed according to maximum power output [437]. Thus the south orientation inclination angle is equal to the latitude position. Following the approach by Held (2010), irradiation on vertical surfaces pointing south was utilised for PV systems assumed to be installed vertically on facades. LTI was converted to vertical irradiance for this installation type.

The irradiance data utilised for the PV potential assessment is of global scale and has been verified critically by Hilgers (2010). Though low in resolution, the data is considered of reasonable good quality. In contrast to North African wind energy, where good sites are spatially confined, high levels of LTI are available in extended regions especially around the Tropic of Cancer. However, a higher resolution of input data may reveal more geographic distinction of irradiation levels. Conversely, irradiance on vertical surfaces oriented south, as utilised for facade installed PV systems, is decreasing from north to south in the North African region.

Compared to previous studies, exclusion zones for PV ground installations are analysed in greater detail and more conservatively, not allowing any parallel land use and thus protecting precious cultivated lands in North Africa. Additionally, forests, protected areas, urban areas, hydrologic and geomorphologic features as well as roads and railways have been excluded for PV, leaving 72.5 % of the total North African surface to be suitable for PV ground installations. The available area for PV rooftop and facade installations was approximated by country specific population figures (according to the approach by Held (2010) and data given in Table (Anex.) 2) combined with the spatial distribution of settlements and industrial buildings.

The technical analysis points out that temperature sensitivity of solar-cells has a strong negative impact on PV module efficiencies. Thin-film modules depict a better performance under high temperature, Saharan climates, than PV modules consisting of crystalline silicone solar-cells. Safety threats from dysfunctional PV modules are not to be expected due to the North African climatic challenges, although degradation will most likely be higher. The lump sum performance ratio of 80 % involves all losses of the PV system. The reference module efficiency of 14.6 % is obtained according to the current market share of mono-/multi- crystalline and thin-film systems including highest technology specific module efficiencies of several manufacturers as depicted in Table (Anex.) 10. While thin-film technology exhibits lower module efficiency than crystalline technologies, the forecasted technological transition from crystalline silicon to thin-film PV [438] was incorporated by an assumed stagnation of this reference module efficiency. As described, PV thin-film modules are better suited for hot North African climates and will as a result most likely expand its market share in this region.

The market for CPV technologies is currently gaining momentum: manufacturing capacities are being scaled up and CPV parks installed [439]. Since CPV operates with direct solar irradiance [440] received in high proportions at the global “sunbelt” [441], thus the zone around the Tropics, it should receive more attention in future Cost-Supply assessments.

PV ground systems have a generation potential of 28,380 TWh/a; a potential of 314 TWh/a is computed for rooftop PV installations, while PV systems mounted vertically on facades could render 20.6 TWh/a in the entire North African region. Since the potential of rooftop and facade installed PV systems is assumed in relation to population figures which depict a strong growth tendency (please refer to Figure 3), the generation potential of these PV installation types is forecasted to grow as well.

Financial parameters were adopted individually for each installation type. Holding highest irradiation levels, Libya, Algeria and Egypt depict lowest LCOE of PV ground installed systems with 13.5 ct€/kWh. LCOE of 14.3 ct€/kWh of PV ground installations is reached in Morocco, Western Sahara and Tunisia. On a North African level, 6749 TWh/a are achievable at marginal LCOE of 14 ct€/kWh, thus currently not competitive to wind energy and other common energy technologies [442]. Due to higher initial investment and less irradiance at installation sites, the LCOE of rooftop installed PV systems about 2.5 ct€/kWh higher than LCOE of PV ground systems. The LCOE of PV-facade-systems is far from being competitive with minimum values of 60 ct€/kW.

LCOE forecasts are conducted applying the Learning Curve methodology [443] under the assumption of declining initial investment and additionally declining O&M costs as percentage of the initial investment [444]. By 2030, almost the entire generation potential of PV ground mounted systems is forecasted to have marginal LCOE of 6ct€/kWh; thus 27,271 TWh/a and over 5 times the combined electricity demand of North Africa [445] and the EU [446] forecasted for 2030. Rooftop PV systems will improve their competitiveness similarly to ground mounted systems with a forecasted generation potential of 136 TWh/a at marginal LCOE of 6 ct€/kWh in 2030. By the assessment in hand, PV facade systems will not reach LCOE in the range of power mix LCOE until 2050.

The Learning Curve methodology is highly reliant on forecasted cumulative installed capacity and assumed learning rates [447]. Inherent uncertainties in the calculation avoided.

Grid Infrastructure

The inclusion of the existing infrastructure, a novelty in this type of analysis, earmarks North African regions, which are of favourable wind regimes or solar irradiance combined with grid and transport infrastructure proximity. After an initial phase experiencing the construction of PV and wind parks of medium sizes (capacity of 20-90 MW), grid premiums per generated kWh are assumed to drop due to more extended new wind and PV park installations (capacity of 1000-3000 MW). The adjacent grid is assumed to have the same capacity usage as the connected park, thus grid connection for PV parks (max. 2088 FLH) is generally higher than for wind parks (max. 5000 FLH). Consequently, in the assessment for PV parks, the inclusion of grid proximity into the assessment confines the formerly extended areas with uniformly high solar irradiation to high yielding areas along the existing infrastructure. Still 74 TWh/a, 20 % of the entire North African electricity demand in 2020 [448], could be supplied at marginal LCOE of 14.1 ct€/kWh.

By including grid premiums for wind parks, resulting site specific wind energy LCOE are arranged in a steeper curve. Tunisia and Morocco are relatively small countries combined with an extensive electricity grid, as depicted in Figure 42. As a result, the inclusion of a grid premium has a low effect on Tunisian and Moroccan wind energy LCOE and economic availability of generation potentials. Egypt and Western Sahara possess many high yielding sites in proximity to the electricity grid as well. In the case of Algeria, high yielding sites coupled with proximity to the existing electricity grid are situated in southern-central Algeria. Good wind sites in proximity to the existing grid are situated as well in north-eastern Libya around the town of Benghazi. The entire North African demand of 2020, i.e. 360TWh/a [449], could be covered at marginal LCOE of 3.5 ct€/kWh including the grid premium for midsize wind parks, thus much lower than the adopted power mix LCOE of 6 ct€/kWh.

For wind energy, costs of transmission to connection points on the European coastline have been approximated in order to evaluate the competitiveness of North African wind energy on the European electricity market. The site-specific LCOE including grid premiums for large sizes wind parks is topped-up with average country-specific grid premiums [450] incorporating cost and electrical losses incurred by the transmission from North Africa to European shores. In 2020, a share of 20 % of the North African electricity demand [451] is assumed to be covered by North African wind energy of lowest LCOE. An additional 20 % of the European demand [452] could be covered at marginal wind energy LCOE of 6.9 ct€/kWh in 2020. Thereby, wind energy LCOE is well below the LCOE of North African CSP-electricity exported to Europe (min. 17 ct€/kWh [453] ) and below European electricity retail prices (ca. 10 ct€/kWh in 2007 [454] ). However, due to simplified cost calculation, resulting LCOE are merely illustrating feasibility and tendency. Higher accuracy in the equation of site-specific transmission premiums is required here.

The produced maps for PV and wind energy may be further utilised to identify high potential regions under grid proximity and transport infrastructure proximity constraints.

Further Topics

In comparison to further technologies - CSP and conventional/nuclear power plants - PV and wind energy technologies have the following advantages: PV and wind energy parks consist of many units which can be added subsequently, thus at a scalable operating park capacity [455]. Wind energy enables a dual use of land [456]. PV systems can be installed on roofs and facades [457], thus not requiring land used otherwise. According to Kost and Schlegl (2010), a much faster implementation of PV and CPV projects provides a competitive edge over CSP plants. Water usage of wind energy is 0 m³/MWh, PV requires less water for cleaning than dry-cooled CSP [458].

However, the fluctuating nature of wind [459] and PV energy is a drawback. For a secure electricity supply with high penetration of PV and wind energy, next to further RES technologies, several implications are disclosed in the following:

- Geographical spread of generation units mitigates fluctuations in the supply of PV and wind energy[460] thereby enhancing PV and wind energy’s capacity credit [461].
- Both, North African PV and wind energy potentials are not as seasonally affected as European potentials and are consequently suitable to complement European RES supply (Figure 57 and Figure 58).
- The more flexible a power network, thus controllable capacities that can cover PV and wind energy supply depressions, the higher the possible share of wind and PV energy [462].
- Storage capacities enhance flexibility [463] but so far only pumped hydro accumulation storage and to a minor degree compressed air energy storage are economically viable [464].
- With higher LCOE than wind energy and similar LCOE to PV energy, CSP is a valuable technology to complement the total RES supply with controllable RES power output [465].
- Since wind, CSP [466] and PV energy can currently not compete with lowest LCOEof fossil-fuelled power plants, there are economic limitations for the penetration of these RES.

As disclosed in Section 2.1, Egypt and Morocco have so far shown strong commitment in reaching their targeted shares of RES penetration. Both countries possess and expand their sizeable hydroelectric capacities (Table 3) which could dip in when PV and wind energy supply levels are low, thereby enabling an overall high share of volatile RES in the electricity supply [467]. The Moroccan and Egyptian targets in electricity supply by RES appear achievable at LCOE competitive to current electricity retail prices. Algeria is starting to harness its very high potentials in both wind and PV energy. Although Tunisian wind and PV energy potentials are much more limited than those of Algeria they are still very suitable for the planned RES targets. The high target of RES share in Libya is possible, yet not much has been done to achieve it (see Section 2.1.4). Hydroelectricity production was low in Tunisia and Algeria was not to be found in Libya, as given in Table 3 for 2009. While sizeable shares of hydroelectricity for controllable RES-supply may not be possible in these countries due to water scarcity, the further development of compressed air energy storage could provide industrial scale energy storage capacities and thereby enhance the flexibility of the power network.

Limiting the emission of Greenhouse Gases (GHG) through RES deployment is essential in mitigating the ongoing climate change [468]. For example, one GWh of electricity generated by the U.S. American non-baseload power plant fleet by RES emits 760 tons of carbon dioxide, 16 tons of methane and 9 tons of nitrous oxide is achieved [469]. Furthermore, 6 TWh/a of zero-emission power substitute the emissions of one coal fired power plant with an equivalent of 4,023,304 metric tons CO2 annually [470].

PV and wind energy are proven low-carbon technologies with high potentials to reduce total GHG emissions [471]. The planned installation of 7200 MW wind energy in Egypt until 2020 is expected to mitigate 12 million tons of CO2 annually and a 12 % emission reduction from the business-as-usual scenario [472]. For the forecasted German power mix of 2020, one GWh produced by wind power would reduce CO2 emissions by 745 t/a, SO2 emissions by 90 t/a and NOX emissions by 109 t/a [473].

