Copy of en Feasibility Study Wind Park Ashegoda 2006

Feasibility Study for Wind Park Developmentin Ethiopia and Capacity Building

Ashegoda Wind Park Site

Final Report

LI/GE6 25 0447 August 2006

Feasibility Study for Windpark Development in Ethiopia and CapacityBuilding

August 2006, Final Report - page ii

LI / GE6 25 0477 final report ashegoda

Deutsche Gesellschaft für Technische

Zusammenarbeit (GTZ) GmbH

Div. for Environment and Infrastructure

TERNA Wind Energy Programme

Dag-Hammarskjöld-Weg 1 5

D-65760 Eschborn

Telefon +49 (0) 61 96 79-0

Telefax +49 (0) 61 96 79-11 15

Contact partner: Tim-Patrick Meyer

[email protected]

Co-financing:

Cooperation partner: Ethiopian Electric Power Corporation

P.O. Box 1233

Addis Ababa

Ethiopia

Telefon +251-11-5534949

Telefax +251-11-1574071

Contact partner: Kebede Walelu

Wind Power Study Project Coordinator

[email protected]

Consultant:

Lahmeyer International GmbH

Friedberger Strasse 173

61118 Bad Vilbel

Telefon +49 (0) 6101-55-1275

Telefax +49 (0) 6101-55-1826

Contact person: Martin Nietzer

[email protected]


August 2006, Final Report - page iii


Project team

International experts

Team leader Dr. Kris Drabik

Deputy team leader Martin Nietzer

Electrical grid expert Ernesto Martínez-Telo

Wind turbine expert Rüdiger Kipke

WindPro / Micrositing expert Michael Friedrich

Road Survey expert Jürgen Hoffmann

Diesel expert Samuel Karres

Economist & CDM expert Rosa Tarragó

Institutional expert Werner Meyer / Dr. Romeo Pacudan

Local experts

Environmental expert & local co-ordinator Dejene Woldemariam

Socio economist Melessew Shanko Fitamo

Energy expert Mulugeta Sergawie

Backstopping

KLIMM expert Dr. Oliver Heil

Grid connection expert Enrique Salazar

Environmental expert Harald Kaschube

Capacity Credit expert Frank Umbach

Windfarms O&M expert Dr. Patric Kleineidam

Quality assurance

Richard Lawless

Dr. Patric Kleineidam

Tobias Leschinsky

Approval Bungo Ezawa

Bad Vilbel, August 2006


August 2006, Final Report - page 4


1 Executive Summary

The Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH and their Co-

operation partner Ethiopian Electric Power Corporation (EEPCo) are developing the first

wind park projects in the central and northern parts of Ethiopia. After the wind measure-

ment campaign with measurements at a height of 10 m a.g.l, carried out by the GTZ

TERNA program, four sites with favourable wind conditions were pre-selected, and addi-

tional wind measurements with 40 m height are under progress. The total envisaged in-

stallation capacity of the projects is nominated to be approximately 40-60 MW at each

site. Within this report, a complete technical and economical feasibility study for the site

Ashegoda will be elaborated.

The feasibility study comprises, besides a wind potential analysis, of a conceptual techni-

cal layout of the wind parks at the proposed sites Ashegoda and Mesobo-Harena includ-

ing environmental impact assessment and an economic/financial analysis. The latter in-

cludes a Capacity Credit assessment, an Economic Analysis, a Clean Develop-

ment Mechanism (CDM) assessment, a Financial Analysis and a Framework

Analysis for Wind Energy in Ethiopia.

1.1 Approach

The feasibility study is based on available wind resource data from the wind measure-

ments on site, LI s long term experience in the wind energy sector and in wind park plan-

ning, as well as on information obtained from the site visit during the first missions to

Ethiopia in 2006. After first inquiries by international experienced turbines manufacturers

with proven turbine technology and a pre-assessment of the results of the requested Ex-

pression of Interest, four turbines types were selected for the feasibility study. Several

potential wind park layouts were developed for the project site Ashegoda:

- Scenario 1: 86 x ENERCON E-48 800 kW turbines, rotor diameter 48 m, hub height 57 m; total capac-

ity of 68.8 MW

- Scenario 2: 86 x VESTAS V52 m 850 kW turbines, rotor diameter 52 m, hub height 60 m; total capacity

of 73.1 MW

- Scenario 3: 86 x GAMESA G58 850 kW turbines, rotor diameter 58 m, hub height 60m; total capacity

of 73.1 MW

- Scenario 4: 86 x ENERCON E-53 800 kW turbines, rotor diameter 53 m, hub height 57 m total capacity

of 68.8 MW as a potential option for 2007, offered by Enercon-India. (The turbine type is currently in the

planning stage, a first prototype will be erected in 2006).




The cost estimates were derived by LI with an internal cost databank and the relevant

wind power project experience world wide.

1.2 Site Description Ashegoda

The proposed wind park site is situated in the northern Ethiopian highland at an altitude of

2400 m a.g.l close to the descent to the coastal plain. The whole area, foreseen for the

construction of the wind park, is covered with small bushes and grass. The land is mainly

used for extensive goat farming, and partly for agricultural use.

The wind park consists of two main areas; the western area is located on two low ridges in

an approximately north-south orientation while the eastern area is spread over a more

distinct mountain range in north-south orientation with several branches. In between the

two areas a lower plain area can be found. South-east of the eastern part of Ashegoda

wind farm, the upper branches of a valley descending to the coastal plain are reaching the

highland plain.

The orographical terrain conditions can be classified as medium complex. The slopes of

the ridges where the wind park has been proposed and the valley towards the coast are

adding some complexity to the vicinity of the wind park site while the remaining area is flat

or modestly hilly.

1.3 Wind Resource Assessment

A wind potential study has been performed in the course of the feasibility study, taking into

account data from two measurement stations on-site. The first one of 10 m height, erected

in January 2005 and the second one with anemometers at heights of 10 m and 40 m a.g.l,

erected in mid-September 2005. Both masts had been equipped with THIES first class

cup anemometers and a wind direction vane (at 10 m a.g.l.).

To predict the long-term wind speed on site, long-term correlations using the MCP (Meas-

ure-Correlate-Predict) method have been performed, using NCEP (US-National Centers

for Environmental Prediction) reanalysis wind data of a period of 25 years. The findings of

this correlation were adapted to the measured wind speeds in order to make them long-

term representative. The long term average wind speed for Ashegoda site at a height of

40 m a.g.l. is 8.11 m/s.




1.4 Energy Production Estimation

The calculations of the expected energy yield for the proposed wind farm at Ashegoda,

which is based on the long-term corrected measured wind data, are executed with the

international standard wind park planning software combination WindPro/WAsP.

The resulting energy yield for the different scenearios, displayed as Energy P 75 value

which is a common value used by international banks financing wind parks, is displayed in

Table 1-1.

Table 1-1: Summary of main estimation result for the different scenarios

Scenario Scenario 1 Scenario 2 Scenario 3 Scenario 4

Type of Turbine ENERCONE 48 VESTAS V 52 GAMESA G58 ENERCON

E 53

Turbine Capacity [kW] 800 850 850 800

Number of WTG [-] 86 86 86 86

Installed park capacity [kW] 68,800 73,100 73,100 68,800

Hub Height [m] 57 60 60 57

Rotor Diameter [m] 48 52 58 53

Specific Rotor Area [m2/kW] 2.26 2.50 3.11 2.76

Gross energy production P-75[MWh/y]

227,155 226,816 274,505 261,271

Wind park array losses [%] 5.8 5.0 5.3 5.7

Turbine availability [%] 95.0 95.0 95.0 95.0

Electrical losses [%] 2.0 2.0 2.0 2.0

Miscellaneous losses [%] 0.1 0.1 0.1 0.1

Net Output [MWh/y] 197,392 198,771 239,804 227,278

Specific Energy Production[kWh/y/m2]

1,268 1,088 1,055 1,198

Full load hours [h/a] 2,869 2,719 3,280 3,303

Capacity Factor [%] 32.8 31.0 37.4 37.7

Due to the favourable wind speeds on site, and the orientation of the wind park rows al-most perpendicular to the main wind direction leading to array losses which are low for a wind park of the proposed size - a very good wind park performance, resulting in excellentfull load hours, and capacity factors is achieved.




1.5 Site Access

The turbines and the equipment will be shipped to the port of Djibouti. From here the tur-

bines can be transported to the project site by road; the main road to Mekelle passes

Ashegoda site at a distance of approximately 10 kilometres. The road from Djibouti via

Mile, Chifra to Weldiya is a gravel road and the condition is predominantly good. For the

road between Weldiya and Ashegoda no condition report was delivered. Nevertheless, it

will be necessary for the turbine manufacturer or EPC contractor to further investigate the

road condition of the section between Weldiya and Ashegoda in the next phase of the

project development.

At Ashegoda project site, a wide unpaved access road leads to the proposed western and

eastern wind park site. No obstacles like trees or signposts exist. All of the internal access

roads of the wind park have to be constructed newly. The existing access dirt road has to

be reinforced.

Additionally a road survey study covering the link Ashegoda - (Mekelle -) Djibouti was car-

ried out by the consultant showing the possibility of transportation for wind turbines of the

800kW class in general. There are only two bottlenecks along the route which should be

further investigated. In the mountainous terrain, between Alamata and Korem there are

two sections where the gradient exceeds the limit for a transport with the weight of a 40 m

long, Megawatt-class blade load. Furthermore, on these points the radii of the curves also

do not allow the transport of these blades. These bottlenecks currently do not allow for the

transportation of large wind turbines in the Megawatt-range.

Besides this, it was stated in the report that the whole transport from Djibouti to the wind

park site at the plateau will take approximately 11 days for a return trip per major compo-

nent.

The transportation of the components is a major obstacle in the project, as the hauIage for

the entire wind turbines will take about one and a half years when using 10 seperate

trucks.

Moreover, the transport of the equipment is a major issue since for the turbine transport

the national road has to be temporarily blocked.




1.6 Environmental Impact Assessment

The most significant impact to local inhabitants will be the loss of approximately 20 hec-

tares of farmland which is needed for construction of wind turbine towers, connecting

roads and buildings and which has to be compensated to the affected people.

The investigation of noise emission of the wind turbines shows no exceedance of the re-

quired threshold values for the surrounding settlements.

According to the Environmental Study carried out by the Feasibility Study team of EEPCo,

the project site is located outside of any state protection area nor is a serious impact on

bird migration routes, flora and fauna to be expected. In this respect it has to be stressed

that the wind park is located on 2300 to 2450m a.s.l. and consequently in a treeless and

hostile environment to flora and fauna.

1.7 Capacity Credit Assessment

Within the Capacity Credit (CC) assessment it is relevant to evaluate the effects of the

implementation of wind energy in large scale on the system, the results have been calcu-

lated for four scenarios defined as:

Scenario I 48 MW installed at the Ashegoda site

Scenario II 68.8 MW installed at the Ashegoda site

Scenario III 116.8 MW installation of both wind parks

Scenario IV 360 MW installation of wind energy in Ethiopia

The main outcomes of the CC, Economic Analysis, CDM Assessment, Financial Analysis

and Framework Analysis are summarized in the following subsections.

The Capacity Credit Assessment has provided comparatively very good results. In the

Capacity Credit (CC) Assessment the power system capacity (or firm capacity) that can

be replaced by wind power was calculated. Also, the influence of wind power on the Inter-

connected System was measured.

Two approaches were used to calculate the Capacity Credit: Year-Round Capacity Credit

Approach and Peak Load Capacity Credit Approach. The differentiation between year

round and peak load capacity credit is especially valid for Ethiopia and other systems with

large hydro storage capacities, as it is less important for the year round CC when the wind

blows.




The Year-Round Capacity Credit Approach has produced the following results:

CC of 11.12 MW in respect to hydropower capacity, i.e., a 23 % for the Ashegoda

Wind Park, a

CC of 26.46 MW in respect to hydropower capacity, i.e., a 38 % for the Ashegoda site;

a

CC of 36.51 MW in respect to hydropower capacity, i.e., 31.69 % for both sites; a

CC of 83.90 MW in respect to hydropower capacity, i.e., 23.31 % when considering

two large sites of 180 MW each located at the Mesobo Harena and Ashegoda sites.

The Peak Load Capacity Capacity Credit Approach has produced the CC results for the

30 %, 20 %, 10 %, 5 % and 1 % highest load hours. For reference, the CC for the 30 %

highest load hours is:

CC of 30.0 % for the Mesobo - Harena Wind Park, a

CC of 47.2 % for the Ashegoda Wind Park,

CC of 40.3 % for both sites;

CC of 36.7 % when considering two large sites of 180 MW each located at the

Mesobo - Harena and Ashegoda sites.

Further, the influence of wind power on the Interconnected System was measured through

changes on the Load Duration Curve (LDC) of ICS, Load Following and Spatial Smooth-

ing Effect. The LDC reflects how the seasonal and daily wind distribution matches to the

load pattern of the ICS. This is especially important to get an idea of the amount and ca-

pacity category (base, intermediate or peak capacity) that have to be installed within a

system with a demand represented by a certain LDC. When comparing the LDC and a

LDC reduced by the wind energy distributed to the system, an illustration of the wind influ-

ences on the system is obtained, showing when peak capacity, base load capacity is

needed or when wind has to be dumped. Wind energy is highly required at peak load

hours being the CC 68 MW.

Load following indicates whether an additional effort has to be considered regarding

power balancing when introducing wind energy to the Ethiopian system. The calculations

have shown that no further effort has to be realized.

Finally, the Spatial Smoothing Effect, which is the effect that numerous wind parks at dif-

ferent sites reduce the fluctuation of wind energy distributed to the grid, has been dealt.

Fluctuation of wind power output is reduced the more dispersed the installed capacity.

Because of this effect, the installation of more than one wind park is recommended.




1.8 Economic Analysis

The results produced in the economic analysis show that the wind park in all four scenar-

ios is highly economically feasible and these are:

A 30.23 % Economic Internal Rate of Return (EIRR) for the Scenario I (Enercon

E48) has been produced. For this Scenario the Net Present Value (ENPV) is

91.59 million USD and the Benefit/Cost (B/C) ratio is 1.81 calculated at 10 % dis-

count rate.

A 33.44 % EIRR is expected to be produced with the Enercon turbines of larger

size type E53 (Scenario II). The Scenario II ENPV is estimated at 115.86 million

USD and the Benefit/Cost (B/C) ratio at 1.98.

A 27.76 % EIRR is produced with Scenario III (Vestas V52). The ENPV is

87.42 million USD, whereas the Benefit/Cost (B/C) ratio is 1.73 calculated at 10 %

discount rate.

The highest EIRR (35.35 %) is produced with the Scenario IV (Gamesa G58). The

ENPV is logically the highest, i.e., 124.78 million USD and the Benefit/Cost (B/C)

is 2.04 calculated at a 10 % discount rate.

The avoidance of CO2 emissions when installing the wind park instead of a diesel power

plant is very high, estimated at 164,437 CO2 tonnes per year totalling 3,288,741 CO2 ton-

nes in the Scenario with the highest power generation (239,804,000 GWh/year).

The scenario analysis carried out for Scenario IV shows that the variable with the highest

impact on the EIRR is the investment cost. When decreasing investment cost by 10 % the

EIRR increases from 35.35 % to 41.46 %.

1.9 Clean Development Mechanism Assessment

The assessment has shown that the Ashegoda Wind Park project is suitable to be regis-

tered as a CDM activity. By doing so, and considering the Scenario IV (Gamesa G58) with

the highest wind power generation (239,804,000 GWh/year), it results in an annual emis-

sion reduction potential of 4,058.42 tCO2 per year the first crediting period and totalling

73,700.84 tCO2. This has been calculated considering a 3x7 crediting period and a 10 %

reduction on the EF at the end of each crediting period. In terms of revenues, and with

prices for Certified Emission Reductions (CERs) for CDM-projects as high as 6 USD/CER,




the potential additional cash-flow to be generated through the application of CDM during

the 20 years operational period is 442,205.1 USD.

Due to the relative low Emission Factor of the Ethiopian grid system, this amount is rela-

tively modest.

1.10 Financial Analysis

The financial analysis differs from the economic analysis in that, in the financial analysis,

the wind park is viewed as an enterprise and in the economic analysis the wind park is

evaluated from the point of view of the national economy of the country. As in the eco-

nomic analysis, four different Scenarios have been evaluated. The main assumptions

considered in the financial analysis are:

electricity price of 6 USDc/kWh increases by 2 % annually;

Certified Emission Reduction (CERs) credits of CDM at a price of 6 USD/CER;

the project has also been considered to be free of taxes;

no inflation has been considered on O&M costs, whereas a major overhaul be-

tween the 10th and 11th years of operation has been considered as well as wind

farm decommissioning costs in year 21.

Under these assumptions and especially due to the strong favourable energy production

estimates in the location of the wind park, the Ashegoda Windfarm is financially feasible in

two Scenarios. A project has been considered financially feasible if the NPV has been

higher than zero, the IRR is higher than the discount rate (10 %) and the minimum DSCR

stays above the desired mark of 1.20x. Not surprisingly, in projects considered feasible,

the specific generation costs are under the current power tariff levels of 6 USDc/kWh. Fur-

ther, the financial addition of CDM has been measured in terms of impact on IRR.

The two scenarios considered financially feasible are the Scenario IV (Gamesa G58)

and Scenario II (Enercon E53).

In the financial analysis sensitivity testing, energy generation and sales tarif had the major

impact on project results carried on Scenario IV (Gamesa G58). A further parameter with

a high impact on project IRR is the investment cost.

Finally, the specific generation costs of the Ashegoda Wind Park are in all Scenarios

lower than for the reference Diesel Power Plant and higher than current Hydropower

Plants in operation or under construction in Ethiopia.




The levelized costs are:

Scenario I (Enercon E48) 6.44 USDc/kWh

Scenario II (Enercon E53) 5.82 USDc/kWh

Scenario III (Vestas V52) 6.84 USDc/kWh

Scenario IV (Gamesa G58) 5.60 USDc/kWh

whereas for the alternative supply options the levelized costs are:

Dire Dawa DPP 12.90 USDc/kWh

Halele-Werabese HPP 3.00 USDc/kWh

Finchaa HPP 3.40 USDc/kWh




1.11 Conclusions and Recommendations

Due to the promising wind conditions at the project site at Ashegoda and the available

open space at the large proposed area, a realisation of the project in general is feasible.

Considering the need of Ethiopia to diversify its power generation currently highly depend-

ing on hydropower, the prevailing energy crisis due to decreasing rainfalls and the in-

creasing power demand, a short term supply solution has to be implemented.

In the short-run it is necessary to increase the current power generation mix, to cover in-

creasing and unmet power demand and to avoid dependence on fuel imports. Wind and

diesel power generation are the two fast-track implementation alternatives considered by

EEPCo.

Between both, the results show that the implementation of the wind park project is the

most economic and financial feasible power generation alternative to be implemented in

the short run in Ethiopia.

The calculated net energy output at the P75 level for the wind park Ashegoda is in range

of 197,392 MWh/y to 239,804 MWh/y at hub heights between 57 m and 60 m. The capac-

ity factors of 31.0% to 37.7 % surpass average values in comparison with other interna-

tional projects, even when taking into account the reduced performance of the wind tur-

bines due to the low air density (at the altitude of 2,000 m), because of the high average

wind speed on site. Depending on the wind turbine type considered, the Internal Rate of

Return oscillates between 16.64 % and 11.00 %. Additional benefits can be generated

through the avoidance of CO2 emissions, estimated at 1,359,387 CO2 tones for the diesel

power plant.

The Ashegoda Wind Park implementation is also recommendable from the point of view of

the Capacity Credit. For Ashegoda a comparatively high Capacity Credit of 38 %

(26.46 MW of hydropower capacity) can be generated.

Even higher are the Capacity Credit results when combining the Ashegoda Wind Park

with a large scale implementation of wind energy in Ethiopia.

The realisation of the project could have other secondary positive effects for Ethiopia.

Since the Ashegoda Wind Park fulfils the conditions to be registered as a CDM-activity,

potential benefits of a CDM registration could be generated. Due to the relative low Emis-

sion Factor of the Ethiopian grid system, the amount to be generated is relatively modest.

Thus, the registration as a CDM project is financially hardly feasible.

Concerning the technical part of Ashegoda wind park, the selected turbine types of the

manufacurers VESTAS, GAMESA and ENERCON are suitable for the planned project




and can be considered as well engineered and proven technologies. For delivery, installa-

tion and commissioning of the turbines, a first Expression of Interest from the manufac-

turer ENERCON-India is available, which shows the interest in general for wind power

projects in Ethiopia by a foreign turbine manufacturer.

As stated in this report, some barriers were identified concerning the project implementa-

tion:

- Transportation of the large number of turbines will take several months. We recom-

mend the development of a detailed transportation concept in advance.