Medgrid[474], the Desertec Industrial Initiative [475], the Clean Technology Fund [476], the Green Climate Fund [477] and further national and international institutions provide technological, financial and political support in the deployment of GHG mitigating energy technologies.

8.1.1 Conclusion

The extensive assessment of this diploma thesis revealed that electricity generation potentials of wind and PV energy in North Africa can cover several times the combined energy demand of North Africa [478] and Europe [479] today and in the future [480]. From a cost perspective, North African wind energy is found to be available already at costs below grid parity [481]. This diploma thesis, further discloses that North African PV energy is not yet competitive to the LCOE of the prevalent power mix [482]. However, energy generated by North African roof and ground installed PV systems is forecasted to reach competitiveness in the short- to midterm future. Both, PV and wind energy can conclusively cover the targeted shares of RES in North Africa and Europe. Already today wind energy may even reduce the power network’s LCOE. Combined with CSP, hydroelectric power and novel energy storage technologies, PV and wind energy are suitable to contribute in the transformation towards low-carbon power networks under the constraints of stable electricity supply.

8.2 Outlook

The Cost-Supply assessment of PV and wind energy in North Africa disclosed that sizeable shares of energy supply by PV and wind energy are achievable at viable LCOE. However, future challenges have to be addressed.

Within “Energy Technologies 2050” Wietschel et al. (2010) advise politicians to pursue a broad support strategy for a spectrum of RES technologies. In regard to onshore wind power R&D should further focus on wind condition forecast and remote sensing as well as improvements in aeroacoustics and aerodynamics and wind power grid integration. PV support should focus on the development of new materials and their industrial processing to achieve solar-cell efficiency and lifetime improvements. While demand for utility scale storage technologies and grid infrastructure increases in all scenarios involving a high RES penetration attention, political attention and R&D efforts are urgently required for these fields [483].

Besides their potential to mitigate the ongoing climate change [484], PV and wind energy are valuable in further aspects:

Exporting electricity to the European market may create revenue for North African economies. Selling Carbon Emission Reduction (CER) certificates would create additional benefits, as depicted by the example of Tunisia [485].

Chapter 2 disclosed that resources are allocated heterogeneously in the North African region, with Morocco currently relying largely on fuel and electricity imports. Consequently, increases in power network LCOE due to high shares of domestically produced energy by RES may be outweighed by the strategic benefits: Less dependency on external price fluctuations for electricity and fossil fuel imports as well as reduced exposure to supply bottlenecks. The very low water usage of PV and wind energy technologies compared to all further assessed technologies can contribute to the reduction of water consumption and thereby mitigate water scarcity [486].

However, the decision for a transition towards low-carbon power networks needs to be accompanied by a strong political, social and scientific will. Henning Kagermann, president of the German National Academy of Science and Engineering, applied the quote by Thomas A. Edison - found at the beginning of the study in hand - when addressing the challenges faced by the ongoing German energy transition [487]: “Without realistic plans of implementation, without permanent monitoring, without the disposition to constantly scrutinise the path to the target, without the obligation to reach the target, visions will ultimately remain an utopia.”

May the North African countries have the will to pursue their energy targets and execute the envisioned energy transition!

9 Annex

9.1 Tables and Figures

Table (Anex.) 1: Electricity demand of the North African region [488]

illustration not visible in this excerpt

Table (Anex.) 2: Population of the North African region

illustration not visible in this excerpt

Sources: For the year 2011 [489], 2015 [490] ; for the years 2025 and 2050 mean of [491] and [492], year 2030 linearly interpolated between 2025 and 2050

Table (Anex.) 3: Wind Parks of North Africa

Abbildung in dieser Leseprobe nicht enthalten[493] [494] [495] [496] [497] [498] [499] [500] [501] [502] [503]

Source: [504] & evaluation by the author; *planned or under construction; **planned but not realised; *** assumption by the author

Table (Anex.) 4: Geographic restrictions and land use factors for WT & PV systems

Abbildung in dieser Leseprobe nicht enthalten [505] [506] [507] [508] [509] [510]

Table (Anex.) 5: IUCN Protected Areas Categories System [511]

illustration not visible in this excerpt

Table (Anex.) 6: BirdLife International - Global IBA Criteria [512]

illustration not visible in this excerpt

Table (Anex.) 7: Details on employed wind turbines and processing tools prior to the inclusion of the performance ratio

Abbildung in dieser Leseprobe nicht enthalten [513] [514]

Table (Anex.) 8: Remaining area in km ² and as the percentage of the area included after the geographic assessment of wind energy

Listed by country after the exclusion of economically not suitable areas yielding <1300 FLH; compared by turbine type and hub height, wind regime including standard form parameter k = 2.0 (standard Rayleigh distribution) and k = {2.0; 2.5, 3.0}.

illustration not visible in this excerpt

Table (Anex.) 9: Input values for LCOE calculation for wind turbine installation

WT: Ex-Works wind turbine; Investment incl. BOS comprises costs for Ex-Works WT, transport, installation and is exclusive of tax, land rental and grid connection if not else stated; O&M of Investment [%]: Annual O&M costs as percentage of initial investment

Exchange Rates: 5 year average until 2011-12-30 [515]: US$/€ 1.3906; AU$/€ 1.5876

Abbildung in dieser Leseprobe nicht enthalten [516] [517] [518] [519] [520] [521] [522] [523] [524] [525] [526] [527]

Table (Anex.) 10: Calculation of average best market module efficiency

Abbildung in dieser Leseprobe nicht enthalten [528] [529] [530] [531] [532] [533]

Table (Anex.) 11: Input values for LCOE calculation for PV ground, rooftop and facade installation

Investment including module costs, BOS; excl. tax, grid connection transport, land rental Exchange Rates: 5 year average until 2011-12-30 [534]: US$/€ 1.3906; AU$/€ 1.5876

Abbildung in dieser Leseprobe nicht enthalten [535] [536] [537] [538] [539] [540] [541] [542] [543]

Wind Power - Cost Supply Curves

illustration not visible in this excerpt

Figure (Anex.) 1: Wind power Cost-Supply curve for 2012 - baseline scenario; Enercon-E- 82, 120 m hub height, variable k

illustration not visible in this excerpt

Figure (Anex.) 2: Wind power Cost-Supply curve for 2050; Enercon-E-82, 120 m hub height, variable k

illustration not visible in this excerpt

Figure (Anex.) 3: Wind power Cost-Supply curves for 2012, large wind parks

illustration not visible in this excerpt

Figure (Anex.) 4: Map - LCOE of wind power in 2012; Gamesa G-80, 80 m hub height, variable k

illustration not visible in this excerpt

Figure (Anex.) 5: Map - LCOE of wind power in 2030; Enercon E-82, 120 m hub height, variable k

illustration not visible in this excerpt

Figure (Anex.) 6: Map - LCOE of wind power in 2012; including grid premium for large wind park (1000-3000 MW)

illustration not visible in this excerpt

Figure (Anex.) 7: Map - LCOE of wind power in 2030; including grid premium for midsize wind park (20-90 MW)

illustration not visible in this excerpt

Figure (Anex.) 8: Map - LCOE of wind power in 2030; including grid premium for large wind park (1000-3000 MW)

PV Energy - Cost-Supply Results

illustration not visible in this excerpt

Figure (Anex.) 9: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2030

illustration not visible in this excerpt

Figure (Anex.) 10: Cost-Supply curve depicting marginal LCOE for PV ground installations with financial parameters for 2050

illustration not visible in this excerpt

Figure (Anex.) 11: PV ground system FLH and excluded area

illustration not visible in this excerpt

Figure (Anex.) 12 Map - LCOE of PV ground systems in 2012

illustration not visible in this excerpt

Figure (Anex.) 13: Map - LCOE of PV ground systems in 2030

illustration not visible in this excerpt

Figure (Anex.) 14: Map - LCOE of PV ground systems in 2050

illustration not visible in this excerpt

Figure (Anex.) 15 Map - LCOE of PV ground systems in 2012; including grid premium for large PV parks (1000-3000 MW)

illustration not visible in this excerpt

Figure (Anex.) 16: Map - LCOE of PV ground systems in 2030; including grid premium for midsize PV parks (20-90 MW)

illustration not visible in this excerpt

Figure (Anex.) 17: Map - LCOE of PV ground systems in 2030; including grid premium for large PV parks (1000-3000 MW)

10 References

Aboulnasr, S. (2006): Financing Large Scale Wind Farms in Developing Countries - Zafarana Wind Farm. Egyptian National Committee of the World Energy Council (WEC). http://www.unep.fr/energy/activities/eca/pdf/WS5/WS5-Sherif.pdf (Accessed: 10.11.2011).

Actu-Environment (2011): France-Tunisie: collaboration dans l’énergie photovoltaïque. http://www.actu-environnement.com/ae/news/france-tunisie-energie-solaire- photovoltaique-concentration-cpv-cea-13051.php4 (Accessed: 25.12.2011).

Adiboina, G. (2010): Thin Film Photovoltaic (PV) Cells Market Analysis to 2020. Global Business Intelligence (GBI) Research. http://www.altenergymag.com/emagazine.php?art_id=1557 (Accessed: 30.12.2011).

Agentur für Erneuerbare Energien (2011): Installierte Leistung Windenergie pro km2 - Landesinfo - Föderal Erneuerbar. http://www.foederal- erneuerbar.de/landesinfo/bundesland/MV/kategorie/wind/auswahl/352- installierte_leistun/#goto_352 (Accessed: 21.11.2011).

Aîche-Hamane, L. and Khellaf, A. (2003): Cartographie des ressources éoliennes de L’Algérie, Centre de Développement de Energies Renouvelables. Bulletin des Sciences G é ographiques, Vol. 11, 23-28. (Accessed: 18.11.2011).

Amar, F. and Elamouri, M. (2011): Wind Energy Assessment of the Sidi Daoud Wind Farm - Tunisia. Institute for Engineering Studies of Sfax & INSAT. http://www.intechopen.com/source/pdfs/17122/InTech- Wind_energy_assessment_of_the_sidi_daoud_wind_farm_tunisia.pdf (Accessed: 10.11.2011).

Appleyard, D. (2011): Solar’s Outlook Dazzles | Special Supplement: Large Scale Solar Magazine Article. Renewable Energy World. http://www.renewableenergyworld.com/rea/news/article/2011/05/for-solar- supplementsolars-outlook-stays-bright (Accessed: 2.01.2012).

Arab Union of Electricity (2010): Electrical Networks Arab Countries.

Arab Union of Electricity (2001): Electrical Networks Arab Countries. http://www.auptde.org/NewSite/user/User_Def1.aspx?PID=2014&ID=136.

Barlovento Recursos Naturales and Misr Consulting Engineers (2011): Gulf of EI- Zayt 120 MW Wind Power PlantSpanish Project - Environmental and Social Impact Assessment - Executive Summary. New & Renewable Energy Authority, Egypt. http://www.nrea.gov.eg/1%5Ceng%5CE.S.docx.

Benkhadra, A. (2009): Does Morocco provide a new model for bridging old and new energy systems? Ministry of Energy, Mines, Water and Environment, Morocco. http://www.moroccanembassylondon.org.uk/images/photoConf/Morocco%27s_Annu al_Investment_Conference_2009_-_Minister_Benkhadra_presentation.pdf (Accessed: 7.11.2011).

BirdLife International Birdlife Data Zone - Global IBA Criteria. http://www.birdlife.org/datazone/info/ibacritglob (Accessed: 20.11.2011a).

BirdLife International Birdlife Data Zone - Important Bird Area Search. http://www.birdlife.org/datazone/site/search# (Accessed: 20.11.2011b).

BirdLife International (2003): Windfarms and Birds: An analysis of the effects of windfarms on birds, and guidance oon environmental assessment criteria and site selection issues. Council of Europe. http://www.birdlife.org/eu/pdfs/BirdLife_Bern_windfarms.pdf (Accessed: 19.11.2011).

Bloomberg - New Energy Finance (2009): Wind Turbine Prices Move Down, Shows New Price Index from NEF. Renewable Energy World. http://www.renewableenergyworld.com/rea/news/article/2009/06/wind-turbine-prices- move-down-says-new-price-index (Accessed: 14.01.2012).

Bofinger, D. S., Callies, D., and Scheibe, M. (2011): Studie zum Potential der Windenergienutzung an Land. Bundesverband Windenergie e.V. & Fraunhofer IWES. http://www.wind-energie.de/sites/default/files/download/publication/studie-zum- potenzial-der-windenergienutzung-land/bwe-potenzialstudie_final.pdf (Accessed: 15.11.2011).

Bosch & Partner, Peters Umweltplanung, Klinski, S., and OVGU Magdeburg (2009): Abschätzung der Ausbaupotenziale der Windenergie an Infrastrukturachsen and Entwicklung von Kriterien der Zulässigkeit. Bundesministerium für Umwelt, Naturschutz and Reaktorsicherheit bzw. Forschungszentrum Jülich PTJ. http://www.erneuerbare- energien.de/files/pdfs/allgemein/application/pdf/studie_potenziale_windenergie_bd_1 .pdf (Accessed: 31.12.2011).

Brand, B. and Zingerle, J. (2010): The renewable energy targets of the Maghreb countries: Impact on electricity supply and conventional power markets. Institute of Energy Economics at the University of Cologne (EWI). http://africabusiness.com/2010/12/10/the-renewable-energy-targets-of-the-maghreb- countries-impact-on-electricity-supply-and-conventional-power-markets/ (Accessed: 26.12.2011).

Briney, A. (2012): Sahara Desert. About.com. http://geography.about.com/od/locateplacesworldwide/a/saharadesert.htm (Accessed: 13.01.2012).

Brusaw, S. Solar Roadways - A Real Solution. http://solarroadways.com/ (Accessed: 12.01.2012).

BSW-Solar (2011a): Preisindex. Bundesverband Solarwirtschaft e.V. http://www.solarwirtschaft.de/preisindex (Accessed: 2.01.2012).

BSW-Solar (2011b): Pressemitteilung von SMA: SMA knackt 99-Prozent-Marke beim Wirkungsgrad. Bundesverband Solarwirtschaft e.V. http://www.solarwirtschaft.de (Accessed: 29.12.2011).

Bundesverband WindEnergie (2010): Statistiken. http://www.wind-energie.de/infocenter/statistiken (Accessed: 28.11.2011).

Bürgerinitiative Gegenwind Husarenhof VG Weimar, Urteil v. 9.11.2000 - 1 K 654/00, NuR 2001 S. 536. http://www.gegenwind-husarenhof.de/urteile.html (Accessed: 17.11.2011).

BusinessGreen.com (2010): Bright outlook for world’s largest concentrated solar photovoltaic plant. http://www.businessgreen.com/bg/news/1806737/bright-outlook-worlds-largest- concentrated-solar-photovoltaic-plant (Accessed: 29.12.2011).

C.U.B.E. Engineering Projekt Dernah, Libyen. http://www.cube-engineering.com/referenzen/case-studies/dernah-libyen.html (Accessed: 10.11.2011).

CENER (2009): Atlas Eolien de Tunisie. Agence National pour la Ma î trise de l ’ Energie, Tunis.

CIA (2011): The World Factbook. Central Intelligence Agency. https://www.cia.gov/library/publications/the-world-factbook/ (Accessed: 23.12.2011).

Clean Investment Funds (2009a): Investment Plan for Egypt. Clean Technology Fund. http://www.climateinvestmentfunds.org/cifnet/sites/default/files/Egypt%20CTF%20Inv estment%20Plan%20-%20Endorsed.pdf (Accessed: 30.01.2012).

Clean Investment Funds (2009b): Investment Plan for Morocco. Clean Technology Fund. http://www.climateinvestmentfunds.org/cifnet/sites/default/files/Morocco%20CTF%20 Investment%20Plan%20-%20Endorsed.pdf (Accessed: 30.01.2012).

Cleantech The Giant Wind Gamble. Cleantech magazine, (Issue 03/2011). http://www.cleantechinvestor.com/portal/wind-energy/9552-the-giant-wind- gamble.html (Accessed: 30.01.2012).

Climate Investment Funds Clean Technology Fund. http://www.climateinvestmentfunds.org/cif/node/2 (Accessed: 30.01.2012).

CMAAUV (2010): Resolución del Conseller de Medio Ambiente, Agua, Urbanismo y Vivienda de fecha 12 de mayo de 2010, relativa a la modificación puntual del Plan Especial de la zona 3 del Plan eólico. Conseller de Medio Ambiente, Agua, Urbanismo y Vivienda. http://www.derecho.com/l/bop-castellon/resolucion-conseller-medio-ambiente-agua- urbanismo-vivienda-fecha-12-mayo-2010-relativa-modificacion-puntual-plan- especial-zona-3-plan-eolico/ (Accessed: 17.11.2011).

Crowley, M. (2011): Why CPV? The CPV Value Proposition. CPV Consortium. http://www.cpvconsortium.org/Portals/0/Mark%27s%20presentation.pdf (Accessed: 29.12.2011).

CSNER (2009): Le Solaire photovoltaïque,. Chambre Syndicale Nationale des Energies Renouvelables, Tunis. http://www.csner-tn.com/fr/lien1.php?ID=2 (Accessed: 25.12.2011).

Czisch, G. (2005): Szenarien zur zuküntigen Stromversorgung. Dissertation an der Universität Kassel.

Dalili, N., Edrisy, A., and Carriveau, R. (2009): Renewable and Sustainable Energy Reviews: A review of surface engineering issues critical to wind turbine performance. University of Windsor. http://www.sciencedirect.com/science/article/pii/S1364032107001554#sec4 (Accessed: 7.11.2011).

Danish Energy Agency (2011): Stamdataregister for vindmøller. http://www.ens.dk/da-

DK/Info/TalOgKort/Statistik_og_noegletal/Oversigt_over_energisektoren/Stamdatare gister_vindmoeller/Sider/forside.aspx (Accessed: 21.11.2011).

Decon and Fichtner Consultancy (2008): Environmental Impact Study 200MW WIND FARM At the GULF OF EL ZAYT. Deutsche Energie-Consult Ingenieurgesellschaft (decon) & Fichnter Engineering Services and Consultancy. http://www.bei.org/attachments/pipeline/20070089_eia_study200m_en.pdf (Accessed: 28.11.2011).

Desertec Foundation Desertec Foundation. http://www.desertec.org/en/ (Accessed: 23.01.2012).

DeStatis (2009): Länderübersicht Afrika. Statistisches Bundesamt Deutschland. http://www.destatis.de/jetspeed/portal/cms/Sites/destatis/Internet/DE/Content/Statisti ken/Internationales/InternationaleStatistik/Land/Afrika/Laenderlisten/Afrika,templateId =renderPrint.psml (Accessed: 28.11.2011).

Dii Bringing the Desertec vision into reality. Desertec Industry Initiative. http://www.dii-eumena.com/home.html (Accessed: 24.01.2012).

Dodd, J. (2010): Windpower Monthly - Algeria seals deal for first ever wind farm. WindPower Monthly. http://www.windpowermonthly.com/news/982260/Algeria-seals-deal-first-ever-wind- farm (Accessed: 10.11.2011).

ECB Euro foreign exchange reference rates. European Central Bank. http://www.ecb.int/stats/exchange/eurofxref/html/index.en.html (Accessed: 2.01.2012).

Egyptian Electricity Transmission Company (2010): Samallout / Suez Gulf / Jabal El- Zayt 500 kV Elctrical Interconnection Project - Environmental and Social Assesment and Management Plan. Ministry of Electricity and Energy, Egypt. http://www.eetc.net.eg/Download_Center/projects/attachments_2010_03_31/E.S_Sa mallout_Jabal+El-Zayt.pdf (Accessed: 9.01.2012).

Egyptian Environmental Affairs Agency Important Bird Areas (IBAs) of Egypt. Ministry of State for Environmental Affairs, Egypt. http://www.eeaa.gov.eg/english/main/protect_bird.asp (Accessed: 21.11.2011).

Elgamal, A., Prowell, I., Romanowitz, H., Duggan, J. E., and Jonkman, J. (2010): Earthquake Response Modeling for a Parked and Operating Megawatt-Scale Wind Turbine. U.S. National Renewable Energy Laboratory. http://nees.ucsd.edu/projects/2010-wind-turbine/pubs/2010-nrel-prowell-elgamal- romanowitz-duggan-jonkman.pdf (Accessed: 20.11.2011).

Elsobki, M., Wooders, P., and Sherif, Y. (2009): Clean Energy Investment in Developing Countries:Wind Power in Egypt. International Institute for Sustainable Development. http://www.iisd.org/pdf/2009/bali_2_copenhagen_egypt_wind.pdf (Accessed: 9.01.2012).

Enzili, M. (2011): L’Energie Éolienne Au Maroc (Ressources, Potentiel et Projets). Agence Nationale pour le D é veloppement. http://marokko.ahk.de/fileadmin/ahk_marokko/Veranstaltungen/Delegationsreise_Win d_15.11.2011_Praesentationen/Presentation_ADEREE.pdf (Accessed: 23.12.2011).

Enzili, M. (2007): Les Resources Eoliens du Maroc. Centre de D é veloppment des Energies Renouvelables, Maroc.