- Several manufacturers have been requested for an EoI by the consultant, except of

Enercon India most answers are still pending or under clarification of details.

- One of the main risks lies in the time frame for project construction (twelve months) in

2007. An extension of the construction phase seems difficult due to the extreme time

pressure from the Ethiopian Authorities. However, the timely realisation of the con-

struction works is possible in the event that a turbine supply contract is signed as soon

as possible, and supervision of the construction works is applied.

- The grid connection to the 230 kV level is feasible.



final report ashegoda_v05.09

2 Background

This section describes the situation in Ethiopia, as far as it is relevant for the set up of the

further analysis of wind energy in the Ethiopian power system, and the implications for the

future of the system.

Ethiopia is situated at the Horn of Africa between the 4th and the 15th degree of latitude. It

is landlocked with a total area of 1.13 million km2 (comparison Germany 357.050 km2),

with borders to Djibouti, Eritrea, Kenya, Somalia and Sudan. With a population of

74.8 million the population density reaches 66,7 persons per km2. The Gross Domestic

Product (GDP) per capita was USD 800 in 2005, with 50 % of the population below the

poverty line.

2.1 Power System

Roughly 95 % of Ethiopia s electric energy system is dependent on hydropower. The total

net has an installed dependable capacity of 715 MW in the Interconnected System (ICS)

and 31 MW in the Self Contained System (SCS)1. The ICS is the state wide electricity

network supplying 98 % of Ethiopia s electrical energy. The SCS are small isolated grids,

not connected to the ICS. The ICS is owned and operated by the state owned Ethiopian

Electric Power Corporation. In 2003/04 it had an annual production of 2,317GWh together

with the SCS, which is an increase of 12 % to the previous year. The EEPCo has 717,007

customers in both the ICS and the SCS, to whom 1,847 GWh of energy was sold that

year. The difference between the sold and the produced value indicates a loss of 21 %.

EEPCo states that close to 15 % of the population have access to electricity and it is pro-

jected to increase the electrification rate rapidly. Until today the Ethiopian grid is a se-

cluded grid with no connections to neighbouring countries, even though there are plans,

supported by the World Bank and the African Development Bank, to install a transmission

line to Djibouti and Sudan. Both projects await the final launch2.

Ethiopia, having no significant fossil fuel sources, is fully reliant on fuel imports. For the

whole country fuel for cars and generators is hauled from Djibouti via rail and truck, and

only recently the import of oil from Sudan has been taken up. Ethiopia has proven oil re-

serves of 214,000 bbl (January 2002) and 25 billion m3 in natural gas reserves, but the

1 Ethiopian Electric Power Corporation (EEPCo). EEPCo in Brief. Internet source:http://www.eepco.gov.et/brief.html, last access: January 5th, 2006, 14:10 h.2 African Energy (2006), World Bank Commits to Improve Access in Rural Ethiopia, Support EEPCo sStrengthening. African Energy. Issue 95, p. 8-9.




deposits will not be explored any time soon, even though there are plans to produce plas-

tic products from natural gas and energetic use of the gas is under discussion.

2.2 Hydro Power Capacity

There is 679 MW of installed hydro power capacity in the ICS. According to EEPCo this

capacity cannot be counted on as dependable capacity. For the total system the depend-

able capacity is calculated by EEPCo as 601 MW. Currently eight hydro power plants are

distributing to the grid with rated capacities between 11 and 184 MW of installed capacity.

They have been commissioned between 1960 and 2004. Most of the reservoirs have

been designed as a seasonal storage with a theoretical storage capacity of up to 1,000

GWh (Finchaa HPP) energy equivalents. Due to precipitation and siltation of the reser-

voirs, some of the hydro power plants (HPP) are loosing storage volume resulting in re-

duced energy output throughout the year3.

Another restriction of the hydro system is caused by the variability of rainfall. In years of

low rainfall and drought the amount of water available during the rainy season from July

until September does not allow for the reservoirs be filled up to the maximum. Annual wa-

ter inflow to the Finchaa reservoir from 1963 1997 varies between 293 and 532 Mm3

with a standard deviation of 57 and a mean of 427. These extreme changes in water

availability indicate the problems of the Ethiopian electricity supply. In the past extended

drought has been the reason for extensive load shedding, cutting regions or businesses

off the grid to reduce consumption. Customers were left without electricity for 15 hours a

day for two days a week. This unacceptable level of supply security and the aim of in-

creasing the electrification ratio has forced the planning of new hydro power plants. Cur-

rently the Tekeze HPP with a projected installed capacity of 153 MW is under construc-

tion. Various other HPP with an installed total capacity of around 1,500 MW are projected

to be built up until 20134. How much of this very ambitious plan is going to be commis-

sioned until that date is not yet predictable.

3 Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) (Ed.) (2004): Wind Energy ProgrammeTERNA Information for Project Appraisal: Ethiopia. Internet-source. http://www.gtz.de/de/dokumente/en-windenergy-ethiopia-siteselectionreport-2005.pdf, last access: June 2nd, 2006, 15:55 h.4 Ethiopian Electric Power Corporation (EEPCo) (2004): EEPCo Power System Expansion Master Plan Up-date. Personal communication, April 2004.




2.3 Diesel Power Plants

In 2004 three diesel power plants (DPP) were commissioned to ease energy shortage

caused by drought and to reduce load shedding. The newly installed power plants are

Kaliti (9 MW), Awash 7 Kilo (22.4 MW) and Dire Dawa (38 MW). The fuel used to run the

DPPs is mainly heavy fuel oil (HFO), similar to #6-oil, which has to be imported as de-

scribed earlier. The fuel is subsidised by the government so that market prices have to be

estimated. For the year 2004/2005 (G.C.) the three DPP produced 18.4 GWh, which

equals an average capacity factor of 3 %5 (EEPCo 2005b). These numbers and the fact

that the diesel fuel has to be imported at a high price, indicate already that the production

of diesel power is quite costly for EEPCo.

2.4 Implications

The energy sector in Ethiopia is expanding rapidly, and even with the new hydro power

plants, the problem of the fluctuating water availability will not be solved entirely. Thus, in

order to guarantee security of supply, the power generation system has to be diversified.

The necessary increase of the electrification rate and the corresponding grid expansion,

need additional capacity in the short-run to support the hydro system throughout the year

and especially at the end of the dry season, when water levels are low and demand re-

mains constant. Therefore, a fast-track implementation capacity increase is necessary. As

a short-run solution to cover the increasing and suppressed demand, EEPCo evaluates

two alternatives: wind and diesel power. Regarding wind, the primary question for Ethiopia

is not how much capacity can be replaced, as it is a common issue in European or North

American countries when assessing capacity credit of wind, but the question is how much

capacity shall be added by specific amounts of installed wind capacity.

5 Ethiopian Electric Power Corporation (EEPCo) (Ed) (2005b): Power Plant Production Report 1997. Unpub-lished paper.




Table of Contents

1 Executive Summary ..................................................................................... 4

1.1 Approach .................................................................................................................................. 4

1.2 Site Description Ashegoda....................................................................................................... 5

1.3 Wind Resource Assessment .................................................................................................... 5

1.4 Energy Production Estimation .................................................................................................. 6

1.5 Site Access............................................................................................................................... 7

1.6 Environmental Impact Assessment.......................................................................................... 8

1.7 Capacity Credit Assessment .................................................................................................... 8

1.8 Economic Analysis ................................................................................................................. 10

1.9 Clean Development Mechanism Assessment ....................................................................... 10

1.10 Financial Analysis................................................................................................................... 11

1.11 Conclusions and Recommendations...................................................................................... 13

2 Background ................................................................................................ 15

2.1 Power System ........................................................................................................................ 15

2.2 Hydro Power Capacity............................................................................................................ 16

2.3 Diesel Power Plants ............................................................................................................... 17

2.4 Implications............................................................................................................................. 17

3 Introduction ................................................................................................ 34

4 Site conditions ........................................................................................... 36

4.1 Site Description ...................................................................................................................... 36

4.2 Site Limitations ....................................................................................................................... 38

4.3 Ground and Soil Conditions ................................................................................................... 38

4.4 Access Roads, Availability of Cranes..................................................................................... 40

4.4.1 Road Access .......................................................................................................................... 40

4.4.2 Road Access for larger Turbines up to 2 MW ........................................................................ 42

4.4.3 Railway Transport .................................................................................................................. 43

4.4.4 Available Crane Capacities .................................................................................................... 46

4.5 Internal Access Roads ........................................................................................................... 48

4.6 Enviromental Impact Assessment.......................................................................................... 48

4.7 Legal Constraints ................................................................................................................... 49




4.8 Earthquake Risk ..................................................................................................................... 49

5 Wind Resources ......................................................................................... 51

5.1 Wind Data Collection.............................................................................................................. 51

5.2 Wind Data Analysis ................................................................................................................ 57

5.3 Long-term Correlation ............................................................................................................ 60

5.4 IEC Wind Class ...................................................................................................................... 64

6 Technical Layout of Ashegoda Wind Park............................................... 67

6.1 Wind Potential Map ................................................................................................................ 67

6.2 Wind Turbine Selection .......................................................................................................... 69

6.2.1 Suitable Tower Heights .......................................................................................................... 69

6.2.2 Determination of the Optimal Unit Size .................................................................................. 71

6.3 Turbine Distances .................................................................................................................. 76

6.4 Wind Park Layouts ................................................................................................................. 77

6.4.1 Wind Park Layout Enercon E-48............................................................................................ 78

6.4.2 Wind Park Layout Vestas V52................................................................................................ 79

6.4.3 Wind Park Layout Gamesa G-58 ........................................................................................... 80

6.4.4 Wind Park Layout Enercon E-53............................................................................................ 81

6.4.5 Conclusion.............................................................................................................................. 81

6.5 Turbulence.............................................................................................................................. 82

6.6 Noise Impact........................................................................................................................... 83

6.7 Shadow Impact....................................................................................................................... 85

7 Energy Production Estimation.................................................................. 86

7.1 Meteorology............................................................................................................................ 86

7.2 Software Basics...................................................................................................................... 88

7.2.1 WindPro.................................................................................................................................. 88

7.2.2 WAsP...................................................................................................................................... 88

7.3 Model Input Parameters ......................................................................................................... 89

7.3.1 Orography............................................................................................................................... 89

7.3.2 Roughness ............................................................................................................................. 91

7.3.3 Wind Shear............................................................................................................................. 93

7.4 Wind Turbine Parameters ...................................................................................................... 96

7.4.1 Power Curve and Air Density ................................................................................................. 97

7.5 Losses and Uncertainties ..................................................................................................... 100

7.5.1 Losses .................................................................................................................................. 1007.5.1.1 Park Efficiency ..................................................................................................................... 100




7.5.1.2 Turbine Availability............................................................................................................... 1017.5.1.3 Electrical Losses .................................................................................................................. 1017.5.1.4 Miscellaneous Losses.......................................................................................................... 102

7.5.2 Wind Speed related Uncertainties........................................................................................ 1027.5.2.1 Uncertainties of the WAsP-Model........................................................................................ 1037.5.2.2 Uncertainties of the Wind Data ............................................................................................ 1047.5.2.3 Total Wind related Uncertainty ............................................................................................ 1067.5.2.4 Uncertainties of the Power Curve ........................................................................................ 106

7.5.3 Uncertainties Energy Yield ................................................................................................... 1077.5.3.1 Enercon E-48 ....................................................................................................................... 1087.5.3.2 Vestas V52........................................................................................................................... 1117.5.3.3 Gamesa G 58....................................................................................................................... 1147.5.3.4 Enercon E-53 ....................................................................................................................... 117

7.6 Summary .............................................................................................................................. 120

8 Internal Wind Park Cabling...................................................................... 121

8.1 Cabling Concept................................................................................................................... 121

8.2 Cable Type ........................................................................................................................... 124

8.3 Earthing Network.................................................................................................................. 125

8.4 Determination of Wind Turbine Groups................................................................................ 127

8.5 Switching Station.................................................................................................................. 128

9 Grid Connection ....................................................................................... 129

9.1 Grid Integration Concept ...................................................................................................... 129

9.2 Overhead Line...................................................................................................................... 133

9.3 Transmission Line ................................................................................................................ 134

9.4 Mekele Substation................................................................................................................ 135

10 Estimation of costs .................................................................................. 137

10.1 Investment Costs Estimation................................................................................................ 137

10.1.1 Enercon E 48 Investment Costs........................................................................................... 137

10.1.2 Vestas V52 Investment Costs .............................................................................................. 138

10.1.3 Gamesa G58 Investment Costs ........................................................................................... 139

10.1.4 Estimated Enercon E-53 Investment Costs ......................................................................... 140

10.2 Construction Period.............................................................................................................. 141

10.3 Potential for Local/Regional Input ........................................................................................ 141

10.3.1 Grid Connection by EEPCo.................................................................................................. 141

10.3.2 Civil Works............................................................................................................................ 141

10.3.3 Lattice Towers for Wind Turbines ........................................................................................ 141

10.4 Operation and Maintenance Costs....................................................................................... 142

10.4.1 General Description.............................................................................................................. 142




10.4.1.1 Enercon E-48 Maintenance and Repair Costs ................................................................ 14410.4.1.2 Vestas V52 Maintenance and Repair Costs.................................................................... 14510.4.1.3 Gamesa G58 Maintenance and Repair Costs................................................................. 14610.4.1.4 Enercon E-53 Maintenance and Repair Costs ................................................................ 14710.4.1.5 Comparison of Maintenance and Repair Costs Estimations ........................................... 147

10.4.2 High Technical Availability.................................................................................................... 148

10.4.3 Local and foreign Operation and Maintenance .................................................................... 148

10.4.4 Training................................................................................................................................. 149

10.4.5 Overview Operation and Maintenance Costs ...................................................................... 15010.4.5.1 Scenario I 86 x Enercon E-48 Turbines........................................................................ 15010.4.5.2 Scenario II 86 x Vestas V52 Turbines .......................................................................... 15110.4.5.3 Scenario III 86 Gamesa G58 Turbines ......................................................................... 15210.4.5.4 Scenario IV 86 Enercon E-53 Turbines ........................................................................ 15310.4.5.5 Comparison of different Scenarios .................................................................................. 153

11 Capacity Credit ......................................................................................... 154

11.1 Methodology and assumptions ............................................................................................ 154

11.1.1 Definition of the Capacity Credit........................................................................................... 154

11.1.2 Methodology Overview......................................................................................................... 155

11.1.3 Assumptions ....................................................................................................................... 158

11.1.4 Load Duration Curve Calculations ....................................................................................... 15911.1.4.1 Load Restrictions ............................................................................................................. 16111.1.4.2 Load Duration Curve Analysis & Results......................................................................... 161

11.1.5 Load Following ................................................................................................................... 163

11.1.6 Spatial Smoothing Effect ...................................................................................................... 165

11.2 Year-Round Capacity Credit Approach................................................................................ 166

11.3 Peak Load Capacity Credit................................................................................................... 168

11.3.1 Seasonal Wind and Water Distribution ................................................................................ 168

11.3.2 Daily wind power distribution........................................................................................... 169

11.3.3 Peak Load Capacity Credit Calculation................................................................................ 170

11.4 Results: Capacity Credit....................................................................................................... 173

11.5 Conclusion: Capacity Credit ................................................................................................. 175

12 Economic Analysis .................................................................................. 177

12.1 Methodology & Main Assumptions....................................................................................... 177

12.2 Economic Benefits................................................................................................................ 180

12.2.1 Basic Diesel Power Plant Data ....................................................................................... 180

12.2.2 Avoided Capital Costs ...................................................................................................... 180

12.2.3 Avoided Fuel Costs .............................................................................................................. 183

12.2.4 Avoided non-Fuel O&M Costs.............................................................................................. 189

12.2.5 Avoided Emissions ............................................................................................................ 190




12.2.6 Diesel Summary Assumptions ............................................................................................. 193

12.2.7 Indirect benefits ................................................................................................................. 194

12.3 Economic Costs ................................................................................................................... 194

12.3.1 Investment Costs of the Wind Park...................................................................................... 194

12.3.2 Economic O&M Costs of the Wind Park .............................................................................. 195

12.3.3 Leakage Costs of the Wind Park .................................................................................... 198

12.4 Results: Economic Analysis ................................................................................................. 198

12.4.1 Economic Cash-flow Projections .................................................................................... 198

12.4.2 EIRR and NPV ................................................................................................................... 199

12.4.3 B/C Ratio............................................................................................................................. 199

12.5 Scenario Analysis................................................................................................................. 201

12.5.1 Results: Economic Scenario Analysis.................................................................................. 203

12.6 Conclusions: Economic Analysis ......................................................................................... 204

13 CDM Assessment..................................................................................... 206

13.1 Introduction........................................................................................................................... 206

13.2 Institutional Framework for CDM Projects in Ethiopia.......................................................... 206

13.3 CDM Project Cycle ............................................................................................................... 208

13.4 Emission Reductions attributable to the Ethiopian WPs ...................................................... 209

13.4.1 Step 2. Calculation of the Build Margin ................................................................................ 213

13.4.2 Step 3. Calculation of the Baseline Emission factor ............................................................ 214

13.5 Conclusions: CDM Assessment........................................................................................... 220

14 Financial Analysis.................................................................................... 221

14.1 Methodology & Main Assumptions....................................................................................... 221

14.1.1 Inflation Rate ...................................................................................................................... 222

14.1.2 Rate of Exchange .............................................................................................................. 222

14.1.3 Depreciation Rates............................................................................................................ 223

14.1.4 Dividend Distribution ......................................................................................................... 223

14.1.5 Applicable Taxes ............................................................................................................... 224

14.1.6 Discount Rate (WACC) ........................................................................................................ 22414.1.6.1 Definition .......................................................................................................................... 22414.1.6.2 Methodology .................................................................................................................... 22514.1.6.3 WACC Results ................................................................................................................. 227

14.1.7 Project s Milestones ............................................................................................................. 228

14.1.8 Investment Costs.................................................................................................................. 229

14.1.9 Operation and Maintenance Costs....................................................................................... 230

14.1.10 Land Lease Costs ................................................................................................................ 231




14.1.11 Costs for Mitigation Measures.............................................................................................. 231

14.1.12 Project Financing Structure .................................................................................................. 23114.1.12.1 Equity Finance ................................................................................................................. 23214.1.12.2 Debt Finance.................................................................................................................... 232

14.1.13 CDM Up-Front & Administrative Costs................................................................................. 233

14.1.14 Financial Benefits ................................................................................................................. 23414.1.14.1 Electricity Sales................................................................................................................ 23414.1.14.2 CDM Revenues................................................................................................................ 235

14.2 Results: Financial Analysis................................................................................................... 235

14.2.1 Major Financial Indicators ................................................................................................ 23614.2.1.1 Net Present Value ......................................................................................................... 23614.2.1.2 Financial IRR (FIRR) and Return on Equity (ROE) ................................................. 23714.2.1.3 Return on Equity (ROE) ............................................................................................... 23714.2.1.4 Debt Service Coverage Ratio (DSCR)....................................................................... 23814.2.1.5 Levelized Costs................................................................................................................ 239

14.2.2 Summary of Key Financial Parameters.......................................................................... 242

14.2.3 Conclusions: Financial Analysis ..................................................................................... 242

14.2.4 Financial Statements ........................................................................................................ 24414.2.4.1 Cash-flow projections ................................................................................................... 24414.2.4.2 Profit and Loss Accounts ............................................................................................. 24414.2.4.3 Balance ........................................................................................................................... 244

14.3 Sensitivity Analysis ............................................................................................................... 244

14.3.1 Methodology ....................................................................................................................... 245

14.3.2 Sensitivity Variables.......................................................................................................... 245

14.3.3 Results: Sensitivity Testing.............................................................................................. 24514.3.3.1 Sensitivity Measurement.............................................................................................. 248

15 Framework Analysis for Wind Energy in Ethiopia ................................ 249

15.1 Financing Options ................................................................................................................ 249

15.2 Regulatory and Legal Framework ........................................................................................ 250

16 Conclusion and Recommendations ................................................... 251

17 Annex A..................................................................................................... 253

17.1 Annex A - 1: Aviation corridors at Mekelle airport ................................................................ 253

17.2 Annex A - 2: Enviromental Report of Ashegoda .................................................................. 254

17.3 Annex A - 3: Soil investigation report ................................................................................... 255

17.4 Annex A - 4: Road Authority Report, Weights and Wheel base........................................... 256

17.5 Annex A - 5: Description of available crane by MIDROC (Addis Abeba)............................. 257

17.6 Annex A - 6: Road Map Djibouti - Mekelle ........................................................................... 258

17.7 Annex A - 7: Road Survey Report ........................................................................................ 259

17.8 Annex A - 8: Terms of reference for consultant's work ........................................................ 260




18 Annex B..................................................................................................... 261

18.1 Annex B 1: Correlation Diagrams...................................................................................... 261

18.2 Annex B 2: Noise Impact................................................................................................... 262

18.3 Annex B 3: Shadow Impact ............................................................................................... 263

19 Annex C..................................................................................................... 264

19.1 Annex C 1: Map of Ashegoda Wind Park Layout.............................................................. 264

19.2 Annex C 2: Energy Calculations ....................................................................................... 265

19.3 Annex C 3: Turbulence Calculations................................................................................ 266

19.4 Annex C 4: two-dimensional view of the digital terrain model........................................... 267

19.5 Annex C 5: three-dimensional view of the digital terrain model ....................................... 269

20 Annex D..................................................................................................... 272

20.1 Annex D 1: Technical Information of Enercon E-48.......................................................... 272

20.2 Annex D 2: Technical Information of Vestas V52 ............................................................. 273

20.3 Annex D 3: Technical Information of Gamesa G58 .......................................................... 274

20.4 Annex D 4: Preliminary Technical Information of Enercon E-53....................................... 275

21 ANNEX E : ................................................................................................. 276

21.1 Annex E-1 ENERCON E-48 Cash-Flow Economic Analysis ............................................. 276

21.2 Annex E-2 ENERCON E-53 Cash-Flow Economic Analysis ............................................... 277

21.3 Annex E-3 VESTAS V52 Cash-Flow Economic Analysis .................................................... 278

21.4 Annex E-4 GAMESA G58 Cash-Flow Economic Analysis................................................... 279

22 ANNEX F : ................................................................................................. 280

22.1 Annex F-1 ENERCON E-48 Operating Results ................................................................... 281

22.2 Annex F-2 ENERCON E-48 Profit & Loss Account ............................................................. 282

22.3 Annex F-3 ENERCON E-48 Balance Sheet......................................................................... 283

22.4 Annex F-4 ENERCON E-53 Operating Results ................................................................... 284

22.5 Annex F-5 ENERCON E-53 Profit & Loss Account ............................................................. 285

22.6 Annex F-6 ENERCON E-53 Balance Sheet......................................................................... 286

22.7 Annex F-7 VESTAS V52 Operating Results ........................................................................ 287

22.8 Annex F-8 VESTAS V52 Profit & Loss Account................................................................... 288

22.9 Annex F-9 VESTAS V52 Balance Sheet.............................................................................. 289

22.10 Annex F-10 GAMESA G58 Operating Results..................................................................... 290

22.11 Annex F-11 GAMESA G58 Profit & Loss Account ............................................................... 291

22.12 Annex F-12 GAMESA G58 Balance Sheet .......................................................................... 292




23 ANNEX G :................................................................................................. 293

23.1 Annex G-1 Fuel Factors used in the CDM Assessment ...................................................... 293

List of Figures

Figure 4-1: Map of Ethiopia with the region of Ashegoda wind park .................................36

Figure 4-2: Ashegoda area, digital terrain model, view from the southwest in a height of1000 m above ground. A larger print-out can be found in Annex C-5 ........................37

Figure 4-3: The soil condition on Ashegoda site ...............................................................39

Figure 4-4: Branch from main road to Ashegoda sites ......................................................40

Figure 4-5: Road map of north-eastern Ethiopia (red arrow marks the wind park area) ...42

Figure 4-6: Typical open freight wagon of the Djibouti Ethiopean railway......................45

Figure 5-1: Measurement masts at Ashegoda site ............................................................51

Figure 5-2: Vicinity of measuring mast 13 Ashegoda II (center of figure).........................53

Figure 5-3: Measuring mast 13 Ashegoda II, vertical adjustment......................................53

Figure 5-4: Measuring mast 4 Ashegoda I, anemometer and wind vane at 10 m a.g.l. ....54

Figure 5-5: measurement device installation at mast 4 Ashegoda I ..................................55

Figure 5-6: wind rose of mast 4 Ashegoda I ......................................................................56

Figure 5-7: wind rose of mast 13 Ashegoda II ...................................................................56

Figure 5-8: Correlation diagram for mast 13 Ashegoda II, anemometers at 10m (x-axis)and 40m (y-axis) for wind direction sector 105° to 135°.............................................58

Figure 5-9: NCEP grid points around Mekelle (yellow pins mark the 2.5° -spacing NCEPgrid points, red pins the location of the RISØ delivered NCEP points), the red circleidicates the selected NCEP point ...............................................................................61

Figure 5-10: Wind rose at 40 m a.g.l. for Ashegoda site, mast 1 Ashegoda30m_40m, 10-year corrected. ...........................................................................................................63

Figure 5-11: Summarized Weibull-distribution for mast 1 Ashegoda30m_40m, 10-yearcorrected ....................................................................................................................64

Figure 6-1: Wind potential map of Ashegoda site (red line indicates the road, orange areas mark the villages, yellow line the wind park area) ......................................................68

Figure 6-2: Turbine costs in accordance with several hub heights....................................70

Figure 6-3: Digital height model, view from a height of 1000 m height south-east toAshegoda wind park. A larger print-out can be found in Annex C-5...........................77

Figure 6-4: Ashegoda Wind Park, Layout Enercon E-48..................................................78

Figure 6-5: Ashegoda Wind Park, Layout Vestas V52 ......................................................79

Figure 6-6: Ashegoda Wind Park, Layout Gamesa G-58 ..................................................80




Figure 6-7: Ashegoda Wind Park, Layout Enercon E-53...................................................81

Figure 6-8: Turbulence vs. IEC characteristics (conditions fulfilled)..................................83

Figure 7-1: Wind flow over forest (source: GASCH 1996).................................................86

Figure 7-2: Wind flow energy vs roughness class (source: WindPro manual) ..................87

Figure 7-3: SRTM height contour map of Ashegoda .........................................................90

Figure 7-4: three-dimensional digital elevation model of Ashegoda site, view from south-western direction. .......................................................................................................90

Figure 7-5: Roughness classes (source: WindPro Manual) ..............................................92

Figure 7-6: Roughness map of Ashegoda site (red ellipse: wind park site) ......................93

Figure 7-7: Wind Profile for Ashegoda wind measurement ...............................................95

Figure 7-8: measured power curve with scatter band........................................................97

Figure 7-9: Power curves of the wind turbines under consideration for Ashegoda windpark for an air density of = 1.225 kg/m3..................................................................99

Figure 7-10: Weibull distribution of the correlated wind data of mast 1 Ashegoda 30_40m...................................................................................................................................99

Figure 7-11: Probability of exceedance for Ashegoda wind park, Enercon E-48 layout..109

Figure 7-12: Probability of exceedance for Ashegoda wind park, Vestas V52 layout .....112

Figure 7-13: Probability of exceedance for Ashegoda wind park, Gamesa G58 layout ..115

Figure 7-14: Probability of exceedance for Ashegoda wind park, Enercon E-53 layout..118

Figure 7-15: P75 energy production of the different scenarios of Ashegoda wind park ..120

Figure 8-1: Principle scheme of a ring concept ...............................................................121

Figure 8-2: Principle scheme of a radial concept ............................................................122

Figure 8-3: Scheme of the internal Ashegoda wind park cabling ....................................123

Figure 8-4 Earthing network ............................................................................................125

Figure 8-5 Terminal Tower connected to substation earthmat ........................................127

Figure 9-1 Geographical layout of the Ashegoda wind park grid connection ..................131

Figure 9-2: Principle wind park grid connection layout ....................................................132

Figure 9-3 Wind Park grid connection schemes..............................................................134

Figure 9-4 Mekele substation ..........................................................................................135

Figure 9-5 Single line diagram of Ethiopian power grid section including MEKELEsubstation .................................................................................................................136

Figure 10-1: Development of the O&M cost for Ashegoda..............................................144







Figure 11-1: Load duration curve and residual load duration curve for Ethiopia, The insetshows the 50 highest load hours (own illustration)...................................................160

Figure 11-2: LDCs and power output of other plants, dependable capacity, max. load,max. residual load (own illustration) .........................................................................162

Figure 11-3: Load following .............................................................................................164

Figure 11-4 Water reservoir levels with and without wind energy ...................................167

Figure 11-5: Average load and energy based CC ...........................................................168

Figure 11-6: Seasonal wind and water distribution..........................................................169

Figure 11-7: Daily load and wind distribution match........................................................170

Figure 11-8: CC at different Load Levels.........................................................................172

Figure 12-1: HFO380 price development ........................................................................186

Figure 12-2: Development of the brent crude oil price from 1997 to 2005 ......................187

Figure 12-3: CO2 price development at EEX ...................................................................192

Figure 12-4: Crude price (USD/bbl) from April 2004 to March 2006................................202

Figure 12-5: Scenario analysis: HFO and LFO price development........................203

Figure 14-1 USD / ETB exchange rate development ......................................................223

Figure 14-2: Capital Asset Pricing Model ........................................................................226

Figure 14-3: Meaning of CAPM .......................................................................................226

Figure 14-4: Spider diagram showing results of sensitivity testing..................................246




List of Tables

Table 1-1: Summary of main estimation result for the different scenarios...........................6

Table 4-1 Transport figures of wind turbines .....................................................................44

Table 4-2: Basic dimensions of the required cranes .........................................................47

Table 4-3: Overview of the Access Roads distances ........................................................48

Table 5-1: Basic data of the Ammonit data logger WICOM 32..........................................52

Table 5-2: Basic data of the measuring masts at Ashegoda site ......................................52

Table 5-3: basic wind measured data for Ashegoda site...................................................55

Table 5-4: Correlation coefficients of the measurement masts used for the MCP-Process...................................................................................................................................59

Table 5-5: wind speed data used for MCP-prediction .......................................................59

Table 5-6: Wind direction data used for MCP-prediction...................................................60

Table 5-7: overview of the long term (10-year) wind data at 40 m a.g.l of mast 13Ashegoda II ...............................................................................................................63

Table 5-8: IEC Wind Turbine classification.......................................................................65

Table 5-9: Site classification parameters...........................................................................66

Table 6-1: Specific investment costs for selected turbine types........................................73

Table 6-2: Requests for expressions of interest ...............................................................74

Table 6-3: minimal turbine distances for Ashegoda wind park ..........................................76

Table 6-4: turbulence sub-classes of the selected wind turbines ......................................82

Table 6-5: Noise standards according to German standards ............................................84

Table 6-6: Sunshine probability at Mekelle........................................................................85

Table 7-1: Hellmann-exponents, sector-wise, for the location of met mast1 Ashegoda30m_40m ................................................................................................94

Table 7-2: Sources of the power curves............................................................................97

Table 7-3: Array losses Ashegoda wind park ..................................................................101

Table 7-4: Deviation of Energy due to wind uncertainties ...............................................108

Table 7-5: Energy Calculations for Ashegoda Wind Park, Enercon E-48 layout .............110


Table 7-7: Energy Calculations for Ashegoda Wind Park, Vestas V52 layout.................113


Table 7-9: Energy Calculations for Ashegoda Wind Park, Gamesa G58 layout..............116

Table 7-10: Deviation of Energy due to wind uncertainties .............................................117

Table 7-11: Energy Calculations for Ashegoda Wind Park, Enercon E-53 layout ...........119




Table 8-1 Wind Park division by four areas.....................................................................128

Table 10-1: Total Investment Cost of Wind Farm Ashegoda...........................................137




Table 10-5: Local and foreign O&M tasks .......................................................................149

Table 10-6: Annual Operation & Maintenance Expenses................................................150




Table 11-1: Difference between LDC, ILDC and RLDC ..................................................162

Table 11-2: Frequency of load changes within one hour with and without wind .............165

Table 11-3: Stored energy at Finchaa reservoir ..............................................................167

Table 11-4: Frequency distribution of wind power during hours of highest load .............171

Table 11-5: Frequency of hours with available wind below the Year-Round CC.............172

Table 11-6: Capacity Credit Results................................................................................174

Table 12-1: Technical data of the reference DPP ...........................................................182

Table 12-2: Economic data of the reference DPP ...........................................................183

Table 12-3: Fuel prices at different international ports in USD/metric ton .......................186

Table 12-4: Relevant fuel price data................................................................................188

Table 12-5: Non-fuel variable O&M for a DPP.................................................................190

Table 12-6: Avoided emissions of the DPP .....................................................................191

Table 12-7: Summary of basic assumptions of the reference DPP.................................193

Table 12-8: Investment costs considering shadow prices ...............................................195

Table 12-9: Economic values of Enercon E48 annual O&M costs ..................................196

Table 12-10: Economic values of Enercon E53 annual O&M costs ................................196

Table 12-11: Economic values of Vestas V52 annual O&M costs...................................197

Table 12-12: Economic values of Gamesa G58 annual O&M costs................................197

Table 12-13: Results economic analysis Ashegoda wind park .......................................200

Table 12-14: Summary results of scenario analysis ........................................................204

Table 13-1: Ethiopian power plant structure....................................................................211

Table 13-2 Baseline Scenario A ......................................................................................215

Table 13-3: Baseline Scenario B1 ...................................................................................216

Table 13-4: Baseline Scenario B2 ...................................................................................216

Table 13-5: Emission reductions & CER revenues for P75, P50 and P90 ......................217




Table 13-6: CDM Assumptions........................................................................................218

Table 13-7: CER generation and CER revenues with Enercon E48 turbines..................218

Table 13-8: CER generation and CER revenues with Enercon E53 turbines..................218

Table 13-9: CER generation and CER revenues with Vestas V52 turbines ....................219

Table 13-10: CER generation and CER revenues with Gamesa G58 turbines ...............219

Table 14-1: Average exchange rates ..............................................................................222

Table 14-2: WACC for the Scenario I (Enercon E48) ......................................................227

Table 14-3: Project s milestones .....................................................................................228

Table 14-4: Investment costs (financial value) ................................................................229

Table 14-5: O&M costs....................................................................................................230

Table 14-6: Financing structure and disbursement .........................................................232

Table 14-7: Annual energy generation ............................................................................235

Table 14-8: Levelized costs for wind energy ...................................................................239

Table 14-9: Assumptions for the DPP & HPP levelized cost calculation.........................241

Table 14-10: Wind, diesel, hydropower levelized costs................................................241

Table 14-11: Summary of key financial parameters ........................................................242

Table 14-12: Financial sensitivity testing results .............................................................246

Table 14-13: Sensitivity indicators & switching values ..............................................248




Abbreviations

a annum,year

ACM0002 Approved consolidated baseline methodology 0002

a.g.l above ground level

a.s.l. above sea level

B/C Benefit/Cost (ratio)

bbl Barrel

bbl Barrel

BM Build Margin

CAPM CapitalAsset Pricing Model

CDM Clean Development Mechanism

CDM Clean Development Mechanism

CER Certified Emission Reduction

CER Certified Emission Reduction

CIRR Commercial Interest Reference Rate

CIRR Commercial Interest Reference Rate

CM Combined Margin

cSt Centi Stokes

DNA Designated National Authority

DOE Designated Operational Entity

DPP Diesel Power Plant

DPP Diesel Power Plant

DSCR Dept Service Coverage Rate

DSCR Dept Service Coverage Rate

ECA Export Credit Agency

ECA Export Credit Agency

EEPCo Ethiopian Electric Power Corporation

EEX European Energy Exchanges

EF Emission Factor

EIRR Economic Internal Rate of Return

EIRR Economic Internal Rate of Return

EMD Energi-og Miljödata

ENPV Economic Net Present Value




EPA Ethiopian Environmental Protection Authority

ERPA Emission Reduction Purchase Agreement

ETB Ethiopian Birr

ETB Ethiopian Birr

FIRR Financial Internal Rate of Return

FIRR Financial Internal Rate of Return

FOCC Financial Opportunity Cost of Capital

FX Foreign Exchange

GDP Gross Domestic Product

GDP Gross Domestic Product

GEF Global Environment Facility

GEF Global Environment Facility

GHG Greenhouse Gas

GHG Greenhouse Gas

HFO Heavy Fuel Oil

HFO Heavy Fuel Oil

HPP Hydro Power Plant

HPP Hydro Power Plant

ICS Interconnected System

ICS Interconnected System

IDC Interest During Construction

IDGTE Institution Diesel and Gas Turbine Engineers

IPP Independent Power Producer

IRR Internal Rate of Return

KP Kyoto Protocol

KW Kilowatt

kW-class Windturbines in the range up to 1000 kW

KWh Kilowatthour

LDC Load Duration Curve

LFO Light Fuel Oil

LHV Lower Heating Value

LI Lahmeyer International

MCM Million Cubic Meter

MDO Marine Diesel Oil

MGO Marine Gas Oil




MW-class Wind turbines above 1MW (1000 kW)

NCEP National Centre for Environmental Prediction

NCV Net Calorific Value

NGO Non-Governmental Organization

NPV Net PresentValue

NREA National Renewable Energy Authority

O&M Operation andMaintenance

ODA Overseas Development Assistance

OM Operating Margin

PDD ProjectDesignDocument

PoE Probability of Exceedance

PPP Power Purchase Parity

PSEMP Power System Expansion Master Plan

RLDC Residual LoadDuration Curve

ROE Return on Equity

SCF Standard Conversion Factor

SCS Self Contained System

SI Sensitivity Indicator

SW Switching Values

UNFCCC United Nations Framework on Climate Change

USD US Dollar

W weights

WACC Weighted Average Cost of Capital

WAsP Wind Atlas Analysis and Application Program

WPs Wind Parks




3 Introduction

Roughly 95 % of Ethiopia s electric energy system is dependent on hydropower. The total

net has an installed capacity of 715 MW in the Interconnected System and 31 MW in the

Self Contained System. Due to the precipitation and siltation of the reservoirs some of the

hydro power plants are loosing storage volume resulting in reduced energy output

throughout the year. Another restriction of the hydro system is caused by the variability of

rainfall. In years of low rainfall and drought the amount of water available during the rainy

season from July until September does not allow for the reservoirs be filled up to the

maximum. These extreme changes in water availability indicate the problems of the

Ethiopian electricity supply.

The energy sector in Ethiopia is expanding rapidly, and even with the new hydro power

plants, the problem of the fluctuating water availability will not be solved entirely. Thus, in

order to guarantee security of supply, the power generation system has to be diversified.

The necessary increase of the electrification rate and the corresponding grid expansion,

need additional capacity in the short-run to support the hydro system throughout the year

and especially at the end of the dry season, when water levels are low and demand re-

mains constant. Therefore, a fast-track implementation capacity increase is necessary. As

a short-run solution to cover the increasing and suppressed demand, EEPCo evaluates

two alternatives: wind and diesel power.

Three diesel power plants were already commissioned in 2004 to easy energy shortage

caused by drought and to reduce load shedding. Ethiopia is fully reliant on fuel imports.

Fuel prices have been steady increasing in the last years, which worsens the economic

feasibility of the installed and new diesel power plants.

Since the beginning of 2005, GTZ is supporting EEPCo in the planning of a grid-

connected wind farm in the range of 40-60 MW. Within its TERNA program, GTZ has car-

ryed out wind measurements at different sites in Ethiopia since January 2005.

Ethiopian Electric Power Corporation (EEPCo) is planning the implementation of wind

parks in several areas of the Federal Democratic Republic of Ethiopia, which are esti-

mated to comprise of up to 200 MW to the year 2012. The first two wind parks, with a ca-

pacity of 40 60 MW each, are expected to come to the grid by the end of 2007 at the

latest. One of the foreseen wind parks shall be located in the northern part of Ethiopia in

the higher mountain areas, outside of larger towns, but close to existing transmissions

lines and roads. For the prediction of the wind resources, wind measurements on-site

have been performed by the GTZ TERNA program through the consultant of GTZ since

2005, which indicate above average wind speeds for the pre-selected sites in the range

of 6.86 to 9.36 m/s at forty meters measurement height. The measurement campaign




started with the investigation of eleven sites; the four most promising sites have been pre-

selected for further wind measurements at a height of 40m a.g.l.

In the next stage of the project development, the preparation of Feasibility Studies were

foreseen, with the sites Mesobo-Harena and Ashegoda being the task of Lahmeyer Inter-

national. The german based GTZ supported EEPCo in the project development by carry-

ing out a tender for Windpark Development in Ethiopia and Capacity Building as stated

in the ToR related to the project. All information concerning the ToR can be found in An-

nex A 7.

The originally foreseen sites of Mesobo-Harena and Nazareth as defined in the ToR, and

as assigned to Lahmeyer International for feasibility analysis, were changed later to the

sites Mesobo-Harena and Ashegoda due to problems with an ETV sending mast on the

Nazareth site. The mast makes the implementation of a wind farm of 40-60 MW, as re-

quested in ToR by GTZ and EEPCo, nearly impossible until end of 2007.

Later on, EEPCo will develop the Feasibility Study for the Nazareth site under supervision

of Lahmeyer International, taking into account the additional time demand. The second

wind park site originally envisaged, to be supervisied by EEPCo at Gondar, was taken out

of the feasibility study due to the limited available space on site, preventing the realisation

of a 40-60 MW wind farm.