EspaceManager (2010): Tunisie: promouvoir l’autoproduction d’électricité par l’énergie solaire photovoltaïque. http://www.espacemanager.com/macro/tunisie-promouvoir-lautoproduction- delectricite-par-lenergie-solaire-photovoltaique.html (Accessed: 25.12.2011).

EWEA (2007a): Avoided emissions and external costs in EU-27 in 2020. European Wind Energy Association. http://www.wind-energy-the-facts.org/en/environment/chapter-5.environmental- benefits-of-wind-energy-in-comparison-to-remaining-electricity-generation- technologies/avoided-emissions-and-external-cost-for-different-wind-deployment- scenarios-in-the-eu27-member-states-in-2020.html (Accessed: 18.01.2012).

EWEA (2011): Cost of Wind Power compared to Other Technologies. European Wind Energy Association. http://www.ewea.org/index.php?id=1643 (Accessed: 30.01.2012).

EWEA (2007b): Large commercial wind turbines. European Wind Energy Association. http://www.wind-energy-the-facts.org/en/part-i-technology/chapter-3-wind-turbine- technology/evolution-of-commercial-wind-turbine-technology/large-commercial-wind- turbines.html (Accessed: 17.01.2012).

EWEA (2009): Wind Energy - The Facts - Contribution of wind power to electricity generation and generation capacity in the EU-27. European Wind Energy Association. http://www.wind-energy-the-facts.org/en/scenarios-and-targets/chapter-3- contribution-of-wind-power-to-electricity-generation-and-generation-capacity-in-the- eu-27/ (Accessed: 15.01.2012).

Exchange-rates.com Moroccan Dirham Exchange Rate Graph - Euro - Historical Exchange Rates. http://www.exchange-rates.org/history/MAD/EUR/G (Accessed: 9.01.2012).

FAO (2007): Digital Soil Map of the World. Food and Agricultural Organisation. http://www.fao.org/geonetwork/srv/en/metadata.show?id=14116 (Accessed: 12.11.2011).

Farrell, J. (2011): Wind Farm Size Hitting a Sweet Spot? Renewable Energy World. http://www.renewableenergyworld.com/rea/blog/post/2011/04/wind-farm-size-hitting- a-sweet-spot (Accessed: 22.01.2012).

Fathi, A. (2008): RES (Wind) Incentives and Barriers in Egypt. New & Renewable Energy Authority, Egypt. http://www.gtz.de/de/dokumente/nrea2008-en-country-case-study-egypt.pdf (Accessed: 23.11.2011).

First Solar First Solar corporate website: Thin film solar modules. http://www.firstsolar.com/Products-and-Services/Products/PV-Modules (Accessed: 30.12.2011).

GAFI (2010): Invest in Egypt - Renewable Energy. General Authority for Investment, Egypt. http://www.gafinet.org/English/SectorsValuePreposition/Renewable%20Energy%20v alue%20proposition-2010.pdf (Accessed: 27.12.2011).

GWEC (2011): Egypt - Total Installed Capacity. Global Wind Energy Council. http://www.gwec.net/index.php?id=122 (Accessed: 10.11.2011).

GWEC (2008): Large scale integration of wind energy into electricity grids. Global Wind Energy Council. http://www.gwec.net/index.php?id=146&L=0 (Accessed: 15.01.2012).

Hamilton, T. (2009): A Cheaper Solar Concentrator. TechnologyReview.com. http://www.technologyreview.com/energy/22204/ (Accessed: 29.12.2011).

Hau, E. (2006): Wind turbines: fundamentals, technologies, application, economics. Springer-Verlag.

Hau, E. (2008): Windkraftanlagen: Grundlagen, Technik, Einsatz, Wirtschaftlichkeit. Springer-Verlag.

Hauser, M. (2003): Dynamische Systeme and Zeitanalyse - Bescheiben von Zeitreihen Kapitel 9. Statistik and Mathematik - WUWien. http://statmath.wu.ac.at/staff/hauser/LVs/MgtSci-DynSysZR/9zrbeschr_hd.pdf (Accessed: 12.01.2012).

Held, A. (2010): Modelling the future development of renewable energy technologies in the European electricity sector using agent-based simulation. Dissertation at the Karlsruhe Institute of Technology.

Hilgers, P. (2010): Calculating Cost-Supply Curves of Concentrating Solar Power Plants in North Africa using a Geographic Information System. Diploma thesis at the Karlsruhe Institute of Technology.

Honsberg, C. and Bowden, S. (2010): PV education: Solar Radiation on a Tilted Surface. PV education. http://pvcdrom.pveducation.org/SUNLIGHT/MODTILT.HTM (Accessed: 12.01.2012).

Hoogwijk, M. (2004): On the Global and Regional Potential of Renewable Energy Sources. Dissertation at the University of Utrecht. http://nws.chem.uu.nl/publica/Publicaties2004/e2004-2.pdf (Accessed: 29.12.2011).

House of Lords (2010): Bill: Wind Turbines (Minimum Distances from Residential Premises). www.parliament.uk. http://www.publications.parliament.uk/pa/ld201011/ldbills/017/11017.1-i.html (Accessed: 17.11.2011).

IEA (2008): Energy Prices & Taxes, Quarterly Statistics, Second Quarter 2008. International Energy Agency.

IEA (2011a): Key World Energy Statistics. International Energy Agency. http://www.iea.org/textbase/nppdf/free/2011/key_world_energy_stats.pdf (Accessed: 25.12.2011).

IEA (2011b): Oil market Report 2011. International Energy Agency. http://omrpublic.iea.org/currentissues/full.pdf (Accessed: 26.12.2011).

IEA (2011c): Renewable Energy - Policy Considerations for Deploying Renewables. http://www.iea.org/papers/2011/Renew_Policies.pdf (Accessed: 15.01.2012).

IEA (2009a): Statistics & Balances. International Energy Agency. http://www.iea.org/stats/index.asp (Accessed: 25.01.2012).

IEA (2010): Technology Roadmap - Solar photovoltaic energy. International Energy Agency. http://www.iea.org/papers/2010/pv_roadmap.pdf (Accessed: 28.12.2011).

IEA (2009b): Technology Roadmap - Wind energy. International Energy Agency. http://www.iea.org/papers/2009/Wind_Roadmap.pdf (Accessed: 15.01.2012).

IEA & OECD (2008): Energy Technology Perspectives 2008 - Scenarios & Strategies to 2050. International Energy Agency & Organisation for Economic Co-operation and Development. http://www.iea.org/techno/etp/ETP_2008_Exec_Sum_English.pdf (Accessed: 24.10.2011).

ISO Maintenance Agency for ISO 3166 country codes - Country names and code elements. International Standards Organization. http://www.iso.org/iso/country_codes/iso_3166_code_lists/country_names_and_code _elements.htm (Accessed: 24.01.2012).

IUCN (2011): IUCN Protected Area Management Categories. International Union for Conservation of Nature and Natural Resources. http://www.iucn.org/about/work/programmes/pa/pa_products/wcpa_categories/ (Accessed: 19.11.2011).

Jacob, T. (2010): Nächster Dammbruch: Erstmals Windpark in Vogelschutzgebiet. Volksinitiative „ Gegen Windräder “. http://www.volksinitiativewindrad.de/index.php?option=com_content&view=article&id =152:naechster-dammbruch-erstmals-windpark-in-vogelschutzgebiet-aufruf-zur- demo-der-windkraftgegner-am-17-von-800-uhr-bis-1000-uhr-auf-dem-brauhausberg- vor-der-einfahrt-des-landtages&catid=25:die-vi (Accessed: 19.11.2011).

James, N. (2001): History of GIS. Undergraduate Thesis at Brown University. http://envstudies.brown.edu/oldsite/Thesis/2001/james/gishistory.html (Accessed: 5.01.2012).

Jarvis, A., Reuter, H. I., and Nelson, A. (2008): Hole-filled seamless SRTM data V4, International Centre for Tropical Agriculture (CIAT). Consortium for Spatial Information (CGIAR-CSI). http://srtm.csi.cgiar.org/ (Accessed: 16.11.2011).

KfW Entwicklungsbank (2011): Projekt - Windenergie - Ägypten - sauberer Strom aus der Wüste. http://www.kfw- entwicklungsbank.de/ebank/DE_Home/Laender_und_Programme/Nordafrika_und_n aher_Osten/Aegypten/Projekt_-_Windenergie.jsp (Accessed: 14.01.2012).

Komoto, K., Ito, M., and van der Vleuten, P. (2009): Energy from the desert, Earthscan.

Köpke, R. (2011): Auf Augenhöhe angekommen - Siemens forciert das ohnehin schon erfolgreiche Windkraft-Gechäft. Dafür soll die neue Zentrale sorgen, die ihren Sitz in Hamburg hat. Energie & Management - Zeitung für den Energiemarkt, Vol. 21/2011.

Kost, C. and Schlegl, T. (2010): Studie Stromgestehungskosten erneuerbare Energien. Fraunhofer Institute for Solar Energy Systems ISE. http://www.ise.fraunhofer.de/veroeffentlichungen/studie-stromgestehungskosten- erneuerbare-energien (Accessed: 9.11.2011).

LAG-VSW (2007): Abstandsregelung für Winenergieanlagen zu bedeutsamen Vogellebensräumen sowie Brutplätzen ausgewählter Vogelarten -

Länderarbeitsgemeinschaft der Vogelschutzwarten (LAG-VSW). Berichte zum Vogelschutz, 44/2007. http://www.drv-web.de/pdf/downloads/bzv44/LAG-VSW-Abstand.pdf (Accessed: 19.11.2011).

Large commercial wind turbines http://www.wind-energy-the-facts.org/en/part-i-technology/chapter-3-wind-turbine- technology/evolution-of-commercial-wind-turbine-technology/large-commercial-wind- turbines.html (Accessed: 17.01.2012).

Li, H. and Chen, Z. (2007): Overview of different wind generator systemsand their comparisons. Institute of Energy Technology, Aalborg University, Denmark. http://sklpe.cqu.edu.cn/KXYJ/ei/ei48.pdf (Accessed: 6.11.2011).

Llorente, R., Lacal, R., and Aguado, M. (2010): Renewable and Sustainable Energy Reviews: Power electronics evolution in wind turbines - A market-based analysis. European Commission & Centro Nacional de Energias Renovables, Spain. http://www.sciencedirect.com/science/article/pii/S1364032111002978 (Accessed: 7.11.2011).

Luque, A. and Hegedus, S. (2011): Handbook of Photovoltaic Science and Engineering, Wiley.

Macknick et al. (2011): A Review of Operational Water Consumption and Withdrawal Factors for Electricity Generation Technologies. U.S. National Renewable Energy Laboratory. http://www.nrel.gov/docs/fy11osti/50900.pdf (Accessed: 23.01.2012).

Madrigal, M. and Stoft, S. (2011): Transmission Expansion forRenewable Energy Scale-UpEmerging Lessons andRecommendations. The World Bank Group - The Energy and Mining Sector Board. http://www.esmap.org/esmap/sites/esmap.org/files/DP%2026%20transmission%20e xpansion%20text%209-15-11web_SMALL.pdf (Accessed: 9.01.2012).

Math Pages Weibull Analysis. http://www.mathpages.com/home/kmath122/kmath122.htm (Accessed: 14.01.2012). Maxing Out Media (2008): Red Sea Chronicles. http://www.maxingout.com/red_sea_windsurfing.htm (Accessed: 12.01.2012).

Medgrid Medgrid. http://www.medgrid-psm.com/en/ (Accessed: 23.01.2012).

MEM Algerie (2007): Guidelines to Renewable Energies. Ministry of Energy and

Mines, Algeria. http://www.mem-algeria.org/fr/enr/Guide_Enr_ang.pdf (Accessed: 26.12.2011).

MEM Algerie Présentation des 20 villages solairesdans le Sud Algérien. Ministry of Energy and Mines, Algeria. http://www.mem-algeria.org/fr/enr/energie%20solaire/pres_20-villages.htm (Accessed: 26.12.2011).

MEM Maroc (2008): Ministère de l’Energie, des Mines, de l’Eau et de l’Environment: Les Energies Renouvelables Au Maroc. Minist è re de l ’ Energie, des Mines, de l ’ Eau et de l ’ Environnement, Maroc. http://www.mem.gov.ma/revue_presse/questPDF/Energies-renouvelables.pdf (Accessed: 7.11.2011).

MERRA (2011): Daily & Montly IAU 2d atmospheric single-level diagnostics 1979- 2011. Goddard Earth Sciences, Data & Information Services Center by NASA. http://disc.sci.gsfc.nasa.gov/daac-bin/FTPSubset.pl?LOOKUPID_List=MATMNXSLV (Accessed: 23.12.2011).

Merziger et al. (2007): Formlen + Hilfen zur Höheren Mathematik, Binomi Verlag.

Meyers and Meneveau (2011): Optimal turbine spacing in fully developed wind farm boundary layers, Wiley. http://onlinelibrary.wiley.com/doi/10.1002/we.469/abstract (Accessed: 8.01.2012).

Mills and Wiser (2010): Implications of Wide-AreaGeographic Diversity for Short- Term Variability of Solar Power. Berkeley National Laboratory. http://eetd.lbl.gov/ea/emp/reports/lbnl-3884e.pdf (Accessed: 23.01.2012).

Mortensen et al. (2005): Wind Atlas for Egypt, NREA, RISOE.

Morthorst, E., Auer, H., Garrad, A., and Blanco, I. (2009): Wind Energy - The Facts, Part III, The Economics of Wind Power, European Wind Energy Association. http://www.wind-energy-the-facts.org/documents/download/Chapter3.pdf (Accessed: 14.01.2012).

Moussi, S. (2008): Le secteur de l’électrcité en Algérie - Situation actuelle et perspectives in Sonelgaz, Algerie.

Müller et. al (2011): Renewable Energy - Policy Consideration for Deploying Renewables. International Energy Agency. http://www.iea.org/papers/2011/Renew_Policies.pdf (Accessed: 22.01.2012).

Muller, J. and Errichello, R. (2002): Oil Cleanliness in Wind Turbine Gearboxes. Machinery Lubrication. http://www.machinerylubrication.com/Read/369/wind-turbine-gearboxes-oil (Accessed: 6.11.2011).

NASA (2010): ReadMe Document for MERRA Data Products. Goddard Earth Sciences, Data & Information Services Center by NASA. http://disc.sci.gsfc.nasa.gov/mdisc/documentation/README.MERRA.pdf (Accessed: 6.01.2012).

National Geographic Maps - National Geographic. http://maps.nationalgeographic.com/maps (Accessed: 25.01.2012).

Neji, L., Borup, M., and Mayer-Spohn, O. (2007): Cost Development of Energy Technologies - an Analysis based on Experience Curves, Final Report of the EC Project NEEDS.

Nemet, G. (2007): Learning Curves for Photovoltaics. International Energy Agency. http://www.iea.org/work/2007/learning/Nemet_PV.pdf (Accessed: 13.01.2012).

Neuhof, F. (2011): Egypt shows faith in renewable energy with wind farm tender. The National. http://www.thenational.ae/business/energy/egypt-shows-faith-in-renewable-energy- with-wind-farm-tender (Accessed: 15.01.2012).

New et al. (2002a): A high-resolution data set of surface climate overglobal land areas. Climatic Research Unit. http://www.cru.uea.ac.uk/cru/data/hrg/tmc/new_et_al_10minute_climate_CR.pdf (Accessed: 7.01.2012).

New et al. (2002b): CRU CL 2.0. Climatic Research Unit. http://www.cru.uea.ac.uk/cru/data/hrg/tmc/ (Accessed: 12.11.2011).

Nikionok-Ehrlich, A. (2011): Handeln statt halluzinieren. Energie & Management - Jahresmagazin 2011.

NIMA (2003): Vector Map 0: Updated and improved version of the National Imagery and Mapping Agency’s (NIMA) Digital Chart of the World (DCW). http://www.mapability.com/index1.html (Accessed: 17.11.2011).

NOAA Tsunami Runup Simplified Database Search. National Geophysical Data Center. http://www.ngdc.noaa.gov/nndc/struts/form?t=101650&s=167&d=166 (Accessed: 20.11.2011).

NREA (2011): Egyptian Renewable Energy Activities and Strategy 20 x ’20. New & Renewable Energy Authority, Egypt. http://www.dii- eumena.com/fileadmin/Daten/AC_2011/Presentations/4a%20Regulatory/Dii_Cairo% 202011_Rakha_Regulatory%20framework%20for%20desert%20energy%20projects %20in%20MENA.pdf (Accessed: 30.01.2012).

NREA (2010): NREA Projects & Strategies. New & Renewable Energy Authority, Egypt. http://www.nrea.gov.eg/english1.html (Accessed: 10.11.2011).

NREL (2006): Solar: monthly and annual average direct normal (DNI), global horizontal (GHI), latitude tilt, and diffuse data and GIS data at 40km resolution for Africa. U.S. National Renewable Energy Laboratory. http://swera.unep.net/index.php?id=35&idx=333 (Accessed: 30.12.2011).

O.N.E. (2011): Grand Projets - Production - Éolienne, Office National de l’Electricité. Office National de l ’É lectricit é du Maroc. http://www.one.org.ma/fr/pages/interne.asp?esp=2&id1=5&id2=54&id3=44&t2=1&t3= 1 (Accessed: 28.11.2011).

O.N.E. (2008): Quand l’énergie... vient du vent, Office National de l’Electricité. Office National de l ’É lectricit é du Maroc. http://www.one.org.ma/fr/pdf/Parcs_eoliens.pdf (Accessed: 7.11.2011).

Optics.org (2010): Concentrating PV: risky business or solar star? http://optics.org/indepth/1/1/14 (Accessed: 2.01.2012).

Pican, E., Omerdic, D., Toal, D., and Leahy, M. (2011): Analysis of parallel connected synchronous generators in a novel offshore wind farm model. University of Limerick & University of Ireland. http://www.sciencedirect.com/science/article/pii/S0360544211006426 (Accessed: 27.01.2012). PV Power Plants (2011): PV Power Plants 2011 - Industry Guide: Size Matters. http://www.pv-power-plants.com/index.php?id=973 (Accessed: 31.12.2011).

Q-Cells (2011): Datenblatt Q-Cells QPEAK. http://www.q-cells.com/uploads/tx_abdownloads/files/Q- Cells_QPEAK_Datenblatt_DE_2011-08_Rev04_WEB.pdf (Accessed: 29.12.2011).

Q-Cells Q-Cells Corporate Website: Solarmodule. http://www.q-cells.com/produkte/solarmodule.html (Accessed: 30.12.2011).

Quaschning, V. (2001): Solar Power - Photovoltaics or Solar Thermal Power Plants? VGB Congress Power Plants 2001.

http://www.volker-quaschning.de/downloads/VGB2001.pdf (Accessed: 27.12.2011).

Ragheb, A. and Ragheb, M. (2010): Wind Turbine Gearbox Technologies. University of Illinois. https://netfiles.uiuc.edu/mragheb/www/Wind%20Power%20Gearbox%20Technologie s.pdf (Accessed: 3.11.2011).

Ragwitz, M., Kost, C., and Gazzo, A. (2011): Middle East and North Africa Region Assessment of the Local Manufacturing Potentialfor Concentrated Solar Power (CSP) Projects. Fraunhofer ISI, Fraunhofer ISE, Ernst & Young. http://arabworld.worldbank.org/content/dam/awi/pdf/CSP_MENA__report_17_Jan201 1.pdf (Accessed: 3.11.2011).

RCREEE (2010a): [...] Assessment of National Regulations and Incentives for Renewable Energy and Energy Efficiency: Country Report Libya. Regional Center for Renewable Energy and Energy Efficiency. http://www.rcreee.org/Studies/Danida_Country_Studies/Libya.pdf (Accessed: 26.12.2011).

RCREEE (2010b): [...] Assessment of National Regulations and Incentives forRenewable Energy and Energy Efficiency: Country Report Libya. Regional Center for Renewable Energy and Energy Efficiency. http://www.rcreee.org/Studies/Danida_Country_Studies/Tunisia_EN.pdf (Accessed: 26.01.2012).

RCREEE Regional Center for Renewable Energy and Energy Efficiency. http://www.rcreee.org/ (Accessed: 26.12.2011).

Republique Algérienne, P. (2004): Conventions et accords internationaux - lois et decrets arretes, decicions, avis, commuiactions et annonces. Journal Officiel de la Republique Alg é rienne. http://www.solarpaces.org/_Libary/AlgerianFeedInLaw.pdf (Accessed: 15.01.2012).

SaharaWind.com Windparks in Nordafrika. http://www.saharawind.com/index.php?Itemid=55&id=38&option=com_content&task= view&lang=de (Accessed: 7.11.2011).

Saleh, I. (2006): Prospects of Renewable Energy in Libya. Al-Fateh University. www.irsol.ch/spse/spse-download.php?pid=20 (Accessed: 27.12.2011).

Schermeyer, H. (2011): Analyse der marokkanischen Ausbaupläne für Windkraftanlagen and CSP Kraftwerke unter regelungstechnischen and wirtschaftlichen Gesichtspunkten. Diploma thesis at the Karlsruhe Institute of Technology.