This report is focusing on the Ashegoda site.




4 Site conditions

4.1 Site Description

Figure 4-1: Map of Ethiopia with the region of Ashegoda wind park

The proposed wind park site is situated in the northern Ethiopian highland close to the

descent to the coastal plain. The whole area, foreseen for the construction of the wind

park, is covered with small bushes and grass. The land is mainly used for extensive goat

farming, and partly for agricultural use.

The wind park consists of two main areas; the western area is located at two low ridges in

an approximately north-south orientation while the eastern area is spread over a more

distinct hill in north-south orientation with several branches. In between the two parts a




lower plain area can be found where a high voltage line is passing in a north / south orien-

tation. The altitude of the wind park ranges from approximately 2,300 to 2,450 m a.s.l. The

considerable elevation above sea level results in a low average air density of 0.922 kg/m3.

South-east of the eastern part of Ashegoda wind farm, the upper branches of a valley de-

scending to the coastal plain are reaching the highland plain.

Figure 4-2: Ashegoda area, digital terrain model, view from the southwest in a height of 1000 m above ground. A larger

print-out can be found in Annex C-5

The client has provided topographical maps in 1:50.000 and 1:12.500 scale which cover

the full project site. Information on the surrounding terrain is taken from LI s world wide

data base and of the visual impressions during the site visit.

The orographical terrain conditions can be classified as medium complex. The slopes of

the ridges where the wind park has been proposed and the valley towards the coast are

adding some complexity to the vicinity of the wind park site while the remaining area is flat

or modestly hilly.




4.2 Site Limitations

The proposed wind park site has basically favourable conditions for implementing a wind

farm: like available space, no housing directly on site, existing road access, a nearby high

voltage line, a substation in approximately 20 km distance, and as well as favourable wind

conditions.

However some restrictions were identified which will limit the available area for the erec-

tion of wind turbines:

Corridors claimed for aviation by Mekelle airport (see Annex A-1)

Obligations of the Environmental Report (see Annex A-2)

Several villages surrounding the wind park area (see Figure 6-1)

The area north of the road is used by the Ethiopien military radar and can not be

used for the wind park Ashegoda west.

The wind park site will not be affected by the civil and military aviation directly but it is

possible that a demand for aviation lights for all turbines will be made.

For a possible future extension of the current wind park layout there are several options.

One option are further negotiations with the Ethiopian military at Mekelle to shift the given

boundary of the wind park towards north. The second option is to add additional wind tur-

bines in the middle of both wind park areas accepting less energy yield in comparison with

the proposed western and eastern hills of the Ashegoda wind park site.

4.3 Ground and Soil Conditions

A preliminary soil condition investigation was performed during the site visit in February

with the preliminary result that the conditions of the entire site are suitable for flat founda-

tions.

After the preliminary investigation LI s subcontractor has provided a report on the geo-

technical investigation for the planned wind park Ashegoda.

The report describes the regional site geology in general as well as the seismic conditions

at the site.




Figure 4-3: The soil condition on Ashegoda site

The soil investigations have further been based on excavated test pits of different depths

between 0.5 m and 1.0 m. Photos of each pit are attached to the specific report.

The report states that the likelihood of the soil becoming fluidised in the event of an earth-

quake is very remote for the investigated area. But it is also stated that the Ashegoda is

marked as Zone 2 within the Seismic Risk Zoning Map of Ethiopia. Therefore for the final

design of the turbines and foundations a seismic check is recommended according to the

regional risk.

Based on a visual assessment and the examined material in the test pits the bearing ca-

pacity of the soil is conservatively estimated based on the experience of the conductors of

the soil investigation. The estimations underline that sufficient soil properties can be ex-

pected to foresee shallow foundations for the wind turbines. For the final design of the

foundations of each wind turbine a more detailed analysis is required at each location of a

wind turbine, as described in the feasibility study report. It is expected that some optimisa-

tions will be possible based on that more detailed investigation.

The complete soil investigation report carried out is shown in Annex A - 3.




4.4 Access Roads, Availability of Cranes

4.4.1 Road Access

Basically, the transportation of the turbine equipment in Ethiopia will be by road. A paved

main road leading from Mekelle towards the south passes the Ashegoda area nearby.

From the main road an unpaved road is heading through the wind park sites Ashegoda

west and Ashegoda east. No obstacles like trees or signposts exist. The site access road

provides no problem for delivering turbines in the range of 800 kW to the wind park site

but has to be improved and levelled for the transportation of components of larger wind

turbine types. Currently this road is under reconstruction, all further roads required for the

construction of the wind park have to be newly built; a preliminary design has been devel-

oped by the consultant. The exact length taken into consideration can be found in An-

nex C 1.

Figure 4-4: Branch from main road to Ashegoda sites

The total distance of the direct link from Djibouti via Dese and Alamta to Ashegoda is ap-

proximately 695 km (715 km distance Djibouti-Mekelle). The road is a gravel road except

the part from Djibouti to Mile which is a paved road over a distance of 196 km.




As stated in detail in the road condition report (2003/2004) from Ethiopian Road Authority,

attached in Annex A-6, the road from Djibouti via Mile and Chifra to Weldiya is a gravel

road, the condition is mainly good and with a specified width of 7 m. For the road between

Weldiya and Ashegoda no condition report has been provided to LI. Because this road is

one of the main roads in the country, it can be assumed that the conditions of this part of

the route are equal to the conditions of the first part from Djibouti to Weldiya. The axle

loads of the trucks carrying the turbine nacelles and the towers do not exceed 12 tons and

are within the limits that the roads as mentioned above can bear.

For the transportation of the turbines to the site a maximum slope of 10% of the road ac-

cess can be accepted. According to the site visit trip and the measured data by GPS the

existing slope of the gravel road from the main road to the site is calculated to 2.9%

maximum. A description of the maximum axle loads and wheel-bases given by the Ethio-

pian Road Authority are shown in Annex A - 6.

The transportation of wind turbines of the 800 kW class as proposed within this report

from Djibouti Harbour to the wind park site provides no major difficulties. The main prob-

lem is the development of a suitable logistics concept. For example, the hauIage of

86 wind turbines will take about one and a half years when using 10 seperate trucks,

whereas the number of trucks in Ethiopia suitable for the transportation of heavy or long

items like wind turbine components has yet to be evaluated.




Figure 4-5: Road map of north-eastern Ethiopia (red arrow marks the wind park area)

4.4.2 Road Access for larger Turbines up to 2 MW

According to our experience with wind power projects in similar countries, e.g. Djibouti

and Nigeria, there are frequent problems with heavy weight and wide load transports in

case of using large wind turbines of the Megawatt class.

To take into consideration the erection of larger wind turbines up to 1.5 MW or 2 MW in

Ethiopia, the experts of LI carried out a road survey study covering the main roads from

Djibouti harbour to the wind park site. This study was done in June by car starting at Me-

kelle and traveling the route to Djibouti harbour. Some obstacles on the way from the

highland to the lowland were found and described in detail in the Annexed road report. In

the mountainous terrain, between Alamata and Korem there are two sections where the

gradient exceeds the limit for a transport with the weight of a 40 m long, Megawatt-class

blade load. Furthermore, on these points the radii of the curves also do not allow the




transport of these blades. These bottlenecks currently do not allow for the transportation

of large wind turbines in the Megawatt-range.

Besides this, it was stated in the report that the whole transport from Djibouti to the wind

park site at the plateau will take approximately 11 days for a return trip per major compo-

nent for turbines in the kW-class.

In summing up we can establish that without road improvement and time extension the

transportation of turbines up to 2 MW can not be carried out sufficiently at present. There-

fore this report is focussing on turbines below the Megawatt-class.

4.4.3 Railway Transport

In Ethiopia only one railway line, connecting Djibouti and Addis Abeba, exists. The railway

is jointly owned by the countries of Ethiopia and Djibouti.

A railway line to Ashegoda does not exist. Consequently, involving the railway in the

transport chain requires the transfer of the turbine components at Addis Abeba to road

vehicles.

The railway, however, can be an option for the first part of the transport but it has to be

noted that the track and rolling stock of the railway are worn out due to a considerable

lack of investment. The railway is unreliable and can not fulfil the traffic demands as re-

quired.

Additionally, several physical limitations hamper the transport of large goods. The informa-

tion in general, given by the railway company describes the situation as follows:

1. Four freight wagons can be hauled per train as a maximum (but the consultant

observed a much longer train)

2. Maximum weight per train is 160 tons (according to the consultants experience

this could be increased by double-heading the trains)

3. Maximum wagon length is 12 m

4. Due to the limitations given by the tunnels and bridges of the line, the loading

gauge is limited - width 3.7 m and height 3.3 m.

5. A maximum axle load was not specified but by observing track and capacity of

the freight wagons LI assumes that this will not exceed 10 t. A maximum meter

weight was not specified as well.

Besides this information from the railway company, the exact figures of the transport

manuals given by several turbine manufactures are shown in the following table:




Table 4-1 Transport figures of wind turbines

length width weight length width height weight length width weight towerm m t m m m t m m t sections

Enercon E 48; 55m 22.8 1.7 2.3 7.42 4.37 4.54 21 16.93 *3.30 30 section I17.70 2.53 20 section II

0 19.92 1.87 15 section III

Enercon*** E 53/1; 72m 25.26 2.44 2.6 ** 4,50 ** ** 21.85 *4.30 37 section I25.75 3.08 22 section II24.50 2.12 15 section III

Gamesa G 58; 55m 28.5 2.35 3.2 8.00 2.3 2.9 24 9.60 3.31 16 section I19.20 3.02 23 section II24.50 2.43 18 section III

Gamesa G 58; 65m 28.5 2.35 3.2 8.00 2.3 2.9 24 19.00 3.62 31 section I19.20 3.02 23 section II24.49 2.43 18 section III

Vestas V 52; 60m 25.3 2.10 1.9 6.681 1.923 2.819 22 14.00 3.62 24.08 section I19.17 3.03 24.50 section II24.45 2.44 20.43 section III

Fuhrländer MD 70; 65m 34.0 2.6 5.2 10.20 3.75 3.85 58.0 13.00 4.00 38.70 section I19.25 3.84 31.70 section II25.40 3.84 30.00 section III23.90 3.84 27.00 section IV

Vestas V 80; 60m 39 3.520 appr. 6.5 10.05 3.37 4.05 61.2 10.39 4.00 31.73 section I23.82 3.59 51.08 section II24.37 2.78 39.41 section III

The nacelle and blades from Gamesa would remain identicalonly tower specification would change according to the countries.

* outside flange diameter** data currently not published*** for the E 53 the dimension of the prototyp are given by Enercon Germany (E53/1 means the first turbine of this type)

Main Measures and Weights of Wind Energy Converter (samples)

Type and hub height Blades Nacelle Tubular steel tower

Figure 4-6 shows a typical open freight wagon with a payload of 22.4 tons and a length of

12 m which can be used for the transport of rotor blades. These have a weight of less

than 10 tons and a length between 22.8 and 40 m; the overlapping length when loading

them on a 12 m length wagon provides no problem as it is common in such cases for rail-

ways world-wide to attach further flat wagons to both ends of the blade-carrying wagon

which act as idler or distance cars. The clearances in tight curves where the overlapping

part of the blades will swing out over the track have to be checked by the railway, how-

ever.

Concerning the Nacelles, only Gamesa G-58 and Vestas V52 nacelles can be transported

on these wagons: for the other manufacturers they either exceed the height clearances or

the allowed payload of the wagons. Concerning the towers, for most sections, either the




diameter is too large (the height of the floor of the wagon has to be added to the height of

the load) or the weight is beyond the limits of the railway equipment; only some of the up-

per parts of the towers are suitable for the transport on the line.

Further parts can be hauled by the railway if lower wagons are available. It is not known to

the consultant if these are available at the railway. Suitable meter gauge wagons, called

Rollwagen, had been made redundant in Germany, Poland and Switzerland in large num-

bers recently.

Figure 4-6: Typical open freight wagon of the Djibouti Ethiopean railway

Due to the unreliable and bad condition of the total railway system a safe transport over

the long distance of the line can not be assured.

In March 2006 it was announced to contract the operation of the railway to the South Afri-

can company Comazar, which will presumably led to a higher transport capacity and im-

proved reliability of the line. But this situation will not be of any help for the transportation

of parts of the proposed wind turbines for the foreseen implementation during 2007.




4.4.4 Available Crane Capacities

In general two cranes are required for the erection of a wind turbine. One is the so called

main crane and the second the helping crane. The main crane has to be dimensioned

large enough to lift all tubular tower sections for commissioning and finally the nacelle and

blades to the top of the turbine, whereas the helping crane mainly gives assistence to the

lifting process, at a lower point of application than the main crane. Therefore a smaller

crane can be used for the supporting crane.

For the erection of the wind turbines it is foreseen to hire available cranes in Ethiopia. As

a first result of the investigations by the consultant concerning available crane capacities it

can be concluded that there is only one possibility to hire an 81.6 tons crane with a boom

of 37.9 m plus a ten meter extension from the company MIDROC at Addis Abeba. Further

details like daily costs, transportation costs to the site and the availability at the foreseen

erection period in 2007 are depending on detailed negotiations with MIDROC after the

final decision of which turbine type to select for the implementation of the wind park. We

assume that the MIDROC crane, with some small extensions of the existing boom, can be

used as the supporting crane. The information provided by MIDROC is shown in An-

nex A-5.

Optionally, a crane from the construction company Bilfinger & Berger stationed in Nigeria

can be hired. The preliminary information about this crane allows the assumption that this

is a feasible option for 2007 and the Ashegoda wind park site.

In addition, it should also be investigated if the utilisation of the available crane capacity at

Zafarana, Egypt it possible. Zafarana is one of the largest wind park sites in the world,

consisting of several hundred of wind turbines in the 800 kW range. However, it has to be

noted that in 2006 and 2007, two extension stages of the wind park Zafarana are fore-

seen, limiting the available crane capacity.

Finally there is another option of buying a new crane by EEPCo to independently solve

the problem with the required crane. This option will need additional money for the in-

vestment phase of the project, and will be several times more expensive than using a

rented or an overhauled crane. We do not recommend this option for the beginning of the

wind park development in Ethiopia, in order to avoid additional costs, and keeping in mind

that during the warranty period of the wind turbines there will be no need for an independ-

ent crane capacity by EEPCo.

Taking into account the information above, it is likely that there are several options to

bring sufficient crane capacities to the wind park site, allowing EEPCo to erect wind tur-

bines in the proposed turbine class.




We recommend the common option of hiring a crane in close cooperation with the turbine

manufacturer and arrange an agreement for the use of one of the manufactures cranes as

the main crane during the implementation phase of the project. The supporting crane

should be hired from MIDROC to avoid further transportation costs caused by the ship-

ment of a crane from other countries to Ethiopia. This concept has been choosen e.g. for

Zafarana wind park.

Additionally, Enercon India presented and offered their craneless repair concept during a

visit in Addis Abeba through its representative. Further detailed information concerning the

proposed craneless concept was not given by Enercon India so far.

To get a general idea, the following table presents the minimum dimensions of the needed

cranes for the erection of the several turbine types:

Table 4-2: Basic dimensions of the required cranes

WTG Hub heights main crane assistance crane

Vestas V 52, hh 55 m 55 m 250 t 100 t

Enercon E-48, hh 57 m 57 m 300 t 80 t

Gamesa G58, hh 60 m 60 m 350 t 100 t

Vestas V 80, hh 85 m 85 m 600 t 250 t




4.5 Internal Access Roads

The access roads have been preliminary designed and can be found at the wind park lay-

out map presented in Annex C-1. It has to be separated between:

access road to be newly built

existing road to be improved

The following table gives the summary of the different internal roads to be built:

Table 4-3: Overview of the Access Roads distances

TypeScenario I:

Length in km

Scenario II:

Length in km

Scenario III:

Length in km

Scenario IV:

Length in km

Paths to be reinforced /

widened5.0 5.0 5.0 5.0

Roads to be newly built 25.0 25.0 25.0 25.0

4.6 Enviromental Impact Assessment

An environmental impact assessment has been carried out by the feasibility study team of

EEPCo.

No serious impacts besides the permanent loss of agricultural land are to be considered

neither for local inhabitants nor for flora and fauna.

In detail:

The land foreseen for the erection of the wind park is mainly used for pasture farm-ing and agriculture. Approximately 20 hectares of farm land will be permanentlylost for farming while further areas will be not available during the constructionphase of the wind park, the loss of grain is estimated to 100 tons. A compensationis to be paid, the exact figures are displayed in the environmental impact report inAnnex A-2.

During the construction period, care has to be taken that waste disposal and sani-tary requirements are properly defined and implemented and precaution arrange-ments are taken to prevent the spread of infectious diseases.

No serious long-term impact to the local fauna is to be expected, no designatednational wild life parks or reserve areas tap the wind park.




No serious long-term impact to the local flora is to be expected.

No serious long-term impact on birds is to be expected, no migration routes ofbirds are known for the area.

No sanctuaries, no known historical or cultural sites exist in the proposed windpark area.

The Environmental Monitoring Unit (EMU) of EEPCo proposes the implementation of an

Environmental Monitoring Plan during all phases of the project. Details can be found in the

Enviromental Assessment Report Ashegoda in Annex A-2.

4.7 Legal Constraints

Legal constraints at site are not known by the consultant at present.

In any case, as kind of compensation, it should be recommended to provide to the af-

fected communities employment within the project (during construction and operation).

4.8 Earthquake Risk

Eastern and Southern Africa covers a region which is prone to a significant level of seis-

mic hazard due to the presence of the East African rift system. A number of destructive

earthquakes have been reported during this century. In Ethiopia, they include the 1960

Awasa earthquake (Ms* = 6.1), the 1961 Kara Kore earthquake which completely de-

stroyed the town of Majete and severely damaged Kara Kore town, the 1969 Serdo earth-

quake (Ms = 6.3), 1989 Dobi graben earthquake (Ms = 6.5) which destroyed several

bridges on the highway connecting the port of Assab to Addis Ababa, the 1983 Wondo

Genet and the 1985 Langano earthquakes which caused damage in parts of the main

Ethiopian rift (Midzi V. al. 1997).

* Ms = surface-wave magnitude




Seismicity of eastern and southern Africa based on the catalogue compiled by Turyomu-

rugyendo (1996). Earthquake epicentres are shown for MS4.0.

If the project is developed in areas classified as earthquake zones, it must be proved that

legislation in Ethiopia obliges the developer, to approach and consult with an engineer or

a statics specialist approved by the authorities during the project phase. In several coun-

tries including, e.g., Italy legislation requires that a local engineer or statics specialist,

approved by the authorities, must review the statics underlying a given construction de-

sign prior to the erection of buildings and this also includes wind turbine foundations in

earthquake zones. If such a review should reveal that a wind turbine foundation must be

modified to suit local seismic conditions, this is done in co-operation with local authorities

in accordance with specific requirements and in this way, it is ensured that the seismic

requirements are met.

Literature:

Turyomurugyendo, G. 1996. Some aspects of seismic hazard in the east and south African region. Unpub-lished. M. Sc. Thesis, Institute of Solid Earth Physics, University of Bergen, Bergen, Norway, 80p.Midzi V. et al. 1997: Seismic hazard assessment in eastern and southern Africa




5 Wind Resources

5.1 Wind Data Collection

Figure 5-1: Measurement masts at Ashegoda site

Two measurement masts have been erected in the area during the year 2005:

Mast 4 Ashegoda I with a measuring height of 10 m a.g.l. in January 2005 and mast 13

Ashegoda II with two different measuring heights (10 m and 40 m a.g.l. respectively) in

September 2005. Both masts are equipped with Thies first class cup anemometers and a

wind direction vane (Thies compact) at 10 m height a.g.l. The anemometers have been

calibrated according to MEASNET standards. Copies of the calibration protocols have

been handed over to the consultant. The data logger used for the measurements are

manufactured by Ammonit with the basic data given in Table 5-1.

4 Ashegoda I

13 Ashegoda II




Table 5-1: Basic data of the Ammonit data logger WICOM 32

Type Version / date SampleRate

StorageIntervall

Data channels

Wind speed Wind direction

AmmonitWICOM 32

V 1.415th September

20041 Hz 10 Minutes

Average, Mini-mum, Maximum, Standard devia-

tion

Average, Mini-mum, Maximum, Standard devia-

tion

No further data (temperature, air density, and humidity) have been collected.

Basic data of the measurement stations are given in the following Table 5-2.

Table 5-2: Basic data of the measuring masts at Ashegoda site

Wind Meas-uring Mast

Wind Mast Coor-dinates

(UTM WGS 84, Zone 37)

Height a.g.l. Records Avail-ability

x y Altitude[m] [m] From Until*)

4 Ashegoda I 562 882 1 484 021 2,391 10 15.01.2005 09.04.2006 96%

13 AshegodaII

565 488 1 483 611 2,418 10 19.09.2005 19.02.2006 100%

*) measurements still running, date shown is the end of data used within this study

With a data recovery rate of above 96 % for all stations the result of the measurement

campaign can be considered as good.