Scholz, Y. (2010): Möglichkeiten and Grenzen der Integration verschiedener regenerativer Energiequellen zu einer 100% regenerativen Stromversorgung der Bundesrepublik Deutschland bis zum Jahr 2050. Deutsches Zentrum für Luft- and Raumfahrt.

Semiconductor Today (2011): First Solar raises CdTe PV cell efficiency record from 16.7% to 17.3%. http://www.semiconductor- today.com/news_items/2011/JULY/FIRSTSOLAR_260711.html (Accessed: 30.12.2011).

Sensfuss et al. (2003): Fluktuationen der Windenergie and deren Vorhersagbarkeit bei einem verstärkten Ausbau des Offshore Anteils in Deutschland bis 2020. Fraunhofer ISI. http://www.isi.fhg.de/isi-de/e/download/publikationen/Sensfuss_IEWT_2003.pdf (Accessed: 24.01.2012).

Sharp Sharp Corporate Website: PV Module. http://www.sharp.de/cps/rde/xchg/de/hs.xsl/-/html/solarmodule.htm (Accessed: 30.12.2011).

Smithsonian National Museum of Natural History Global Volcanism Program | Volcanoes of the World | Volcanoes of Africa and the Red Sea | Africa (northern). http://www.volcano.si.edu/world/region.cfm?rnum=0205 (Accessed: 20.11.2011).

Smithsonian National Museum of Natural History (2011): This Dynamic Planet - World Map of Volcanoes, Earthquakes, Impact Craters, and Plate Tectonics. http://mineralsciences.si.edu/tdpmap/index.htm (Accessed: 20.11.2011).

Soitec (2011): Soitec’s Concentrix technology. http://www.soitec.com/en/technologies/concentrix/ (Accessed: 29.12.2011).

Solaranlagen Portal (2011): Monokristallin oder Polykristallin - Solarzellen im Vergleich. http://www.solaranlagen-portal.com/solarmodule/systeme/vergleich (Accessed: 2.01.2012).

Solarbuzz (2010): Global PV Market. http://solarbuzz.com/facts-and-figures/market-facts/global-pv-market (Accessed: 28.12.2011).

Solarbuzz Solar Electricity Prices. http://www.solarbuzz.com/facts-and-figures/retail-price-environment/solar-electricity- prices (Accessed: 10.08.2011).

Sonelgaz Développement des énergies renouvelables. http://www.sonelgaz.dz/article.php3?id_article=183 (Accessed: 26.12.2011).

Sørensen, B. (1999): Long term scenarios for global energy demand and supply. Four global greenhouse mitigation scenarios. Roskilde University.

STEG (2010): L’Eolien en Tunisie. Soci é t é Tunisienne de l ’ Electricit é et du Gaz. http://www.steg.com.tn/dwl/Eolien_Tunisie.pdf (Accessed: 10.11.2011).

STEG (2008): Reseau electrique tunisien. Soci é t é Tunisienne de l ’ Electricit é et du Gaz.

Steyer, C.-D. (2009): Kampf gegen Windräder. Der Tagesspiegel, Berlin. http://www.tagesspiegel.de/berlin/brandenburg/kampf-gegen- windraeder/1546864.html (Accessed: 17.11.2011).

Strube, O. (2010): Top 10: Ten Largest Solar PV Manufacturers for 2009. Renewable Energy World. http://www.renewableenergyworld.com/rea/blog/post/2010/06/top-10-ten-largest- solar-pv-companies (Accessed: 30.12.2011).

Suleske, A. A. (2010): Performance Degradation of Grid-Tied Photovoltaic Modules in a Desert Climatic Condition. Master thesis at Arizona State University. http://repository.asu.edu/attachments/56112/content/Suleske_asu_0010N_10107.pdf (Accessed: 29.12.2011).

Suntech Suntech - Corporate Website. http://am.suntech-power.com/ (Accessed: 30.12.2011).

SWERA (2005): Wind Speed At 50 m Above The Surface Of The Earth (m/s). NASA Surface meteorology and Solar Energy (SSE). http://swera.unep.net/index.php?id=7 (Accessed: 12.11.2011).

Taiyo Kogyo Corporation (2012): Transparent Photovoltaic Glass. http://www.taiyokogyo.com/tss/ (Accessed: 21.01.2012).

The World Bank (2010): Project Appraisal Document on a Proposed Loan in the Amount of US$ 70 Mio and a Clean Energy Fund Loan in the Amount of US$ 150 Mio. To the Government of Egypt for a EG-Wind Power Development. The World Bank. http://www.icex.es/cambioclimatico/web/egipto/Mastercopy%20of%20Wind%20PAD %20(April%2015).pdf (Accessed: 9.01.2012).

Trieb et al. (2007): Concentrating Solar Power for Seawater Desalination. Deutsches Zentrum für Luft- and Raumfahrt. http://www.dlr.de/tt/Portaldata/41/Resources/dokumente/institut/system/projects/aqua -csp/AQUA-CSP-Full-Report-Final.pdf (Accessed: 23.01.2012).

Trieb, F. (2005): MED-CSP. Deutsches Zentrum für Luft- and Raumfahrt. http://www.dlr.de/tt/desktopdefault.aspx/tabid-2885/4422_read-6573/ (Accessed: 3.11.2011).

Trieb, F., O’Sullivan, M., Pegger, T., Schillings, C., and Krewitt, W. (2009): Characterisation of Solar Electricity Import Corridors from MENA to Europe. Deutsches Zentrum für Luft- and Raumfahrt.

Tzimas et al. (2009): Future Fossil Fuel Electricity Generation in Europe: Options and Consequences. European Commission JRC. http://ec.europa.eu/dgs/jrc/downloads/jrc_reference_report_200907_fossil_fuel_elect ricity.pdf (Accessed: 4.01.2012).

U.S. Department of Defense (1992): Military Specification Digital Chart of the World (DCW). http://earth-info.nga.mil/publications/specs/printed/89009/89009_DCW.pdf (Accessed: 6.01.2012).

U.S. Department of Energy (2011): 2010 Wind Technologies Market Report. http://www1.eere.energy.gov/wind/pdfs/51783.pdf (Accessed: 14.01.2012).

U.S. Department of Energy (2006): Energy Efficiency and Renewable Energy: Solar Energy Technologies Program - Multi-Year Program Plan 2007-2011. http://www1.eere.energy.gov/solar/pdfs/set_myp_2007-2011_proof_1.pdf (Accessed: 29.12.2011).

U.S. EPA (2011): Calculations and References. U. S. Environmental Protection Agency. http://www.epa.gov/cleanenergy/energy-resources/refs.html#coalplant (Accessed: 18.01.2012).

U.S. EPA (2010): Year 2007 GHG Annual Output Emission Rates. U. S. Environmental Protection Agency. http://www.epa.gov/cleanenergy/documents/egridzips/eGRID2010V1_1_year07_GH GOutputrates.pdf (Accessed: 18.01.2012).

UCSDTV (2010): Statements of Elgamal et al. within the Movie „Seismic Testing of Wind Turbines for Renewable Energy“. University of California, San Diego. http://www.youtube.com/watch?v=2IN486JYkt8 (Accessed: 20.11.2011).

UN Pop (2009): World Population Prospects 2009. United Nations Population Division. http://www.un.org/esa/population/publications/wpp2008/wpp2008_text_tables.pdf (Accessed: 3.01.2012).

UN Stats (2008): Population by sex, annual rate of population increase, surface area and density. United Nations Statistics Division. http://unstats.un.org/unsd/demographic/products/dyb/dyb2008/Table03.pdf (Accessed: 21.11.2011).

UNFCCC (2011): Green Climate Fund report of the Transitional Committee. United Nations Framework Convention on Climate Change. http://unfccc.int/files/meetings/durban_nov_2011/decisions/application/pdf/cop17_gcf .pdf (Accessed: 25.01.2012).

UNFCCC (2012): Kyoto Protocol. United Nations Framework Convention on Climate Change. http://unfccc.int/kyoto_protocol/items/2830.php (Accessed: 25.01.2012).

US PRB (2011): 2011 World Population Data Sheet. United States Population Fererence Bureau. http://www.prb.org/Publications/Datasheets/2011/world-population-data-sheet.aspx (Accessed: 3.01.2012).

Verwaltungsgericht Minden (2008): Urteil 11 K 1989 / 08. http://www.jusmeum.de/urteile/vg_minden/fbab5d4a60aac621ea04894d1f9c070ad06 9379d85e393490366daf7e42f8a4d (Accessed: 17.11.2011).

Verwaltungsgericht Neustadt/Wstr. (2005): Urteil vom 11. April 2005 - 3 K 1624/04.NW. http://www3.justiz.rlp.de/rechtspr/DisplayUrteil.asp?rowguid=%7BCF4A159C-6DFA- 41DF-A229-E98AA94ED50A%7D (Accessed: 17.11.2011).

Vestas (2011): Delivered Vestas turbines as of 30 June 2011. http://www.vestas.com/en/about-vestas/profile/track-records/results-%E2%80%93- countries-and-regions.aspx (Accessed: 10.11.2011).

Watts, M., Passey, R., and Johnston, W. (2011): PV in Australia 2010, Australian PV Association. International Energy Agency. http://www.apva.org.au/sites/default/files/documents/APVA%20Status%20Reports/P V%20in%20Australia%202010.pdf (Accessed: 2.01.2012).

WDPA (2010): 2010 World Database on Protected Areas Annual Release. http://www.wdpa.org/AnnualRelease.aspx (Accessed: 19.11.2011).

Wietschel and Genoese (2011): Großtechnische Stromspeicherung im Vergleich. Energiewirtschaftliche Tagesfragen, year 61 issue 6.

Wietschel et al. (2010a): Energietechnologien 2050 - Schwerpunkte für Forschung and Entwicklung - Politikbericht, Fraunhofer Verlag.

Wietschel et al. (2010b): Energietechnologien 2050 - Schwerpunkte für Forschung and Entwicklung - Technologiebericht, Fraunhofer Verlag.

Wind Power Program (2011): The wind turbine database. http://www.wind-power-program.com/download.htm#database (Accessed: 17.01.2012).

Wwindea (2011): Half-year Report 2011. World Wind Energy Association. http://wwindea.org/home/images/stories/publications/half_year_report_2011_wwea.p df (Accessed: 15.01.2012).

Yingli Solar Yingli Solar Corporate Website: Produkte. http://www.yinglisolar.com/de/produkte/ (Accessed: 30.12.2011).

[...]


[1] In case no particular source is provided, the respective illustration - figure or table - is based on own calculations and design.

[2] In case no particular source is provided, the respective equation is based on own development.

[3] IEA & OECD (2008) p. 2 ff.

[4] UNFCCC (2012)

[5] UNFCCC (2011)

[6] Climate Investment Funds

[7] EWEA (2009)

[8] Ragwitz et al. (2011)

[9] IEA (2009) (b)

[10] National Geographic

[11] ISO

[12] James (2001)

[13] DeStatis (2009)

[14] Hilgers (2010)

[15] DeStatis (2009)

[16] CIA (2011)

[17] Brand and Zingerle (2010)

[18] IEA (2011) (a)

[19] IEA (2011) (b)

[20] O.N.E. (2008)

[21] IEA (2009) (a)

[22] IEA (2011) (a)

[23] IEA (2009) (b)

[24] Tzimas et al. (2009)

[25] Hilgers (2010)

[26] RCREEE

[27] IEA (2009) (a)

[28] IEA (2011) (c); p. 34f.

[29] Neuhof (2011)

[30] Republique Algérienne (2004)

[31] IEA (2009) (a)

[32] Benkhadra (2009)

[33] IEA (2011) (a)

[34] O.N.E. (2008)

[35] Benkhadra (2009)

[36] MEM Maroc (2008)

[37] Brand and Zingerle (2010)

[38] O.N.E. (2008)

[39] Clean Investment Funds (2009) (b)

[40] Brand and Zingerle (2010)

[41] MEM Maroc (2008)

[42] Brand and Zingerle (2010)

[43] Komoto et al. (2009)

[44] MEM Maroc (2008)

[45] MEM Maroc (2008)

[46] MEM Algerie (2007)

[47] Brand and Zingerle (2010)

[48] MEM Algerie (2007)

[49] Dodd (2010)

[50] Sonelgaz

[51] MEM Algerie (2007)

[52] MEM Algerie

[53] Komoto et al. (2009) p. 139

[54] Brand and Zingerle (2010)

[55] RCREEE (2010) (b)

[56] STEG (2010)

[57] CSNER (2009)

[58] EspaceManager (2010)

[59] Actu-Environment (2011)

[60] RCREEE (2010) (a)

[61] RCREEE (2010) (a)

[62] Saleh (2006)

[63] RCREEE (2010) (a)

[64] C.U.B.E. Engineering

[65] Saleh (2006)

[66] RCREEE (2010) (a)

[67] RCREEE (2010) (a)

[68] NREA (2010)

[69] Neuhof (2011)

[70] NREA (2010)

[71] Clean Investment Funds (2009) (a)

[72] NREA (2010)

[73] GWEC (2011)

[74] GAFI (2010)

[75] NREA (2011)

[76] Elsobki et al. (2009)

[77] NREA (2010)

[78] NREA (2011)

[79] Hau (2006) p. 67 ff.

[80] Hau (2006) p. 13 ff.

[81] Hau (2008) p. 20 f.

[82] Hau (2006) p. 23 ff.

[83] Hau (2008) p. 82

[84] Hau (2008) p. 82

[85] Hau (2008) p. 82

[86] Hau (2008) p. 82

[87] Hau (2006) p. 44 f.

[88] Hau (2008), p. 797

[89] Hau (2006) p. 71

[90] Hau (2006) p. 73

[91] Li and Chen (2007)

[92] Hau (2006) p. 67 ff.

[93] Hau (2008) p. 94 ff.

[94] Hau (2006) p. 122

[95] Hau (2008) p. 251 ff.

[96] Hau (2008) p. 779

[97] Hau (2006) p. 112 f.

[98] Llorente et al. (2010)

[99] Llorente et al. (2010)

[100] Li and Chen (2007)

[101] Li and Chen (2007)

[102] Li and Chen (2007)

[103] A. Ragheb and M. Ragheb (2010)

[104] Hau (2008) p. 251 ff.

[105] Li and Chen (2007)

[106] Li and Chen (2007)

[107] A. Ragheb and M. Ragheb (2010)

[108] Köpke (2011)

[109] Cleantech

[110] Pican et al. (2011)

[111] Li and Chen (2007)

[112] Li and Chen (2007)

[113] Llorente et al. (2010)

[114] GWEC (2008)

[115] Hau (2006) p. 762

[116] Muller and Errichello (2002)

[117] Dalili et al. (2009)

[118] Luque and Hegedus (2011), p. 11 ff.

[119] Luque and Hegedus (2011), p. 64 ff.

[120] Luque and Hegedus (2011), p. 4

[121] Luque and Hegedus (2011), p. 64 ff.

[122] Luque and Hegedus (2011), p. 359 ff.

[123] Luque and Hegedus (2011), p. 3 ff.

[124] Quaschning (2001)

[125] PV Power Plants (2011)

[126] IEA (2010)

[127] Solarbuzz (2010)

[128] Adiboina (2010)

[129] Luque and Hegedus (2011), p. 23 ff.

[130] Luque and Hegedus (2011), p. 23 ff.

[131] Trieb (2005)

[132] IEA (2010)

[133] Luque and Hegedus (2011), p. 359 ff.

[134] Crowley (2011)

[135] Soitec (2011)

[136] BusinessGreen.com (2010)

[137] Optics.org (2010)

[138] Hamilton (2009)

[139] Hau (2008), p. 593

[140] Mortensen et al. (2005)

[141] Czisch (2005)

[142] NIMA (2003)

[143] U.S. Department of Defense (1992)

[144] Hoogwijk (2004) p. 897

[145] Hau (2008)

[146] Czisch (2005)

[147] New et al. (2002) (b)

[148] New et al. (2002) (b)

[149] Hoogwijk (2004)

[150] New et al. (2002) (a)

[151] Hau (2008), p. 506

[152] MERRA (2011)

[153] NASA (2010)

[154] MERRA (2011)

[155] SWERA (2005)

[156] SWERA (2005)

[157] Mortensen et al. (2005)

[158] Mortensen et al. (2005)

[159] CENER (2009)

[160] CENER (2009)

[161] Aîche-Hamane and Khellaf (2003)

[162] Hau (2008), p. 506

[163] Aîche-Hamane and Khellaf (2003)

[164] C.U.B.E. Engineering

[165] Enzili (2007)

[166] Enzili (2007)

[167] Merziger et al. (2007), p. 198, adapted

[168] Hauser (2003)

[169] Hau (2008), p. 514

[170] Math Pages

[171] Hau (2008), p. 514

[172] Math Pages

[173] Math Pages

[174] Hoogwijk (2004)

[175] Jarvis et al. (2008)

[176] Trieb et al. (2009)

[177] NIMA (2003)

[178] CMAAUV (2010)

[179] Bofinger et al. (2011)

[180] Steyer (2009)

[181] House of Lords (2010)

[182] O.N.E. (2008)

[183] Japan Studie Zafarana

[184] Bürgerinitiative Gegenwind Husarenhof

[185] Verwaltungsgericht Minden (2008)

[186] Verwaltungsgericht Neustadt/Wstr. (2005)

[187] Bosch & Partner et al. (2009)

[188] NIMA (2003)

[189] WDPA (2010)

[190] Trieb et al. (2009)

[191] Held (2010)

[192] Egyptian Environmental Affairs Agency

[193] BirdLife International (2003)

[194] LAG-VSW (2007)

[195] Jacob (2010)

[196] O.N.E. (2008)

[197] Decon and Fichtner Consultancy (2008)

[198] Barlovento Recursos Naturales and Misr Consulting Engineers (2011)

[199] Decon and Fichtner Consultancy (2008)

[200] BirdLife International

[201] Smithsonian National Museum of Natural History

[202] NOAA

[203] Smithsonian National Museum of Natural History (2011)

[204] UCSDTV (2010)

[205] Elgamal et al. (2010)

[206] Hau (2008) p. 596

[207] Hoogwijk (2004)

[208] Hoogwijk (2004)

[209] UN Stats (2008)

[210] Danish Energy Agency (2011)

[211] Agentur für Erneuerbare Energien (2011)

[212] Bofinger et al. (2011)

[213] Maxing Out Media (2008)

[214] Hau (2008), p. 732 ff.

[215] Meyers and Meneveau (2011)

[216] STEG (2010)

[217] Hau (2008), p. 740

[218] Hau (2008), p. 981f.

[219] Bundesverband WindEnergie (2010)

[220] Morthorst et al. (2009)

[221] Large commercial wind turbines

[222] Wind Power Program (2011)

[223] EWEA (2007) (b)

[224] Hau (2008), p. 428 ff.

[225] Hau (2008), p. 569

[226] Bofinger et al. (2011)

[227] Hau (2008), p. 569

[228] Hau (2008), p. 569

[229] Tzimas et al. (2009)

[230] Trieb et al. (2009)

[231] Bofinger et al. (2011)

[232] Hoogwijk (2004), p. 126

[233] Enzili (2007)

[234] Morthorst et al. (2009)

[235] Aboulnasr (2006)

[236] STEG (2010)

[237] Czisch (2005)

[238] Hoogwijk (2004)

[239] Held (2010)

[240] Kost and Schlegl (2010)

[241] Schermeyer (2011)

[242] Hau (2008), p. 835 ff.

[243] KfW Entwicklungsbank (2011)

[244] The World Bank (2010)

[245] Kost and Schlegl (2010)

[246] Elsobki et al. (2009)

[247] Trieb et al. (2009)

[248] Trieb et al. (2009)

[249] Kost and Schlegl (2010)

[250] Held (2010)

[251] Schermeyer (2011)

[252] Hau (2008), p. 622

[253] Bloomberg - New Energy Finance (2009)

[254] Neji et al. (2007)

[255] Morthorst et al. (2009)

[256] IEA (2009) (b)

[257] IEA (2009) (b)

[258] Wwindea (2011)

[259] Kost and Schlegl (2010)

[260] Kost and Schlegl (2010)

[261] EWEA (2011)

[262] Schermeyer (2011)

[263] Luque and Hegedus (2011), p. 62f., p. 702f. & p. 912f.

[264] Luque and Hegedus (2011), p. 63

[265] NREL (2006)

[266] Held (2010)

[267] Held (2010)

[268] Honsberg and Bowden (2010)

[269] Luque and Hegedus (2011), p. 1008

[270] Trieb (2005), p. 41 ff.

[271] Trieb (2005), p. 41 ff.

[272] NIMA (2003)

[273] Brusaw

[274] Trieb (2005)

[275] WDPA (2010)

[276] Sørensen (1999)

[277] Held (2010), p. 72

[278] NIMA (2003)

[279] Hilgers (2010)

[280] Trieb et al. (2009)

[281] Held (2010)

[282] Strube (2010)

[283] IEA (2010)

[284] Adiboina (2010)

[285] Hoogwijk (2004)

[286] IEA (2010)

[287] Luque and Hegedus (2011), p. 296

[288] Luque and Hegedus (2011), p. 296

[289] Luque and Hegedus (2011), p. 297

[290] Q-Cells (2011)

[291] Briney (2012)

[292] Luque and Hegedus (2011), p. 104

[293] Luque and Hegedus (2011), p. 295f.