Mast 13 Ashegoda II was erected as a wired steel tubular tower) manufactured in Ethio-

pia. Although not strictly designed according to IEC 61400-121 Power Performance

Measurement of Grid connected wind turbines the measuring mast 13 Ashegoda II and

the mounting of anemometers and wind vane allows a wind measuring campaign with

acceptable quality levels with the exception of the horizontal adjustment of the upper parts

of the mast (see Figure 5-3) which is not optimal and has to be considered in the uncer-

tainty analysis in chapter 7.5.2.2.




Figure 5-2: Vicinity of measuring mast 13 Ashegoda II (center of figure)

Figure 5-3: Measuring mast 13 Ashegoda II, vertical adjustment




For mast 4 Ashegoda I, a locally available wooden telephone pole has been used (Figure

5-4). The anemometer has been mounted on top of the mast with the aim to measure the

wind flow without influence of the mast itself (Figure 5-5); the vertical distance to the mast

however is small and an interference can not be excluded to all possibility of doubt. In

addition, the wind direction is not properly detected when the wind comes from the upwind

side of the mast due to the mounting of the wind vane close to the pole. However, as this

direction is not the main wind direction, the influence of the tower to the measurements

can be considered as not very significant.

Figure 5-4: Measuring mast 4 Ashegoda I, anemometer and wind vane at 10 m a.g.l.




Figure 5-5: measurement device installation at mast 4 Ashegoda I

The average wind speeds and the frequency distribution of the measurements are pre-

sented in the following table.

Table 5-3: basic wind measured data for Ashegoda site

4 Ashegoda I

10 m a.g.l.

13 Ashegoda II

10 m a.g.l.

13 Ashegoda II

40 m a.g.l.

Wind speed [m/s]

Frequency[%]

Wind speed [m/s]

Frequency[%]

Wind speed [m/s]

Frequency[%]

N 5.51 3.77 2.53 0.40 3.16 0.40

NNE 3.87 1.75 2.79 0.46 3.48 0.46

ENE 6.58 2.85 3.67 0.60 4.57 0.60

E 6.98 6.64 5.62 6.56 6.94 6.56

ESE 9.83 48.92 8.02 87.81 9.35 87.81

SSE 8.45 16.80 6.24 3.43 7.18 3.43

S 3.92 1.50 2.87 0.36 3.14 0.36

SSW 3.91 0.54 2.47 0.11 2.89 0.11

WSW 3.97 0.42 1.84 0.06 2.20 0.06

W 5.22 1.05 1.67 0.06 1.94 0.06

WNW 6.09 6.62 2.01 0.05 2.82 0.05

NNW 6.09 9.14 1.65 0.11 2.24 0.11

total 8.26 100.0 7.69 100.0 8.99 100.0




In graphical format the data are presented as wind roses with the frequency graph scaled

down to enhance clearness:

0.00

2.00

4.00

6.00

8.00

10.00N

NNE

ENE

E

ESE

SSE

S

SSW

WSW

W

WNW

NNW4 Ashegoda 10 m

Frequency

Figure 5-6: wind rose of mast 4 Ashegoda I

0.00

5.00

10.00

15.00

20.00N

NNE

ENE

E

ESE

SSE

S

SSW

WSW

W

WNW

NNW

13 Ashegoda 10 m

13 Ashegoda 40 m

Frequency

Figure 5-7: wind rose of mast 13 Ashegoda II




5.2 Wind Data Analysis

The raw data gathered directly from the data loggers can not be used unprocessed for the

estimation of the energy yield as it may contain erroneous and incomplete recorded data.

Apart from inadequate installation (anemometer not mounted horizontally, wind vane not

aligned to north direction, faulty cable connections, exhausted batteries) the equipment is,

as with every technical device, subject to spontaneous and unpredictable malfunction

leading to data gaps or incorrect data values. Furthermore, external effects like extreme

weather conditions or vandalism may cause similar losses of data and data quality. The

data has therefore to be checked carefully with erroneous values to be sorted out or to be

replaced by suitable ones. Data losses over longer periods of time also distort the infor-

mation.

Lahmeyer International has comprehensively checked the values of the time series of

each measurement mast according to the following steps, removing faulty data:

- the average wind speed at each 10-minutes interval has to be lower than the maxi-

mum value and higher than the minimum, the same for the wind direction

- detecting non plausible peaks and spikes in the time series (i.e. one value of 20 m/s

surrounded by values of 5 m/s)

- carefully checking and interpreting zero values if data loss or calm

- checking low values if being measured value or an offset

- checking low values if being measured or indicating a bearing fault

In a second step, the time series of the further anemometers on site has been used to

cross-check wind speed and direction data for plausibility and for identification of further

conspicuous values by comparing data of different anemometers for the same time pe-

riod.

In order to predict wind data for the gaps in the time series left after the removal of erro-

neous values), the Linear Regression and Wind Index MCP (Measure-Correlate-Predict)

Method has been applied.

During the MCP-Process, the correlation between two time series is used to fill in the gaps

in a time series by extrapolating missing data by using the appropriately adapted - data

for the same time period of a second time series. The MCP procedure can accordingly

be used to extend a time series by addition of adapted data of a second reference time

series. Condition for applying the MCP process is an adequate correlation coefficient

between the time series and commonly shared periods of time.




In case of the measurement campaign at Ashegoda, missing or erroneous data of the

40 m height anemometer of mast 13 Ashegoda II, in operation since 19 th September

2005 has been extrapolated using the MCP method. This anemometer has been used as

the mast is situated within the area of the proposed wind park and 40 m is close to the

proposed hub height of the wind turbines.

Data of the mast 4 Ashegoda I (anemometer at 10 m height a.g.l.) which has delivered

data since January 2005, as well as data from mast 13 Ashegoda II (anemometer at 10 m

height a.g.l.) have been chosen for the MCP-Process. The available data of Mekelle Air-

port has not been selected as there is no common time period to the current measure-

ments, consequently preventing a correlation with the on-site data.

Figure 5-8 shows a correlation diagram as an example; the diagrams for the whole MCP-Process are displayed in Annex B 1.

Figure 5-8: Correlation diagram for mast 13 Ashegoda II, anemometers at 10m (x-axis) and 40m (y-axis) for wind direction

sector 105° to 135°




Table 5-4: Correlation coefficients of the measurement masts used for the MCP-Process

REFERENCEMAST

SITE MAST CORRELATIONCOEFFICIENT

(R)

TIME SHIFT (min) to

13 Ashegoda II40 m-

anemometer

13 Ashegoda II_40mV40ave


95.7% 0


4 Ashegoda I_10m V10ave

80.2% 0

During the MCP process, the time period of the time series of mast 13 Ashegoda II (40 m

anemometer) has been extended from 4.8 months (19.09.2005 11.02.2006) to 14.0

months (15.01.2005 09.04.2006).

The data coverage has decreased from 100% to 95% due to the data availability of 95.7%

of mast 4 Ashegoda I which contributes the majority of data (see Table 5-5) but the quality

of the time series has been increased considerably by eliminating and replacing data

found to be erroneous during the quality check, furthermore the data period has nearly

tripled. Table 5-4 and Table 5-5 provide the ratio between the measured original data of

the 40 m anemometer, and the amount of data extrapolated from the other appropriate

anemometers.

Table 5-5: wind speed data used for MCP-prediction

Data taken from mast Number of measured values

measured data of mast13 Ashegoda II (V40ave)

20857 10-minutes average values

predictions from data of mast13 Ashegoda II ( V10ave)


predictions from data of mast4 Ashegoda I ( V10ave)


Total measurements and predic-tions





Table 5-6: Wind direction data used for MCP-prediction

Data taken from mast Number of measured values

measured data of mast13 Ashegoda II (V40ave)


predictions from data of mast13 Ashegoda II ( V10ave)

none

predictions from data of mast4 Ashegoda I ( V10ave)


Total measurements and predic-tions


5.3 Long-term Correlation

As described in chapter 5.1, the wind measuring campaign at Ashegoda was started in

January 2005 thus delivering data for a period of slightly more than one year while an op-

erational lifetime of at least 20 years is expected for a modern wind turbine. Wind condi-

tions however are not stable over the years and oscillate around the long term average.

Consequently, the result of a 12-month measuring period can not be considered as long

term representative. It is therefore common practice to perform a long term examination of

the wind conditions of the respective site. According to international standards, this has to

be done by comparing and adjusting the measured wind data by means of suitable

mathematical methods with data obtained from long-term measurements - usually mete-

orological stations. A period of at least 10 years (15 to 20 years is preferred) can be con-

sidered as long-term .

In case of Ashegoda site, no suitable data from local meteorological stations were pro-

vided to LI and, as mentioned before, data from an anemometer on the military part of

Mekelle Airport could not be used because of the lack of a common time period to the

measurement data of the proposed wind farm site. Therefore, NCEP reanalysis wind data

of a period of 35 years are used to determine the long-term representative value of the

measurement data. The applied NCEP-Data are created by the US-National Center for

Atmospheric Research and the National Centers for Environmental Prediction (NCEP /

NCAR). Basically, these reanalysis data are the results of a global climate computation

model which uses a large number of filtered and converted historical meteorological data

from 1970 onwards and observations from surface, ships, aircrafts, satellites and other

data sources to recalculate meteorological parameters like wind speed which is given

four times per day from 1970 onwards and being updated every month.




NCEP-data is available for certain grid points at a height of 50 m and 500 m a.g.l. in a

resolution of 2.5 degrees world wide. For the purpose of the Ethiopian wind park evalua-

tion program, further NCEP data for several further points has been provided by RISØ

national laboratory to GTZ which has been handed over to LI. These points give a better

description to the wind climate of the Ethiopian highland close to the drop to the coastal

plain and are given for a period of 25 years in a height of 10 m a.g.l.

Figure 5-9 displays the nearest NCEP 2.5 degree-spacing grid points to Ashegoda re-

spective Mekelle and the RISØ-provided grid points.

Figure 5-9: NCEP grid points around Mekelle (yellow pins mark the 2.5° -spacing NCEP grid points, red pins the location of

the RISØ delivered NCEP points), the red circle idicates the selected NCEP point

To predict the long term wind speed, NCEP data of the grid point East 39.375 and North

12.3808 (red circle in Figure 5-9) were used as having the best correlation coefficient




when compared to East 39.375 North 14.2855. An analysis of the data shows that during

the last ten years no significant variation of the wind speed occurred.

For the common time period shared by both NCEP data and the measured wind speed

data on site the average wind speeds have been calculated and set into ratio, providing a

long-term-coefficient c(25Years) for the wind speed:

min10min10).(

)25(

25 40*)25(40*10

1040 aveVYearscaveV

aveV

aveVaveV

PCNCEP

YearsNCEP

Years

with the individual figures as follows:

average velocity of NCEP at 10m a.g.l, during the common period of measure-

ments and predictions with mast 13 Ashegoda II at 40m a.g.l. (i.e. the measuring

period):

smaveV PCNCEP /935.310 ).(

25 years long term average velocity of NCEP data at 10 m a.g.l:

smaveV YearsNCEP /917.310 )25(

Correction factor, relation between average wind speeds of the measuring period

and the 25-year period

995.0935.3917.3

10

10)25(

).(

)25(

PCNCEP

YearsNCEP

aveV

aveVYearsc

The long-term-coefficient c(25Years) is close to 1.0; it has consequently not been applied

to the correlated 14.0-month time series of the 40 m anemometer of mast 13 Ashegoda II

as the uncertainties of both the measurement and the MCP process are considerably

higher than the deviation of the correlation factor from 1.00.

The long term (25 years) corrected time series for the 40 m anemometer of mast13 Ashegoda II is thus equal to the correlated 14.0-month time series of the 40 m ane-mometer of mast 13 Ashegoda II as displayed in Table 5-6 and Figure 5-10.




Table 5-7: overview of the long term (10-year) wind data at 40 m a.g.l of mast 13 Ashegoda II

Sector Weibull A Weibull k Wind speed

[m/s] [m/s]

N 5.19 3.63 4.68

NNE 4.64 2.33 4.11

ENE 5.64 1.83 5.01

E 8.31 3.93 7.52

ESE 10.09 5.95 9.36

SSE 7.56 2.43 6.70

S 3.49 2.86 3.11

SSW 7.13 1.86 6.33

WAW 3.14 7.17 2.94

W 3.87 3.05 3.46

WNW 6.55 2.68 5.82

NNW 6.93 2.93 6.19

total 9.03 3.32 8.11

0

2

4

6

8

10

12

14N

NNE

ENE

E

ESE

SSE

S

SSW

WSW

W

WNW

NNW13 Ashegoda II 40 m

Frequency

Figure 5-10: Wind rose at 40 m a.g.l. for Ashegoda site, mast 1 Ashegoda30m_40m, 10-year corrected.




Figure 5-11: Summarized Weibull-distribution for mast 1 Ashegoda30m_40m, 10-year corrected

5.4 IEC Wind Class

Wind turbines are subjected to environmental conditions which may affect their loading,

durability and operation. To ensure an appropriate level of safety and reliability, the envi-

ronmental parameters shall be taken into account during the selection of appropriate wind

turbines.

The environmental conditions may be subdivided into normal and extreme external condi-

tions. The normal conditions generally concern long-term structural loading and operating

conditions, while the extreme external conditions represent the rare, but potentially critical,

external conditions such as short-term gusts. Wind turbines are grouped into classes ac-

cording to IEC 61400-1, Rev 3 depending to their ability to withstand defined wind condi-

tions. These classes are characterised by the 10-min average value of the extreme wind

speed with a transgression probability once every 50 years and by the long-term annual

mean wind speed at hub height criteria (required by Rev. 2 of IEC 61400-1 but no longer

used by Rev.3), the sub classes A, B and C (the latter implemented by the latest edition of

the norm, Rev.3) refer to the turbulence intensity at a wind speed of 15 m/s at hub height.




Table 5-8: IEC Wind Turbine classification

WT classes I II III

Vref [m/s] 50.0 42.5 37.5

A I15 0.18 0.18 0.18

B I15 0.16 0.16 0.16

C I15 0.12 0.12 0.12

The figures given in the table are the maximum values for the respective wind classes

Whereas:

Vref : 50-year 10-minute averaged extreme wind speed

A : designates the sub-class for higher turbulence characteristics

B : designates the sub-class for medium turbulence characteristics

C: designates the sub-class for lower turbulence characteristics

I15 : characteristic value of the turbulence intensity at 15 m/s

Vref is determined by applying the corresponding WindPro software tool (using a fitted

Gumbel-distribution, a statistical distribution function) on base of the Ashegoda wind

measurement data.

The prediction is delivering Vref = 24.05 m/s with an uncertainty of 1.97 m/s at the measur-

ing height of 40 m a.g.l.

According to IEC 61400 1, Rev 3 the wind speed is extrapolated to hub height using

equation (1)

)/()( hubhub zzxvzv(1)

The power law exponent has been determined to 0.118 by analysing the correlated

wind data at the heights of 10 m and 40 m. The power law exponent may differ for other

parts of Ashegoda wind park but can be determinded for the position of the measurement

mast only. However, the location of the mast is representative for the site.

This leads to the following:




The results are equal to IEC wind class III for both heights when taking into account the

uncertainties by adding them as a worst-case scenario to the calculated values. As the

boundary value for wind class III is 37.5 m/s, a sufficient distance to the calculated value

is given in case the power law exponent is considerable higher for some points as as-

sumed.

Turbulence is a factor which causes stress and fatigue to several components of a wind

turbine, among them blades, bearings and gearbox. It is usually described as turbulence

intensity which is defined as the standard deviation (which is calculated from measured

wind speed data) divided by the average wind speed. The standard deviation is a statisti-

cal measure describing the deviation of the data points in a set from the average value (in

this case, from the measured average wind speed in the 10-minute-intervall). The average

turbulence intensity of Ashegoda site, measured at the measuring mast 13 Ashegoda II is

10.4%.

Table 5-9 shows the results of the wind class investigation.

Table 5-9: Site classification parameters

50-year 10-minute averaged

extreme wind speed at hub heightvEWS (50) = 25.2 m/s, uncertainty = 1.9 m/s

turbulence intensity 10.4 %

This leads Ashegoda site to be classified as IEC wind class III c, the lowest wind class

meaning that every wind turbine on the current market ist suitable for Ashegoda wind park

in terms of wind class.

40 m Hub Height 60 m Hub Height

Vref

[m/s]

Uncertainty

[m/s]

Vref

[m/s]

Uncertainty

[m/s]

24.5 1.97 25.2 1.9




6 Technical Layout of Ashegoda Wind Park

Lahmeyer International was requested by the client to develop a wind park layout for ap-

proximately 40 - 60 MW installed capacity at Ashegoda. This micro-siting has been done

using the long term experience of LI in international wind projects by means of the wind

industry standard wind farm planning software WindPRO (see chapter 7.2.1).

6.1 Wind Potential Map

Based on the long-term corrected data of the wind measurements on site, a wind potential

map has been calculated (using WindPro and WAsP [see chapter 7.2.2] software), provid-

ing information for the distribution of the average wind speeds over the area foreseen for

the implementation of the wind park (Figure 6-1) and therefore being a base for the wind

park layout development process.




Figure 6-1: Wind potential map of Ashegoda site (red line indicates the road, orange areas mark the villages, yellow line the

wind park area)

North




It can be clearly seen that the ridges west and east of have considerably more favourable

wind conditions than the plain in between, caused by the shadowing effects of the ridges

and the speed-up effect atop a hill described in chapter 7.1.

As mentioned in chapter 4.2, the military installations north of the road (red line) are limit-

ing the wind park area to the region south of the road until it reaches the eastern ridges. It

has to be noted that the exact alignment of the road is not known in the western part, as it

is not covered by the topographical map provided by EEPCo and had to be estimated by

GPS-trackpoints and the impressions gained on site. The villages around the site (orange

areas) are furthermore restricting the available space, as noise emissions from the wind

turbines and the shadow flicker caused by the moving turbine rotors cause stress to the

inhabitants of these villages if specific values are exceeded, see chapters 6.5 and 6.7.

6.2 Wind Turbine Selection

6.2.1 Suitable Tower Heights

Typically, wind speeds are higher with increasing levels above ground. For that reason

higher towers can exploit higher wind speeds so that the annual energy production can be

increased correspondingly. The counteracting effect is the respective additional invest-

ment cost for the tower and the foundation. In Ethiopia, the available crane capacities are

a further limiting factor.

The tower heights for wind turbines are chosen to find a good combination of energy yield

which is increasing with tower height, and costs for tower and foundation, which are in-

creasing with tower heights. Furthermore for larger turbines, the tower heights are also

increasing in general. Especially in lower wind speed areas the tendency towards higher

hub heights can be found.

The following table will present the costs of several turbines types in connection with dif-

ferent hub heights showing the effect of higher hub heights described above:




cost analysis

0.30

0.35

0.40

0.45

0.50

0.55

0.60

40 50 60 70 80 90 100 110 120

hub heights

/kw

h

Vestas V52

Gamesa G58

Siemens Bonus

Enercon E-48

MD70 / MD77

Figure 6-2: Turbine costs in accordance with several hub heights

This calculation was done with official market prices for the wind turbines of the named

manufacturer and the production estimation is connected to the wind conditions at the

Ashegoda wind park site. On the x-axis the standard hub height of the selected turbines in

meters are plotted, and on the y-axis the specific investment costs in /kWh are pre-

sented, calculated by the estimated investment costs devided by the expected annual

energy output (P75-value).

The use of larger hub heights, above 60 m and up to 100m will cause higher costs for

towers, foundations and the erection period, as well as higher crane costs and is not a

satisfactory option for the Ashegoda site.

At the site which is here under discussion, a quite low wind speed characteristic has been

found which has been characterised as wind class III b according to the IEC regulations.

As explained before, the terrain at the site can be described as complex. Therefore the

extrapolation of the wind speed at higher hub heights compared to the height of the

measuring mast is associated with comparatively high uncertainties. These uncertainties

could be reduced using higher measuring masts or higher order models for the flow calcu-

lations like mesoscalic models like KLIMM.




Since none of these options have been used at this site, a hub height has been chosen

which is fulfilling the following two conditions:

- high hub height

- which allows qualified wind regime predictions

Based on LI s experience a meaningful prediction of the wind regime is possible for a site

of these characteristics up to hub heights of 50 to 60 m. We recommend taking towers in

the range of 50-60 m in order to fulfil both conditions. Nowadays these towers are of nor-

mal height for turbines up to 1MW and will be available for the proposed wind turbines.

Therefore the hub heights used for the further evaluation have been chosen in this range.

Based on our information no sufficient crane capacity is available in Ethiopia in any case.

Therefore it has been assumed in the financial model to integrate the crane mobilisation

from outside of Ethiopia. Please refer also to chapter 6.2.2 Determination of optimal unit

size .

6.2.2 Determination of the Optimal Unit Size

The selection of the wind turbine type, suitable for the wind energy application in Ethiopia,

is depending on several criteria, such as:

- Transportation to the foreseen wind park site

- Available space on site

- Orographical conditions on site which may prevent the installation of larger tur-bines in the Megawatt - range

- Local experiences with regular operation and maintenance with wind turbines

- Distances between site and turbine manufacturer who will perform the mainte-nance within the warranty period

- Wind turbines types yet installed in the county

- Energy Yield

- Turbine types available for the Ethiopean market

Wind turbines in the Multi-Megawatt range require a higher level of maintenance to be

performed by the turbine manufacturer - compared to proven turbine types in the range

below one Megawatt. In addition, proven in the past, almost all wind turbine manufactur-

ers are not willing to offer turbines in the Megawatt class for countries which are just en-

tering the wind energy market.




In case of Ethiopia, the project under reference will be the first wind park project to be

implemented in Ethiopia. It is therefore strongly recommended to start with proven wind

turbines in the range below one Megawatt (around 800 kW is todays standard) for the

project. This will have the following advantages when compared to wind turbines with lar-

ger capacity:

- Regular O&M can be performed by local experts and thus a higher availability can

be expected.

- More offers from wind turbine manufacturers will be available in the case a tender-

ing for 800 kW range wind turbines is performed, enabling the Client a more de-

tailed selection.

- The investigation of the transport logistic has been proven that it will be possible to

transport wind turbines in the 800 kW range. For bigger turbines, a detailed road

survey has to be performed as mentioned in section 4.4.2.

- It can be expected that the delivery time for turbines in the Megawatt range will be

longer when compared with smaller turbines. For example, several manufacturers

are not able to deliver any turbines in the Megawatt range in 2006 and 2007.

Consequently, it was decided to focus on wind turbines in the 800 kW range.

The specific investment costs in Euro per generated MWh per year for two selected 2 MW

wind turbines compared to the 0.8 MW turbines of the same manufacturer are displayed in

the following Table 6-1:




Total Invest-Cost [EUR]

Total Invest-Cost [EUR]

plusConstruction

Interests

SingleCross En-ergy Yield

P50 [MWh/y]

Wind Farm Net Energy

YieldP50[MWh/y]

Specific In-vestment

Costs[EUR/MWh/y]

V52 85,148,914 85,148,914 2,706 210,026 405E-48 79,988,914 79,988,914 2,652 205,835 389V80 HH 60m 84,498,914 88,723,860 5,704 173,068 513E-70 HH 65m 85,858,914 90,151,860 6,462 196,067 460

Table 6-1: Specific investment costs for selected turbine types

The lower total number of V80 / E-70 wind turbines compared to the number of E-48 / V52 turbines is due to the effect thatthe distances between Multi-Megawatt turbines have to be considerable larger which may cause a reduced total installedcapacity and thus a lower total energy production compared to a 800 kW / 850 kW wind turbines layout.

This calculation is clearly showing the higher specific investment costs of the Multi-

Megawatt-turbines compared to the 800 kW turbine class. As in contrast to most regions

in Europe the space available at Ashegoda wind park site is sufficient to set up the in-

stalled capacity as requested with the 800 kW wind turbines, this turbine class is preferred

for the project especially when considering that transportation of 2 MW turbines to Ashe-

goda is not possible and taking into account their one-year longer delivery time compared

to the 800 kW turbines.

Requests for Expressions of Interest on supply of wind turbines for wind parks in Ethiopia

had been submitted to the manufacturers given in the following Table 6-2.

Additional Assumptions for V80 and E-701) 50%/ WTG additional transportation Costs2) 60.000 EUR additional Costs for Crane3) 5 days for installation per Turbine instead of 34) 30% additional costs per Foundations5) 5% additional interrests due to extended construction periode6) No Electrical Losses are considered7) Prices for E-70 + V80 from 2005

ParametersPark efficiency Availibility No. Turbines

E-48 hub height 57m 0.95 0.95 86V52 hub height 60m 0.95 0.95 86V80 hub height 60m 0.97 0.92 34E-70 hub height 65m 0.97 0.92 34

Energy Yields [MWh/WTG/y]E-48 hub height 57m 2,652E-70 hub height 65m 6,462V52 hub height 60m 2,706V80 hub height 60m 5,704




Table 6-2: Requests for expressions of interest

Manufacturer

(contact person)

Contact(s) Remarks

Nordex:

Mr. DwengerMarch 06

Nordex is currently completely booked and only 10-12 MD-70 (1.5 MW) turbines would be available for delivering in

2007

REpower Systems:

Mr. FrickeMarch 06 Ethiopia is not in REpower s country portfolio no interest

at present.

Fuhrländer AG:

Mr. Kretz

March, April, May and July

2006

Fuhrländer still needs for further activities a detailed state-ment of EEPCo concerning the projects like time schedule,

project financing and project security.

Vestas:

Mr. Henriksen and Mr. Sondergaard

April, May and June 06

The discussion with Vestas is still ongoing. Vestas standard wind turbines are designed and certified for installation up to 1,000m above sea level maximum. Therefore, the technical support department of Vestas is currently still evaluating how they can quote for that project.

ENERCON:

Mr. HochMarch, April and May 06

Ethiopia is not a key market for Enercon Germany before 2012.

Enercon India:

Mr. Raman

April, May, June and July

06

A first EoI was sent to Ato Kebede in Dec. 05, offering the E-48 with two different hub heights. A detailed offer by En-ercon India for 2 wind park sites is offered to be send by end of July. Enercon India is currently manufacturing the E-48 only.

GE Energy:

Mr. Said (Nairobi)March 06

An enquiry for 1.5MW turbines in general has not been an-swered until now, smaller turbines are currently not avail-able




Suzlon Energy:

Mr. Patel

May and June 06

An enquiry for 600kW and 950kW turbines has not been answered until now

Siemens:

Mr. KruseMay 06 New markets like Ethiopia are not priority markets for Sie-

mens at present.

Gamesa Eolica:

Mr. Artiago

May and June 06

Gamesa is currently completely booked and they would be able for delivery of turbines from the beginning of 2008 soonest.

SeeBA Energie-systeme:

Mrs. Lefevre

March and April 06

SeeBA is a manufacturer of lattice towers for several tower heights and turbine types, e.g. Nordex-, Fuhrländer-, RE-power- and Vestas turbines. This means that SeeBA s an-swer as an associated supplier depends on one of the above mentioned manufacturer s decisions finally.

This reduces the manufacturers coming into question to Enercon India, Vestas and

Gamesa, the latter with the restriction of being not able to deliver turbines in 2007 as en-

visaged for the project.

For the project, the following turbine types have been considered consequently:

- Enercon E-48, 800 kW turbine

- Vestas V-52, 850 kW turbine

- Gamesa G-58, 850 kW turbine as an option for delivery in 2008

- Enercon E-53, 800 kW turbine, (only an option for delivery in 2007 and offered byEnercon-India)

Note: According to the latest information from Enercon, the company is currently develop-

ing a variant of the E-48 wind turbine with a rotor diameter of 53 m and a rated power of

800 kW. The prototype of the Enercon E-53 has been erected in August 2006 in Germany

and serial production is expected to start in 2007. This wind turbine has been included as

an option within this study.




6.3 Turbine Distances

For the micro-siting certain minimum distances between the individual wind turbines have

to be observed. A common rule of thumb specifies three to five rotor diameters in cross

wind directions (less than three is possible under some circumstances) and six to eight

rotor diameters in main wind direction as a minimum spacing between the individual tur-

bines. The minimum distance of three times or less the rotor diameter in cross wind direc-

tion is only feasible in case the wind direction is strictly perpendicular to the row of wind

turbines which can be, due to the given orientation of the ridges on site, be achieved for

some parts of the site, and then only if there is no additional rows of turbines within a con-

siderable distance. The smallest distances in cross wind direction for Ashegoda wind park

shown in Table 6-3 are the result of the layout development iteration process (the layouts

are presented in chapter 6.4) carried out under the condition that the wake losses do not

fall below an average level of at least 85 % for the individual turbines, which has been

considered as necessary for the economical operation of the wind farm. Depending on the

location of the individual wind turbine and the ambient conditions (topography, location of

nearby wind turbines, number of wind turbines towards the main wind direction) the dis-

tance between two adjacent turbines can be larger.

The closest distances between the individual wind turbines are attached in Annex C 2.

Table 6-3: minimal turbine distances for Ashegoda wind park

Turbine type Minimal distance between turbines in cross wind direction

[m] [rotor diameter]

Enercon E-48 170 3.6

Enercon E-53 175 3.3

Vestas V52 185 3.5

Gamesa G58 185 3.2




6.4 Wind Park Layouts

Figure 6-3: Digital height model, view from a height of 1000 m height south-east to Ashegoda wind park. A larger print-out

can be found in Annex C-5

The park layouts have been developed by LI on base of the wind potential map and the

topographical situation, taking into account the limiting factors already discussed in chap-

ters 6.1 and 6.3.

The wind potential map (see Figure 6-1) provides information of the areas with the most

favourable wind conditions (the areas shown in greenish colours, followed by light blue);

wind speeds at the plain (dark blue) are up to 2 m/s lower compared to the ridges, result-

ing in a concentration of the wind turbines on top of the latter.

In order to minimise the wake losses, single lines of wind turbines with considerable dis-

tance between the individual rows are more preferable than clusters of turbines, leading to

the layouts shown in Figure 6-4 to Figure 6-6. Detailed layout maps including internal

roads and cabling are attached in Annex C 1, the exact turbine coordinates are given in

Annex C - 2.

The orientation of the central part of the eastern ridge nearly parallel to the main wind di-

rection (southeast) prevents the more intensive usage for wind turbines as these will be

situated behind each other in the main wind direction while the village at the eastern edge




of the wind park area limits the length of the southeasternmost row of turbines. Further

wind turbines at the northern end of the row will cause the exceedance of the noise level

limits for the village. Similar limitations exist for the northern end of the westernmost row

of turbines.

6.4.1 Wind Park Layout Enercon E-48

Figure 6-4: Ashegoda Wind Park, Layout Enercon E-48




6.4.2 Wind Park Layout Vestas V52

Figure 6-5: Ashegoda Wind Park, Layout Vestas V52




6.4.3 Wind Park Layout Gamesa G-58

Figure 6-6: Ashegoda Wind Park, Layout Gamesa G-58




6.4.4 Wind Park Layout Enercon E-53

Figure 6-7: Ashegoda Wind Park, Layout Enercon E-53

6.4.5 Conclusion

The four wind turbine types under consideration allow the implementation of a wind park

of the envisaged installed capacity within the foreseen area. Similar size of the turbines

and a rotor diameter in a close range (from 48 m [Enercon E-48] to 58 m [Gamesa G58])

leads to a similar basic park design which has then been optimized in terms of park effi-

ciency and energy production.




6.5 Turbulence

To ensure the close spacing of the wind turbines will not affect (decrease) the lifetime of

the turbine and its components, a turbulence calculation is necessary which has been

carried out by LI. The turbulence of the wind flow is a factor which causes stress and fa-

tigue to several components of a wind turbine including blades, bearing and gearbox. It

consists of the so called ambient turbulence applied to the wind flow by the coarseness of

the earth (vegetation, buildings, rocks etc.) and the turbulence added by the other wind

turbines of a wind park.

The impact of the turbulence to each individual wind turbine has been calculated and ana-

lysed by means of the WindPro software package using the Frandsen Turbulence Model6

(as recommended in IEC 61400 1, Rev 3) for GFK-Materials (rotor) with the results of

the analysis, presented in Annex C 3, compared to the limits given by IEC 61400 1,

Rev 3.

The calculated annual average, direction weighted turbulence for each individual wind

turbine has to be lower than the critical values of 16 % turbulence intensity for a IEC class

A turbine and 14% for a class B turbine at a wind speed of 15 m/s. Furthermore, the cal-

culated annual average, direction weighted turbulence curve has to remain below the the

IEC A and B curves for the whole range of wind speeds occurring on site. The IEC

wind clases of the selected wind turbines are as follows:

Table 6-4: turbulence sub-classes of the selected wind turbines

Turbine type Turbulence sub-class

Enercon E-48 A

Enercon E-53 A

Vestas V52 A

Gamesa G58 A

In case of Ashegoda wind park, at 15 m/s no exceedance of the critical A value of 18%

has been calculated for the chosen layouts but the IEC A curves will be exceeded for

wind speeds of more than 15 m/s for the majority of the selected wind turbine positions

when using the Frandsen-model.

6 Sten Frandsen and Morton L. Thorgersen: Integrated Fatigue Loading for wind turbines in wind farms by combining ambi-ent turbulence and wakes




New studies show that the Frandsen-Model overestimates the turbulence applied to the

turbines. For the calculations, the Empirical Turbilence Dutch TNO Laboratory 1993 Model

has been selected alternatively; the results show then no exceedance of the limits.

A consultation of the manufacturers of the wind turbines for checking the turbulence im-

pact is generally required; for the proposed wind park layouts for Ashegoda site however,

no problems in terms of turbulence intensity are to be expected.

Figure 6-8: Turbulence vs. IEC characteristics (conditions fulfilled)

6.6 Noise Impact

The target of the noise assessment is to investigate the potential noise impact of the wind

turbine operation on sensitive areas in the vicinity of the wind farm. The advisable dis-

tances between residences and the proposed wind turbine sites depend on a variety of

factors including local topography, eventually background noise and the size of wind farm

development. Official demands with regard to noise limit values for the operation of a wind

park in Ethiopia are not specified. Therefore a prediction of the sound produced by the

proposed wind farm in the surrounding area and an optimisation of the micrositing was

made in accordance to the strict German noise limit regulations.

The calculation method is specified in ISO 9612-2 and implemented in the WindPro soft-

ware used for the estimation of the noise effects.

The sound emission data used in the calculation and the sound power level of the turbine

bases on information given by the turbine manufacturers.

Correspondingly the following standard values for noise emission are considered depend-

ing on the utilisation of the area:




Table 6-5: Noise standards according to German standards

Utilisation Noise emission [dB(A)]Day time

06:00 22:00

Night time

22:00 06:00Regimen and hospital

areas 45 35

Exclusive residential ar-eas 50 35

General residential areas 55 40

Village centres, mixed utilisation with small

trades60 45

Working areas 65 50

Industrial areas 70 70

Considering that identified noise sensitive areas can be assigned to the Village centres

with mixed utilisation , the limiting noise standard for the operation of the wind farm is an

impact level of 45 dB (A) at night time.

The results of the calculations are showing no conflict in terms of noise level, the bound-

ary levels for the noise emissions during the night are not exceeded for the emission

points (houses of a village nearest to the wind park, churches) in the vicinity of the pro-

posed wind farm.

The detailed results can be found in the Annex B 2.




6.7 Shadow Impact

When the sun is just above the horizon, the shadows of the wind turbine generators can

be very long and can move across houses (windows) for short periods of time. If this hap-

pens for longer period, it causes stress to the inhabitants.

The exact position and time period of shadow can be calculated very accurately for each

location, taking into account the structure of topography and the movements of the sun.

Official Boundary levels are not existing for the shadow flicker effect. In Germany, a com-

monly accepted value is the maximum of 30 hours shadow caused by the wind turbines

per year, and 30 minutes shadow per day.

WindPro software has been used for the calculation of the shadow impact.

For the Enercon E-48 wind turbines, no exceedance of the limits has been calculated,

whereas the Enercon E-53, Vestas V52 and Gamesa G58 wind turbines will cause an

exceedance of the limits for the village immediately east of the wind park. However; the

estimations are carried out for the worst case that the sun is always shining, 365 days

per year. An additional calculation has been performed using the real sunshine probabil-

ity . This data has been estimated by considering the (average) rainy days per year in

northern Ethiopia as days without sunshine; this approach can be considered as ade-

quately conservative; no exceedance of the limits of the maximum shadow impact can be

expected now. The estimated sunshine probability is displayed in Table 6-6.

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

Rain days 0 0 1 1 4 5 8 7 1 1 0 0

Probability of rain 0.00 0.00 0.03 0.03 0.13 0.17 0.26 0.23 0.03 0.03 0.00 0.00

Probability of sunshine 1.00 1.00 0.97 0.97 0.87 0.83 0.74 0.77 0.97 0.97 1.00 1.00

Table 6-6: Sunshine probability at Mekelle

The detailed results can be found in Annex B - 3.




7 Energy Production Estimation

The calculation of the wind resources on-site and the corresponding energy production

are based on the processed wind data collected by the measuring masts at Ashegoda

(see chapter 5.1) with the energy yield estimation carried out by means of the WindPro

and WAsP software packages.

7.1 Meteorology

Generally, wind flows are large-scale balancing air movements in the atmosphere be-

tween high pressure and low pressure-areas, in some cases superposed by local balanc-

ing movements. The pressure differences are caused, and driven by, the solar heating of

the Earth s atmosphere, landmasses and water bodies. Warm air is rising, leaving a low

pressure area at the surface and causing high pressure areas within higher layers of the

atmosphere where the air is cooling down and falling to lower layers with respective pres-

sure effects.

Since the balancing air movements also run near the earth s surface they are significantly

affected by the surface roughness and the orographic terrain structure.

Roughness is a quantitative description for the friction of the surface causing a slowdown

of the near-surface air flow. Water and land without vegetation has a low roughness and

therefore a minimal effect to the wind, forests or cities have a high roughness.

Figure 7-1: Wind flow over forest (source: GASCH 1996)

The influence of the roughness to energy of the wind flow is shown in Figure 7-5.




Figure 7-2: Wind flow energy vs roughness class (source: WindPro manual)

On top of a hill, the air flow is compressed and thus the wind has a higher energy density

compared to the same air flow over flat terrain. Furthermore, valleys and similar terrain

structures can act as a blast pipe, concentrating the air flow. Behind mountain ridges and

inside valleys the air flow is disturbed by recirculation and shading effects.

from: Harald Frater, Wetter und Klima. Phänomene der Erde

The influence of the topography is shown in the energy calculation results in Annex C 2

in Production Analysis




7.2 Software Basics

7.2.1 WindPro

The WindPRO philosophy is object orientated projecting. A wind park project consists of a

number of objects, whereby the wind turbines are key elements. Objects are: wind tur-

bines, wind monitoring stations, local obstacles etc. Some of the objects directly deter-

mine the wind energy, others focus on the environmental aspects and yet some others

can also influence the feasibility of the project. The characteristics of the objects are com-

bined with the terrain data such as the surface roughness maps and the height level con-

tour lines and serve as basic input data for the energy calculations with WAsP, which is

the WindPRO integrated meteorological simulation model.

7.2.2 WAsP

WAsP is a software program for predicting wind climate and energy yield of wind turbines.

The predictions are based on wind data measured on site or from stations in the same

region and considers the effects of the surrounding terrain to the wind flow (topography,

surface description, obstacles).




WAsP uses the wind atlas methodology. In a first step the influence of the terrain to the

wind flow is calculated, creating a generalised regional wind climate.

In a reverse process, this generalised regional wind climate (called wind statistic) is then

applied to topography, surface description and obstacles at the vicinity of each individual

wind turbine, providing the wind flow at this point even if the wind data has measured in

some distance which can be, depending on the terrain, up to several kilometres.

For the energy yield calculation as well as for the wind prediction analyses within this

study the standard WASP-WindPro model has been generated using a digital terrain

model in a radius of at least 20 km around the centre of the site. A digitised roughness

map has been evaluated also in radius of at least 20 km. In a radius of about 1 km all ma-

jor obstacles have to be considered in the model, but in the case of Ashegoda site there

are no such obstacles.

7.3 Model Input Parameters

Besides the processed wind data, important input parameters for the energy yield calcula-

tion model are as described the terrain description, comprising of the orographical (topog-

raphical) model and the description of the earth´s surface (roughness description).

7.3.1 Orography

The orographical terrain data has been gathered from the SRTM -Shuttle Radar Topogra-

phy Mission data base which is provided by the US Geological Survey. The mission has

scanned the earth´s surface between 60° latitude north and 54° latitude south, with a

resolution of 90 x 90 m. The data in between these points has been interpolated using the

WindPro software to gain the topographical model of Ashegoda area, with a height con-

tour density of 5 metres, as shown in Figure 7-3(a larger print-out can be found in Annex

C-4). Orographical data of the 1:12.500 scale topographical map provided by EEPCO has

been used for the wind park area itself. The gained data has been compared with the

Ethiopian topographical maps in 1:50.000 scale to ensure the plausibility of the orographi-

cal terrain data.