[294] Crowley (2011)

[295] Suleske (2010)

[296] Suleske (2010)

[297] Luque and Hegedus (2011), p. 14

[298] U.S. Department of Energy (2006)

[299] BSW-Solar (2011) (b)

[300] Czisch (2005)

[301] Crowley (2011)

[302] Hoogwijk (2004)

[303] Kost and Schlegl (2010)

[304] Held (2010)

[305] Czisch (2005)

[306] Hoogwijk (2004)

[307] Held (2010)

[308] Kost and Schlegl (2010)

[309] Solaranlagen Portal (2011)

[310] Hoogwijk (2004)

[311] Solaranlagen Portal (2011)

[312] Held (2010)

[313] Held (2010)

[314] Neji et al. (2007)

[315] Neji et al. (2007)

[316] Nemet (2007)

[317] Appleyard (2011)

[318] IEA (2010)

[319] IEA (2009) (a)

[320] Hau (2008), p. 622

[321] Kost and Schlegl (2010)

[322] Kost and Schlegl (2010)

[323] Trieb et al. (2009)

[324] U.S. Department of Energy (2006)

[325] Held (2010)

[326] Elsobki et al. (2009)

[327] Neji et al. (2007)

[328] Kost and Schlegl (2010)

[329] Müller et. al (2011)

[330] Luque and Hegedus (2011), p. 1008

[331] Taiyo Kogyo Corporation (2012)

[332] Hilgers (2010)

[333] Hilgers (2010)

[334] Elsobki et al. (2009)

[335] Brand and Zingerle (2010)

[336] NIMA (2003)

[337] Arab Union of Electricity (2010)

[338] Arab Union of Electricity (2001)

[339] STEG (2008)

[340] Moussi (2008)

[341] Madrigal and Stoft (2011), p. 13

[342] Enzili (2011)

[343] Exchange-rates.com; Half year Average Moroccan DH to Euro: 11.2318

[344] Elsobki et al. (2009)

[345] ECB; 5 year average USD to EUR: US$/€ 1.3906

[346] Elsobki et al. (2009)

[347] ECB; 5 year average USD to EUR: US$/€ 1.3906

[348] Madrigal and Stoft (2011), p. 13

[349] ECB; 5 year average USD to EUR: US$/€ 1.3906

[350] Madrigal and Stoft (2011), p. 100

[351] The World Bank (2010)

[352] Egyptian Electricity Transmission Company (2010)

[353] ECB; 5 year average USD to EUR: US$/€ 1.3906

[354] Elsobki et al. (2009)

[355] NIMA (2003)

[356] PV Power Plants (2011)

[357] Farrell (2011)

[358] Dodd (2010)

[359] Elsobki et al. (2009)

[360] Elsobki et al. (2009)

[361] Madrigal and Stoft (2011), p. 100

[362] IEA (2011) (c); p. 34f.

[363] IEA (2008)

[364] Tzimas et al. (2009)

[365] Scholz (2010)

[366] Scholz (2010)

[367] Trieb et al. (2009)

[368] Hilgers (2010)

[369] Hilgers (2010)

[370] Quaschning (2001)

[371] Trieb et al. (2007)

[372] Crowley (2011)

[373] Macknick et al. (2011)

[374] Schermeyer (2011)

[375] Schermeyer (2011)

[376] Trieb (2005)

[377] Hilgers (2010)

[378] Hilgers (2010)

[379] Kost and Schlegl (2010)

[380] Crowley (2011)

[381] Kost and Schlegl (2010)

[382] Elsobki et al. (2009)

[383] Mills and Wiser (2010)

[384] Mills and Wiser (2010); (*): 5 year average USD to EUR: US$/€ 1.3906 (Source: ESB)

[385] Mills and Wiser (2010)

[386] GWEC (2008)

[387] Sensfuss et al. (2003)

[388] GWEC (2008)

[389] GWEC (2008)

[390] Sensfuss et al. (2003)

[391] GWEC (2008)

[392] SWERA (2005)

[393] Czisch (2005)

[394] SWERA (2005)

[395] Czisch (2005)

[396] SWERA (2005)

[397] Crowley (2011)

[398] GWEC (2008)

[399] Scholz (2010)

[400] Schermeyer (2011)

[401] Schermeyer (2011)

[402] IEA (2011) (a)

[403] Wietschel and Genoese (2011)

[404] Wietschel and Genoese (2011)

[405] Wietschel et al. (2010) (b); p. 560 ff.

[406] Hau (2008) p. 635

[407] GWEC (2008)

[408] GWEC (2008)

[409] Medgrid

[410] Desertec Foundation

[411] Dii

[412] Desertec Foundation

[413] Schermeyer (2011)

[414] Benkhadra (2009)

[415] Brand and Zingerle (2010)

[416] Brand and Zingerle (2010)

[417] RCREEE (2010) (a)

[418] NREA (2010)

[419] Tzimas et al. (2009)

[420] Schermeyer (2011)

[421] Czisch (2005)

[422] Hoogwijk (2004)

[423] Hau (2008), p. 622

[424] Elgamal et al. (2010)

[425] UCSDTV (2010)

[426] Smithsonian National Museum of Natural History (2011)

[427] Llorente et al. (2010)

[428] IEA (2009) (b)

[429] Hau (2008)

[430] Czisch (2005)

[431] Held (2010)

[432] Hoogwijk (2004)

[433] Bofinger et al. (2011)

[434] Trieb et al. (2009)

[435] Morthorst et al. (2009)

[436] Scholz (2010)

[437] Luque and Hegedus (2011), p. 491

[438] Adiboina (2010)

[439] BusinessGreen.com (2010)

[440] Crowley (2011)

[441] Trieb et al. (2007)

[442] EWEA (2011)

[443] Neji et al. (2007)

[444] U.S. Department of Energy (2006)

[445] Trieb et al. (2009)

[446] Tzimas et al. (2009)

[447] Neji et al. (2007)

[448] Trieb et al. (2009)

[449] Trieb et al. (2009)

[450] Hilgers (2010)

[451] Trieb et al. (2009)

[452] Tzimas et al. (2009)

[453] Hilgers (2010)

[454] IEA (2008)

[455] Kost and Schlegl (2010)

[456] Hau (2008), p. 622

[457] Luque and Hegedus (2011)

[458] Macknick et al. (2011)

[459] Schermeyer (2011)

[460] GWEC (2008)

[461] Mills and Wiser (2010)

[462] GWEC (2008)

[463] GWEC (2008)

[464] Wietschel et al. (2010) (b)

[465] Schermeyer (2011)

[466] Schermeyer (2011)

[467] GWEC (2008)

[468] Climate Investment Funds

[469] U.S. EPA (2010)

[470] U.S. EPA (2011)

[471] Climate Investment Funds

[472] Clean Investment Funds (2009) (a)

[473] EWEA (2007) (a)

[474] Medgrid

[475] Dii

[476] UNFCCC (2011)

[477] UNFCCC (2011)

[478] Trieb et al. (2009)

[479] Tzimas et al. (2009)

[480] Scholz (2010)

[481] Kost and Schlegl (2010)

[482] Kost and Schlegl (2010)

[483] Wietschel et al. (2010) (a); p. 9ff.

[484] Climate Investment Funds

[485] RCREEE (2010) (b)

[486] Clean Investment Funds (2009) (a)

[487] Nikionok-Ehrlich (2011)

[488] Trieb et al. (2009)

[489] US PRB (2011)

[490] UN Pop (2009)

[491] UN Pop (2009)

[492] US PRB (2011)

[493] SaharaWind.com

[494] SaharaWind.com

[495] SaharaWind.com

[496] O.N.E. (2011)

[497] Dodd (2010)

[498] Amar and Elamouri (2011)

[499] Amar and Elamouri (2011)

[500] Amar and Elamouri (2011)

[501] C.U.B.E. Engineering

[502] Aboulnasr (2006)

[503] Fathi (2008)

[504] Vestas (2011)

[505] Jarvis et al. (2008)

[506] FAO (2007)

[507] NIMA (2003)

[508] WDPA (2010)

[509] BirdLife International

[510] NIMA (2003)

[511] IUCN (2011)

[512] BirdLife International

[513] Wind Power Program (2011)

[514] Wind Power Program (2011)

[515] ECB

[516] Czisch (2005)

[517] Hoogwijk (2004)

[518] Morthorst et al. (2009)

[519] Hau (2008), p. 837

[520] Held (2010)

[521] Bloomberg - New Energy Finance (2009)

[522] Bloomberg - New Energy Finance (2009)

[523] U.S. Department of Energy (2011)

[524] Schermeyer (2011)

[525] Madrigal and Stoft (2011)

[526] Madrigal and Stoft (2011)

[527] KfW Entwicklungsbank (2011)

[528] Suntech

[529] Sharp

[530] Q-Cells

[531] Yingli Solar

[532] Semiconductor Today (2011)

[533] First Solar

[534] ECB

[535] BSW-Solar (2011) (a)

[536] Held (2010)

[537] Solarbuzz

[538] Hoogwijk (2004)

[539] Czisch (2005)

[540] U.S. Department of Energy (2006)

[541] Watts et al. (2011)

[542] IEA (2010)

[543] Held (2010)

Ende der Leseprobe aus 222 Seiten

Details

Titel
Calculating Cost-Supply Curves of Wind Power and Photovoltaic Energy in North Africa using a Geographic Information System
Hochschule
Karlsruher Institut für Technologie (KIT)  (Fraunhofer-Institut für System- und Innovationsforschung)
Note
1.1
Autor
Jahr
2012
Seiten
222
Katalognummer
V196656
ISBN (eBook)
9783668368293
ISBN (Buch)
9783668368309
Dateigröße
12458 KB
Sprache
Englisch
Anmerkungen
Schlagworte
LCOE, Wind, Wind power, PV, PV power, North Africa, Tunisia, Egypt, Morocco, Lybia, Western Sahara, Algeria, Renewable Energy
Arbeit zitieren
Felix Gless (Autor:in), 2012, Calculating Cost-Supply Curves of Wind Power and Photovoltaic Energy in North Africa using a Geographic Information System, München, GRIN Verlag, https://www.grin.com/document/196656

Kommentare

  • Noch keine Kommentare.
Blick ins Buch
Titel: Calculating Cost-Supply Curves of Wind Power and Photovoltaic Energy in North Africa using a Geographic Information System



Ihre Arbeit hochladen

Ihre Hausarbeit / Abschlussarbeit:

- Publikation als eBook und Buch
- Hohes Honorar auf die Verkäufe
- Für Sie komplett kostenlos – mit ISBN
- Es dauert nur 5 Minuten
- Jede Arbeit findet Leser

Kostenlos Autor werden