Figure 7-3: SRTM height contour map of Ashegoda

The height data provides a three-dimensional digital elevation model of the wind farm siteas shown in Figure 7-4 with a larger print-out to be found in Annex C-5

Figure 7-4: three-dimensional digital elevation model of Ashegoda site, view from south-western direction.




7.3.2 Roughness

Abbildung 7-1: Surface of the project site

For the model simulations the roughness classification for the surface in close proximity to

the wind farm site is derived from topographical maps, data obtained during site visits as

well as aerial photos of the region. A basic description of the roughness classes is given

in the following (roughness length is a second roughness description unit), it has to be

noted that the roughness class is a defined value which can not be measured directly. The

roughness length describes the height where the wind speed in a logarithmic wind profile

is becoming zero; the coarser the surface, the higher the roughness length.




Figure 7-5: Roughness classes (source: WindPro Manual)

The roughness description for Ashegoda site is presented in the following Figure 7-6.




Figure 7-6: Roughness map of Ashegoda site (red ellipse: wind park site)

7.3.3 Wind Shear

The upper anemometers of the wind measurement had been installed at a height of 40 m

above ground while the hub height of the proposed wind turbines is between 55 m and

60 m above ground level.

To estimate the wind regime at hub height of the wind turbines (the WAsP model simpli-

fies the wind speed distribution over the rotor as concentrated to the hub height) the wind

speed is extrapolated according to the following formula:




old

newoldnew h

hvv

Whereas is the Power Law Exponent or Wind Gradient which is depending on the

roughness of the terrain, the thermical lamination of the atmosphere, the wind speed itself

and the height above ground. For Ashegoda wind park, the Exponent has been calculated

sectorwise by means of the WindPro-Software package from the measured wind data at

the heights of 10 m a.g.l. respectively 40 m a.g.l.

The wind gradients for the twelfe wind distribution sectors which had been used for the

generation of the wind statistic are presented in Table 7-1, the calculated wind profile (in-

crease of the wind speed with height) of sector south-southeast (main wind direction) as

an example is displayed in Figure 7-7.

Under consideration of terrain and roughness, the wind profile for the area of each indi-

vidual wind turbine is then calculated by WAsP.

Table 7-1: Hellmann-exponents, sector-wise, for the location of met mast 1 Ashegoda30m_40m

Sectors Hellmann-Exponent

N 0.152

NNE -0.0218

ENE -0.2513

E -0.102

ESE 0.0288

SSE 0.0657

S -0.0483

SSW 0.5547

WSW -0.0303

W 0.0009

WNW 0.2859

NNW 0.3181

Average 0.1184




Figure 7-7: Wind Profile for Ashegoda wind measurement

The extrapolation from 40 m measuring height to 60 m hub height can be considered as

reasonable but a further increase of the wind speed for hub heights of 80 m and more is

not stringent as the power law is not suitable for height extrapolations over this range.




7.4 Wind Turbine Parameters

Four scenarios have been selected:

1. 61 x ENERCON: E-48; Hub Height 57 m

Nominal power: 0.800 MW

Control system: Pitch

Rotor diameter: 48 m

2. 60 x VESTAS: V-52, Hub Height 55m




3. 57 x GAMESA: G-58; Hub Height 60m




4. 61 x ENERCON: E-53; Hub Height 57 m optiononal







7.4.1 Power Curve and Air Density

The Power curve of a wind turbine is an important parameter, describing the relation be-

tween the wind speed on site and the respective electrical energy output.

Power curves and ct-values (a parameter for the calculation of the wake effect) of the tur-

bines under consideration are given in Annex C - 2 and are applied for the energy calcula-

tion. The parameters are provided as follows:

Table 7-2: Sources of the power curves

Enercon E-48 Enercon E-53 Vestas V52 Gamesa G58

Origin CalculatedEnercon India

CalculatedEnercon Germany

CalculatedVestas

CalculatedGamesa

Date 07/03/2006 September 2005 11-2004 4/10/2002

Note guaranteedpower curve

guaranteed power curve

guaranteedpower curve

Official power curve

Power curves which had been measured by independent institutions are of higher quality

than calculated ones. Due to the fluctuations of both the characteristics of the wind turbine

components, and the measuring conditions power curves of different measurements dif-

fering slightly between each other. Furthermore, the measurement does not provide the

explicit power curve used for the calculations; the values are scattered around the mean,

as displayed in Figure 7-8 (source: German measuring protocol of Windtest KWK).

Figure 7-8: measured power curve with scatter band




Several manufacturers are thus providing power curves which are calculated from the

results of several measured ones; the performance of these calculated power curves

might be contractually guaranteed by the manufacturers.

The power curves of the Enercon E-48 and Vestas V-52 wind turbines are such power

curves while the power curves of the Gamesa G-58 is a non-measurement based calcu-

lated power curve. The E-53 is currently in the planning stage thus the power curve is only

calculated theoretically but will be revised after the measurement of the prototypes.

During the calculation of the energy yield, the power curves, given for the standard condi-

tions of air density = 1.225 kg/m3 are adapted to the air density of each individual turbine

location at hub height, with the transformed power curves for the average air density at

Ashegoda site to be found in Annex C 2.

The air density at Ashegoda is calculated by WindPro for each individual wind turbine

according to the site conditions, height above sea level plus the hub height of the turbines

of 57 m / 60 m and an annual average temperature level of 17.7°C; the air density ranges

from 0.914 kg/m3 to 0.926 kg/m3. The temperature data is taken from the Tekeze River

Basin Integrated Development Master Plan, Vol XI, Water Resources, Climatology,

May 1998. As verification, Asmara meteorological station was chosen as the nearest sta-

tion in the data base implemented in WindPro which is located at an area similar to

Mekelle (Ethiopean Highlands close to the descent to the coastal plain) and provides an

annual average temperature level of 16° C which is within the same range.




0

100

200

300

400

500

600

700

800

900

1 3 5 7 9 11 13 15 17 19 21 23 25

wind speed [m/s]

po

wer

[kW

h]

Enercon E-48 Vestas V52 Gamesa G58 Enercon E-53

Figure 7-9: Power curves of the wind turbines under consideration for Ashegoda wind park for an air density

of = 1.225 kg/m3

As can be clearly seen, the Enercon E-53 and Gamesa G-58 wind turbines are generating

more energy in the wind speed range from about 6 m/s to 12 m/s which occur more fre-

quently then the other wind speed ranges, see the Weibull-distribution in Figure 7-10. This

is mainly caused by be larger rotor diameter compared to the Enercon E-48 and Vestas

V52 wind turbines.

Figure 7-10: Weibull distribution of the correlated wind data of mast 1 Ashegoda 30_40m




7.5 Losses and Uncertainties

Meteorological phenomena can only be predicted to a certain limited degree. As a conse-

quence it is not possible to make an exact forecast of the wind conditions even if long-

term reference data (which can only represent the past) is used.

Furthermore, data collection and processing is always afflicted with errors and inaccura-

cies as is every mathematical or physical model used to describe or predict real proce-

dures.

To compensate the inaccuracies in modelling approach and basic input data, it is advis-

able to use factors of safety to adjust, or discount the final output.

Two blocks determine the factors of safety: losses and uncertainties.

7.5.1 Losses

Losses are found on the whole energetic transformation chain from the rotor (kinetic en-

ergy) to the substation (electrical energy). The losses are simple add-ups to the total re-

duction of the calculated energy yield. In detail:

7.5.1.1 Park Efficiency

After passing the rotor of a wind turbine, the wind has a decreased speed due to the ki-

netic energy taken away by the rotor and increased turbulence caused by the rotating

rotor and the difference in speed compared to the undisturbed flow. Until the speed differ-

ence to undisturbed flow is not equalised, the result is a lower energy yield for the wind

turbines following in the direction of the flow. These losses are called array or wake

losses.

The calculation of the wake losses of the wind turbines causing the so called shadowing

effect between the wind turbines has been carried out using the wake model PARK

which is part of the WindPRO software with the array losses of the individual wind turbine

layouts of Ashegoda wind park calculated as follows:




Table 7-3: Array losses Ashegoda wind park

Turbine type Array losses

Enercon E-48 5.8%

Vestas V52 5.0%

Gamesa G58 5.3%

Enercon E-53 5.7%

The values are within an acceptable range for a wind park with several rows of closely

spaced wind turbines.

7.5.1.2 Turbine Availability

The turbine availability is the percentage of a year (i.e. 8760 hours) where the turbine is

able to generate electrical energy while being connected to the grid.

Reasons for the non-availability of a wind turbine are various, and include downtimes for

regular maintenance and servicing, component failures (including defect sensors), over-

heating of components, repairs or exchange of components, as well as errors and down-

times of the superior electrical grid.

The turbine availability is set to 95 % as a standard value according to LI´s long term ex-

perience, taking into account that no experience in the operation of windturbines in Ethio-

pia exists at present, and that all local staff have to be well educated and trained in the

first operational years of the windpark. The proposed value is comparable with average

values achieved in other wind parks abroad, and particularily in new wind markets where

the total infrastructure for a high level operation of wind farms have to be build first. Later

on and after the first years of sufficient turbine availability averages, the aim should be to

raise the level up to 97 % which could be seen as a good value for the first wind parks in

Ethiopia. To reach 98 % turbine availability as in high developed countries like in Europe,

well educated and experienced staffs are needed.

7.5.1.3 Electrical Losses

The electrical losses depend on the resistance of the conductors and on the current inten-

sity. To assess the current intensity of the wind park the following methodology is used:

the duration curve of the wind park for one year is approximated by a two stage approach.

For 10% of the year (876 hours) full power is assumed and for the rest of the year (7,884

hours) the load is estimated to 25% of full load.




For this power pattern of the wind park, the turbine current is derived and the losses for

the internal wind park cabling, and for the wind park cabling connection to the substation

are calculated relative to the total energy output. For electrical losses LI has calculated

the following values:

- internal park cabling: 0.2%

- internal park transmission lines: 0.5%

- turbine transformers: 1.1%

- transmission line 230kV: 0.5%

- substation transformer 0.5%

The total electrical losses of the park are estimated at 2.8% of that amount of electricity

which is produced by the wind turbines.

7.5.1.4 Miscellaneous Losses

In addition to the transmission losses and lost production due to reduced availability a

number of other losses should be taken into account.

The aerodynamic turbine performance described by the power curve is strongly depend-

ing on the profile and surface of the wind rotor blades. Blade fouling from dirt or insects on

the surface of the blades lead to non-expected change of airfoil characteristics and to

lower energy yield.

The average air density of 17° in 2.400 m above sea level provides no thermal problems

for the wind turbines. No effect of the low air density is to be expected than the reduced

energy production.

High wind control losses are caused by the turbine cut-in and cut-out strategy. The tur-

bines will cut-off when cut-out wind speed is reached and will not re-cut-in until wind sped

is below a defined wind speed level, lower than the cut-out level.

A reduction factor of 0.1% to the gross energy output was applied to take these effects

into account.

7.5.2 Wind Speed related Uncertainties

Uncertainties cover the inaccuracy of the data processing from the measurement, the in-

ternal data processing and the long-term prediction. A percentage value describes the




standard deviation of scattering results around the expected true value. For the energy

calculation, these wind speed-related uncertainty values have to be transformed to the

energy production level. Additional uncertainties have to be determined for modelling and

mathematical algorithms.

7.5.2.1 Uncertainties of the WAsP-Model

The WAsP (Wind Atlas Analysis and Application Program) software is a proven tool used

in the wind industry for more than 15 years. As every model it has limitations and uncer-

tainties mainly due to the simplifications behind it which had been done to handle the cal-

culations on desktop computers in an acceptable time frame. Mesoscalic meteorological

models require powerful computers and a calculation time of several days. However,

WAsP has been used for a considerable time worldwide and the uncertainties have been

evaluated over the years. In case of Ashegoda, the existence of modestly shaped hills

and ridges lead to a suffiecient quality of the calculation.

Transfer Wind to Energy

The discrete wind flow from discrete wind directions is simplified to 10-minute average

values for 12 direction sectors and statistically preprocessed before being applied to

the power curve. The uncertainty of this step can be set to 1%.

Site modelling

Consists basically of two input parameters: the topographical model and the surface

description (roughness description). The topographical data is gained from the digital

1:12.500 map delivered by EEPCO to LI for the wind park area itself and from the

SRTM (Shuttle Radar Topographical Mission) Satellite height data base for the vicin-

ity of the site. Quality and resolution of both data are good, the uncertainty is low. As

the surface structure of the earth in the area around Mekelle is not complex, it can be

described with acceptable accuracy.

The uncertainty of the Site Model is set accordingly to the below-average value of

2.5%.

Flow modelling

WasP has been developed for use in areas with only modestly shaped hills which is

the case at Ashogoda site. The uncertainty of the flow modelling is set to the medium

value of 3.0% .

Wake modelling




The uncertainty of the selected wake model (N.O.Jensen) is low and as the area of the

wind farm itself is nearly plain (no additional uncertainties due to the influence of the

terrain) it can be set to 0.5%.

wind-related uncertainties for the WAsP-model

SourceUncer-tainty Comments

Transfer Wind to Energy 1.0% typical value

Site Model 2.5% below average

Flow Model 3.0% medium average value

Wake Model 0.5% typical value

Uncertainty WAsP 4.1%

To calculate the total uncertainty all single uncertainties can be considered as stochasti-

cally independent and the commonly used way of estimating the joint uncertainty of inde-

pendent (un-correlated) uncertainties is to calculate the RMS (root mean square) value.

Total wind speed related uncertainty WAsP = 041.0005.003.0025.001.0 2222

7.5.2.2 Uncertainties of the Wind Data

The reliability of the WAsP calculation is highly dependent on the quality of the input pa-

rameters of which wind data is the most important one. The collection and processing of

wind date is subject to several uncertainties.

Anemometer calibration

Anemometers should be calibrated in order to secure that the measured wind

speed equals the actual wind speed. The anemometers of the Ethiopian wind

measurement campaign have been calibrated according to MEASNET standards,

the calibration protocols have been handed over to LI. The assignment of the indi-

vidual calibration protocols to the individual anemometers of the measuring cam-

paign is not possible but as nothing significantly conspicuous has been detected

the uncertainty of the calibration process can be set to the average value of 1.5 %.




Anemometer characteristics

Describes the uncertainty of the quality the anemometer detects the wind flow and

processes the values to digital data. Can be set to the lower value of 0.5% as cali-

brated first class anemometers have been used.

Mounting error

The anemometer has to be vertically mounted. The uncertainty describes the ef-

fect if this is not done properly. As can be seen in Figure 5-3, mast 13 Ashegoda II

is not properly vertical aligned. For the other mast the mounting is more accurately.

The average uncertainty for the mounting error is set to 1.0 %.

Data recording

Describes the uncertainties related to processing and storage of the data provided

by the anemometer and the wind vane in the data logger. Set to 0.4 %.

Terrain description

Describes the uncertainty of the influence of the terrain to the measurement. In-

clined wind flow and strong turbulence can not be measured accurately by a cup

anemometer. At Ashegoda site, mast 4 Ashegoda I is located close to a steep de-

scent in main wind direction (height difference 20 m), mast 13 Ashegoda II is situ-

ated on a small plateau on a modestly shaped ridge. The uncertainty for the terrain

description is set to the average value of 1.5 %.

Long term correlation

The data used for the long term correlation as well as the MCP-Process includes

uncertainties; as long-term reference data of 25 years is available the uncertainty

for this category can be set to the moderate value of 3.0 %. This can be decreased

when using measured long-term data near the site (NCEP data are recalculated

data).

To calculate the total uncertainty all single uncertainties can be considered as stochasti-

cally independent and the commonly used way of estimating the joint uncertainty of inde-

pendent (un-correlated) uncertainties is to calculate the RMS value. The total uncertainty

of 4.1 % refers to the wind speed at hub height for each single turbine.




Uncertainties for Wind Database

Source Uncertainty Comments

Anemometer calibration 1.5% typical average value,

Anemometer characteris-tics 0.5% typical lower value,

Mounting error 1.0% increased value

Data recording 0.4% typical value

Terrain description 1.5% modestly shaped terrain

Long term correlation 3.0% data base satisfactorily

Uncertainty Wind 3.9%

Total uncertainty Wind Data = 039.003.0015.0004.001.005.0015.0 222222

7.5.2.3 Total Wind related Uncertainty

The total wind related uncertainty is the RMS value of Total uncertainty WAsP and Total

uncertainty Wind which is 5.6%.

7.5.2.4 Uncertainties of the Power Curve

It has to be considered that the power curve used for the gross energy calculation is also

subject to uncertainties which had been described in chapter 7.4.1. Due to the non linear

relation of mean wind speed and energy these uncertainties can not be integrated into the

uncertainties of wind conditions but have to be dealt with separately. In chapter 7.5.3 the

calculation of the uncertainties of the energy yield for the different wind turbine types is

performed; the uncertainty of the power curve is, assuming the power performance of the

turbine as independent of the energy deviation due to wind uncertainties, connected to the

uncertainties of the energy yield by the following equation:

Total uncertainty = 22 curvepoweryuncertaintyieldenergyyuncertaint




The turbine supplier usually gives a guarantee of 95 % of the energy values, which leads

to an uncertainty to the predicted figures of 5 %; the actual guarantee value has to be ne-

gotiated with the manufacturer, the uncertainty can be adapted accordingly. This figure

has been taken for every wind turbine type as it is sufficiently conservative for both calcu-

lated and measured power curves.

7.5.3 Uncertainties Energy Yield

The interpretation of uncertainty in energy yield from the total uncertainty in wind speed is

not straightforward. The theoretical cubic relation of wind speed and energy does not give

a correct description of the phenomena.

For the long term mean wind speed averaged over all turbine locations at hub height the

average wind speed value is derived from the wind data processing. The uncertainty is

equivalent to a reduction to the mean wind speed when considering the worst case. To

translate this reduced mean wind speed into energy yield the parameters of a Weibull

distribution are adapted and this new Weibull distribution is then applied to the individual

turbine power curves. The results for the considered wind park layout can be found in the

examinations in the following tables Deviation of Energy due to wind uncertainties , indi-

cating an energy deviation due to the uncertainty in wind speed assessment.




7.5.3.1 Enercon E-48

The transformation of the wind speed related uncertainties into the energy related uncer-

tainty by appropriately reducing the wind speed at hub height leads to the following results

displayed in Table 7-4:

Table 7-4: Deviation of Energy due to wind uncertainties

Meanwind

speed[m/s]

Deviationwind

speed

A-factor[m/s] k-factor

EnergyYield

[MWh/y]

DeviationEnergy

calculated WAsP 8.78 0.0% 9.75 3.62 252,175 0.0%

reduced wind speed; uncertainties taken into

account8.29 5.6% 9.20 3.62 217,290 13.83%

Assuming the power performance of the turbine as independent of the energy deviation

due to wind uncertainties, the total uncertainty for energy yield can be determined from

the uncertainties of wind conditions (13.83 %) and power curve (5%) to 14.71 %

( 22 5)83.13( ).

The analysis of uncertainties is an important step for the risk assessment of the project.

From the predicted annual energy and from the total uncertainty on the energy level of

16.5 % the probability of exceeding of certain energy yields can be calculated by statistical

methods. Applying a Gauss process for the statistic analysis, the calculated gross annual

energy can be understood as the mean annual energy yield having the highest rate of

probability of all single results. The uncertainty shall be understood as standard deviation

of the expected results around the most probable event.

Figure 7-11 displays the probabilities that a certain amount of annual electricity production

is exceeded. Gross annual energy describes the energy yield as calculated and net an-

nual energy the energy yield considering the losses and uncertainties.




150,000

160,000

170,000

180,000

190,000

200,000

210,000

220,000

230,000

240,000

250,000

260,000

270,000

280,000

290,000

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%

gross annual energy net annual energy

Probability of Exceedance

50 %

75%

90%

MWh/y

197,392

177,82

219,133

Figure 7-11: Probability of exceedance for Ashegoda wind park, Enercon E-48 layout

Besides the uncertainties for wind conditions and power curve, the losses for electricity

transmission (2.8 %) and reduced availability of the turbines (95 %) have also to be con-

sidered as constant factors, reducing the estimated energy yield.

For the Enercon E-48 800kW wind turbine described within section 6.2 the energy calcula-

tions, the results for different levels of exceedance are displayed on the following table:




Table 7-5: Energy Calculations for Ashegoda Wind Park, Enercon E-48 layout

Enercon E-48 800 kW; 57 m hub height at Ashegoda Wind Park

Turbine TypeEnercon

E-48Enercon

E-48Enercon

E-48Enercon

E-48

Turbine Capacity kW 800 800 800 800

Number of WTG 86 86 86 86

Installed park capacity MW 68.8 68.8 68.8 68.8

Hub Height m 57 57 57 57

Rotor Diameter m 48 48 48 48

Specific Rotor Area m2/kW 2.26 2.26 2.26 2.26

Probability % 50 75 90 95Gross energy

productionMWh/y 252,175 227,155 204,637 191,160

Wind park array losses % 5.8 5.8 5.8 5.8

Turbine availability % 95.0 95.0 95.0 95.0

Electrical losses % 2.8 2.8 2.8 2.8

Miscellaneous losses % 0.10 0.10 0.10 0.10

Net Output MWh/y 219,133 197,392 177,824 166,113Specific Energy

ProductionkWh/m2 1,408 1,268 1,143 1,067

Full load hours h/a 3,185 2,869 2,585 2,414

Capacity Factor % 36.4 32.8 29.5 27.6




7.5.3.2 Vestas V52





Meanwind

speed[m/s]

Deviationwind

speed


EnergyYield

[MWh/y]

DeviationEnergy

calculated WAsP 8.84 0.0% 9.81 3.62 248,854 0.0%


account8.35 5.6% 9.26 3.62 218,642 12.14%



the uncertainties of wind conditions (12.14 %) and power curve (5 %) to 13.13 %

( 22 5)14.12( ).



15.10% the probability of exceeding of certain energy yields can be calculated by statisti-

cal methods. Applying a Gauss process for the statistic analysis, the calculated gross an-

nual energy can be understood as the mean annual energy yield having the highest rate

of probability of all single results. The uncertainty shall be understood as standard devia-

tion of the expected results around the most probable event.







150,000

160,000

170,000

180,000

190,000

200,000

210,000

220,000

230,000

240,000

250,000

260,000

270,000

280,000

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%



50 %

75%

90%

MWh/y

198,771

181,38

218,084

Figure 7-12: Probability of exceedance for Ashegoda wind park, Vestas V52 layout




For the Vestas V52 850kW wind turbine described within section 6.2 the energy calcula-





Table 7-7: Energy Calculations for Ashegoda Wind Park, Vestas V52 layout

Vestas V52 850 kW; 60 m hub height at Ashegoda Wind Park

Turbine TypeVestas

V52Vestas

V52Vestas

V52Vestas

V52








productionMWh/y 248,854 226,816 206,981 195,110






ProductionkWh/m2 1,194 1,088 993 936


Capacity Factor % 34.1 31.0 28.3 26.7




7.5.3.3 Gamesa G 58





Meanwind

speed[m/s]

Deviationwind

speed


EnergyYield

[MWh/y]

DeviationEnergy

calculated WAsP 8.87 0.0% 9.84 3.62 299,064 0.0%


account8.37 5.6% 9.28 3.62 265,865 11.10%




( 22 5)10.11( ).



14.02% the probability of exceeding of certain energy yields can be calculated by statisti-






is exceeded. Gross annual energy describes the energy yield as calculated, and net an-





190,000

200,000

210,000

220,000

230,000

240,000

250,000

260,000

270,000

280,000

290,000

300,000

310,000

320,000

330,000

340,000

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%



50 %

75%

90%

MWh/y

239,804

220,49

261,258

Figure 7-13: Probability of exceedance for Ashegoda wind park, Gamesa G58 layout




For the Gamesa G58 850kW wind turbine described within section 6.2 the energy calcula-





Table 7-9: Energy Calculations for Ashegoda Wind Park, Gamesa G58 layout

Gamesa G58 850 kW; 60 m hub height at Ashegoda Wind Park

Turbine TypeGamesa

G58Gamesa

G58Gamesa

G58Gamesa

G58








productionMWh/y 299,064 274,505 252,402 239,174






ProductionkWh/m2 1,150 1,055 970 920


Capacity Factor % 40.8 37.4 34.4 32.6




7.5.3.4 Enercon E-53


tainty, by appropriately reducing the wind speed at hub height, leads to the following re-

sults displayed in Table 7-4:


Meanwind

speed[m/s]

Deviationwind

speed


EnergyYield

[MWh/y]

DeviationEnergy

calculated WAsP 8.79 0.0% 9.75 3.62 286,452 0.0%


account8.30 5.6% 9.20 3.62 251,976 12.04%




( 22 5)04.12( ).



14.88 % the probability of exceeding of certain energy yields can be calculated by statisti-











180,000

190,000

200,000

210,000

220,000

230,000

240,000

250,000

260,000

270,000

280,000

290,000

300,000

310,000

320,000

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%



50 %

75%

90%

MWh/y

227,278

207,56

249,183

Figure 7-14: Probability of exceedance for Ashegoda wind park, Enercon E-53 layout




For the Enercon E-53 800kW wind turbine described within section 6.2 the energy calcula-





Table 7-11: Energy Calculations for Ashegoda Wind Park, Enercon E-53 layout

Enercon E-53 800 kW; 57 m hub height at Ashegoda Wind Park

Turbine TypeEnercon

E-53Enercon

E-53Enercon

E-53Enercon

E-53








productionMWh/y 286,451 261,271 238,608 225,045






ProductionkWh/m2 1,313 1,198 1,094 1,032


Capacity Factor % 41.3 37.7 34.4 32.5




7.6 Summary

The four layouts scenarios for Ashegoda wind park show the following energy yield, re-

lated to the P75 value:

Net Energy Production,198,771




0.000

50.000

100.000

150.000

200.000

250.000

300.000

En

erg

yin

MW

/h

Vestas V52 Enercon E-48 Gamesa G58 EnerconE-53 Probability 75%

Miscellaneous losses

Electrical losses

Turbine availability

Wind park array losses

Net Energy Production

Gross Energy production

226,816 KW/h

Gross energy production

274,505 KW/h


227,155 KW/h


261,271 KW/h

Figure 7-15: P75 energy production of the different scenarios of Ashegoda wind park

The higher energy yield calculated for the Gamesa G58 and Enercon E-53 wind turbines

is mainly related to the larger rotor diameter of these turbines compared to the Ener-

con E-48 and Vestas V52 turbines.

The focusing to the generated energy yield however is not sufficient. Investment costs,

indicated by the ratio specific investment costs ( per kWh) are more significant, for de-

tails refer to the economical part of the Feasibility Study for the presented specific data in

per kWh.




8 Internal Wind Park Cabling

8.1 Cabling Concept

There are two common base alternatives for an internal cabling concept: radial feeders

and ring feeders.

The ring feeder concept is the most reliable concept, based on the n-1 criteria. N-1 criteria

ensure that the disconnection of any equipment of the network is allowed without serious

consequences for the total network. This is valid for the grid devices like cables, trans-

formers, substation busbars, etc. In the case of a cable section fault, the correspondent

cable section will be disconnected automatically and all wind turbines still keep in power

production, supplied through the operative cable sections in both directions, as shown

schematically in the following figure.

Fault

Switching station

Wind turbine

Cable section

- closed disconnector

- opened disconnector

Power flow Power flow

Figure 8-1: Principle scheme of a ring concept

However, the ring concept is more expensive than the radial concept because the double

cable length (preferably even in a separate trenches), additional two disconnectors for

each cable section as well as an additional switching station feeder for each ring in order

to ensure the ring concept.




The radial feeder concept is more economical than ring feeder. There is only one way

cable necessary for each area. In case of a wind turbine or transformer fault, the corre-

sponding device will be disconnected and the remaining wind turbines stay connected to a

feeder and still produce power. The main disadvantage of this concept is the low reliability

in case of a cable section fault (short circuit). The whole feeder will be disconnected for a

time period for the repairing works. The principle scheme of the radial concept is shown in

Figure 8-2.

Fault

Switching station

Wind turbine

Cable section

- opened disconnector

Figure 8-2: Principle scheme of a radial concept

Considering area and the economical aspect of the Wind Park, the radial feeder is reliable

option for connecting all turbines in one line for Ashegoda Wind Park. The experience with

wind parks in the last years shows, that due to the very low error rate of ground cables,

the connection via radial feeders is an economical solution as well.




Figure 8-3: Scheme of the internal Ashegoda wind park cabling

It is advantageous to implement the internal cabling with several radial feeders in order to

increase the generation availability (in case of above described cable section fault only

one feeder will be disconnected and remaining one will still supply the power to grid).

Considering the size of the Ashegoda Wind Park, five similar feeders are recommended to

implement. The switching station is located approximately in the middle of the wind park

as shown schematically in Figure 8-3.




8.2 Cable Type

The internal cable connection of the wind turbines is realised by 33 kV underground XLPE

aluminium cables. The 33 kV is one of the standard medium voltage level used in Ethio-

pia. Underground cables are necessary in wind parks because the trucks and cranes

need free space for operation.

Since the voltage level of the wind turbine generator is 400 V (Enercon E-48), at the base

of each wind turbine, transformer generates the 33 kV for internal park transmission.

The selection of the optimal cable type depends on both the arrangement and grouping of

wind turbines (number and power of turbines) and also the choice of feeder concept. All

wind turbines are divided into five groups: 2 groups consist of 13 wind turbines with ca-

pacity of 10.4 MW each, two groups consist of 18 wind turbines with capacity of 14.4 MW

each and one group with 24 wind turbines with a power capacity of 19.2 MW. This group

allocation is more reliable, from the technical point of view, such as reduction in electrical

losses, reliability and availability in fault cases etc.

The selection of the optimal cable type is standard cable cross sections used according to

Wind turbine specification with respect to its thermal and mechanical stress. Considering

load factor, operating temperature, climatic and operational factors, the suitable cable

cross section is chosen on basis of the standard local cable characteristics.

The values are calculated for one, to two, cable circuit, in the trench, U = 1.05 Urated, cos

phi =0,95 and rated current density from 0.12 to 1.75 A/mm2 for aluminium conductors.

The overhead line cross section of 125 mm2 for the aluminium/steel conductors is suffi-

cient for transmission of required feeder load from each group (See Chapter 8.4).

Due to the selected radial feeder for internal cabling concept, only the 33 kV overhead line

sections of each feeder connected to the switching station shall be designed to be able to

transfer 10.4 MW, 14.4 MW and 19.2 MW respectively.

Due to the selected radial feeder for internal cabling concept, the cable sections shall be

designed to be able to transfer an installed power from 9.6 MW to 10.4 MW. In order to

optimise the cable costs, at least one cross sections of cable design with safety margin of

120 mm², aluminium XLPE conductor cables will be used. The cost estimation is based on

the internal uniform cable design with safety margin: 120 mm² aluminium conductor cable.




8.3 Earthing Network

Due to the design for the wind park Ashegoda, and availability of information about the

Mekelle Substation, the earthing cable to be used in the wind farm shall be at least 95

mm² Copper cable.

The earthing network is composed by the wind turbines earthing system, and the wind

farm earthing connection. In order to have a system at the same electrical potential, each

wind turbine is connected to, at least, one other wind turbine.

Figure 8-4 Earthing network

The earthing network of the wind farm is composed of copper cables of 95 mm² to be in-

stalled directly on the trench.

EARTHING CABLE

EARTHING CABLE




Earthing rods shall be solid, copper-clad steel rods with a minimum diameter of 16 mm

with provision for coupling together with a suitable clamp for connection of the ground

wire. The copper coating shall have a minimum thickness of 0.3 mm. The earthing rod

arrangement shall have the principle layout.

The ground wire shall be directly connected to the pole with bolted connectors, of an ap-

proved material, suitable for use with the ground wire such that galvanic action, i.e.

chemical reaction between copper and galvanised steel, is minimised.

Connections to the substations earthing grid shall be made by compressed clamps or

bolted connectors of approved design.

The earthing of the wind park shall be connected to the existing ground network in the

substation.

The terminal tower should always be connected to the substation earthmat. The reason

for bonding the terminal tower to the substation earthmat is to obtain very low impedance

at the terminal tower in order to prevent a back-flash at the terminal tower in the event of

lightning striking the terminal tower.




Figure 8-5 Terminal Tower connected to substation earthmat

If the terminal tower cannot be bonded to the substation earthmat, the tower footing resis-

tance of the thermal tower on its own should then be reduced to less than 10 ohm in order

to still prevent back-flash at the terminal tower, in the event of lightning striking the termi-

nal tower.

8.4 Determination of Wind Turbine Groups

The proposed wind turbines will be arranged in five groups where two groups divided into

equally installed power capacity of 10.4 MW and the remaining groups each with 2x18

and 1x24 units of 14.4 MW and 19.2 MW respectively like shows in Table 8-1. The divi-

sion into the groups with 13 to 24 units is advantageous from the technical point of view

such as reduction of electrical losses, reliability and availability in fault cases etc.




Area No. of units 0,8 MW Installed PowerMW

Group A 18 14.4

Group B 24 19.2

Group C 13 10.4

Group D 18 14.4

Group E 13 10.4

Total 86 68.8Table 8-1 Wind Park division by four areas

For better visualization, wind turbine groups shown in the Figure 8-3

8.5 Switching Station

Within the proposed internal cabling concept, there are five incoming radial feeders to the

switching station busbars and one outgoing direct connection to the existing high Voltage

overhead transmission line to the Ethiopian power grid.

It is advantageous to locate the 33 kV switching Station in the central area of the Wind

Park (86 units). The length of cable and overhead lines can be reduced with the proposed

implementation and problems regarding electrical parameters such as voltage drop,

higher power losses, cable costs etc. can be avoided.

For the transmission of maximal 68.8 MW, an 80 MVA 230 kV / 33 kV power transforma-

tor is needed. There are five 33 kV overhead lines feeders incoming to the substation

busbar. One of 33 kV and one of 230 kV busbars are needed. The proposed substation

design includes secondary systems and measurement equipment.

In order to optimise the reliability and maintenance costs, at least, a SF6 gas insulated

switching station is recommended for implementation, due to the atmospheric conditions

at the 2400 m altitude of the Ashegoda wind park. The switching station is located cen-

trally between the proposed wind turbine groups and close to the existing 230 kV over-

head transmission line, dividing the wind park line into five radial feeders with 2x13 and

2x18 and 1x24 wind turbines accordingly. The distance between switching station and

wind turbine should be minimum 500 m.




9 Grid Connection

9.1 Grid Integration Concept

Different potential grid integration concepts have been investigated. In order to concep-

tulize installation capacity of 68.8 MW, transmission on the high voltage level should be

considered for approx. 24 km air-distance between the wind farm site and the Mekele

substation. The most practical concept for grid connection is the construction of a sepa-

rate substation at the wind park area (230 kV / 33 kV), where a short high voltage single

overhead transmission line brings the generated electricity in direct connection into the

existing 230 KV overhead transmission line to the existing transformer station at the Me-

kele substation at the 230 kV busbar. The capacity of 68.8 MW is technically possible

within the existing grid at the moment with respect to the data provided by the EEPCo.

This option is favourable for technical as well as economical reasons. A grid connection

via underground cables should be considered as a subordinate option (due to enormous

additional cost) only if the construction of an OHTL is not possible.

Investigated concepts for the grid integration of the Ashegoda Wind Park:

1. Connection to the 15 kV Grid

An economical solution of the wind park integration is to connect the wind park with a

15 kV overhead transmission line to the existing 15 kV busbar at 230 / 132 kV / 15

kV at the Mekele substation but according to the information, provided by EEPCo,

there are no available electrical connection (capacity) at the substation for the

planned installation power of 68.8 MW. However, this option would be cause a great

increase of electrical losses and voltage drops.

2. Connection to the 132 kV Grid via 132 kV Overhead Transmission Line

Transformation from 33 kV to 132 kV directly at the wind park site, transmission with

132 kV to the Mekele substation and connection to the existing 132 kV busbar. Ac-

cording to the planned connection for the Wind Park Mesobo-Harena (48.8 MW) at

this voltage rate and the provided information by EEPCo there is no more available

electrical capacity at the substation for the planned installation power range of 86.6

MW. This option recommends a new investment for a transformer at the Mekele sub-

station as well as the 24 km 132kV overhead transmission line from Ashegoda site to

the Mekele substation.




3. Direct Connection to the 230 kV Grid via existing 230 kV Overhead Transmission

Line

According to our recent inspection of existing overhead transmission line of EEPCo,

the connection to the 230 kV grid via existing 230 kV OHTL can be realised. There

are two possibilities for a direct grid connection of the wind park. The first one is a so-

called T-of Joint connection, where a short high voltage, single overhead transmission

line (one line bay), bring the generated electricity in the Ethiopian national grid. This

option may have a few disadvantages regarding reliability of systems during outages,

operation and maintenance; moreover in case of any fault around the T-Joint area will

cause a loss of the total generated electricity contribution of the wind park.

A appropriate connection type of this wind park with the existing 230 KV overhead

transmission line, as it can be shown on the attached Figure 9-3, is designed to be as

Line In Line Out (LILO) configuration, where two short high voltage, overhead trans-

mission lines (two line bays), bring the generated electricity to the Ethiopian national

grid. This configuration will provide good protection and a reliable system during out-

ages, Operation, and more security during maintenance. For example, when any

faults either on the right line section (Mekele direction) or left line section (Alamata di-

rection) occur, still there will not be total disconnection of the wind park from supply-

ing Power in the Ethiopian national grid.

Having the above mentioned advantages of the LILO type of configuration, careful se-

lection and detailed definition of all the switching equipment, Protection and control

devices and their features for the substations Mekele and Alamata must be done in a

detailed grid study. Moreover, it is important to mention the additional cost of the two

complete in and out going line bays at the wind park station.

Modern wind turbine (pitch-concept wind turbine) installations feature a grid feeding

system that meets the latest grid connection requirements and can therefore easily in-

tegrated in any supply and distribution structure, especially when stipulated require-

ments, such as voltage frequency and reactive power for each individual turbine in a

wind farm have to be considered. The concept offers solutions such as reactive

power management and voltage control for normal operation as well as for critical

situations resulting from network short-circuits or bottlenecks, leading the wind tur-

bines to provide maximum grid compatibility due to their control and operating mode.

Output peaks do not occur due to the closed-loop and open-loop control concept. The

grid feed system allows the wind turbine to operate within a wide range promoting re-

liable operation in weak grids. This enables wind energy converters to support the

electrical grid even at complex locations.

An example for a reliable and modern system is the Enercon technology where the

energy generated in the annular generator is fed to an inverter via a rectifier and a so-




called DC link; it ensures that output power is regulated according to grid specifica-

tions.

In order to provide reliable economical grid operation, power feed timing has to be

regulated. To ensure that this takes place, variable set point values for maximum

permitted power gradients can be specified for the most wind turbines grid feed sys-

tems. For example, when the wind turbine or wind farm is started up, power feed can

be controlled according to requirement. This allows the grid operator to optimise load

flow and grid voltage stability as well as the interaction between power supply com-

panies and consumers.

With the regard to carrying capacity of the existing 230 kV AAAC, 2x180 mm² conduc-

tor, according to the information provided by EPCo., it can accommodate over 150

MW power. Presently it serves to provide power to the northern part of the grid only

as a radial network. But in the near future when the new Tekeze power plant starts to

function in 2008, it will be the main power transmission line from Tekeze power plant

and thus it requires immediate reinforcement. After 2008 with the proper development

on the indicated grid infrastructure and the simultaneous demand rise of the northern

grid consumption it is envisaged that this line will be definitely capable of transmitting

power to the grid from Tekeze and wind park power plants as well.

Figure 9-1 Geographical layout of the Ashegoda wind park grid connection




With regards to the planned installation capacity of total 68.8 MW, generally a transmis-

sion on the high voltage level for Ethiopia in this area (230 kV) should be considered for a

approximately 24 km distance between the wind farm site and the Mekele substation. The

most practical and economical concept for grid connection in this site is the direct connec-

tion to the existing 230 kV overhead transmission line for the planned radial feeders. This

option is favourable for technical as well as economical reasons; in this case the proposed

option would save the cost for the construction of a 24 km new overhead transmission line

to the Mekele substation as well as the investment in a new transformer for Mekele. As

already mentioned a grid connection via 132 kV overhead transmission line should be

considered as a subordinate option (due to significant additional cost) only if the direct

connection at the existing 230 kV overhead transmission line is impossible.

The principle wind park grid connection layout is shown in Figure 9-1. The red dotted line

shows schematically the 230 kV overhead transmission line connecting the switching sta-

tion in the middle of the wind park and the 230 kV/132 kV/15 kV Mekele substation.

Figure 9-2: Principle wind park grid connection layout




9.2 Overhead Line

As considered in chapter 9.1 the grid connection shall be implemented through five over-

head lines feeders (internal connection of the five groups, Table 8-1) to the separate sub-

station 230 kV / 33 kV. The distance between planned feeders and the proposed substa-

tion 230 kV / 33 kV varies approximately from 1 to 3 km where a short high voltage over-

head transmission line (230 kV) brings the generated electricity into the existing high volt-

age transmission overhead line and from there to the transformer station at the Mekele

substation at the busbar 230 KV.

The optimal conductor type is selected for the maximum installed power of the whole ra-

dial feeders. The selection of the optimal conductor type has been done by the standard

overhead line used within the Ethiopian power grid with respect to its thermal and me-

chanical stress. Considering load factor, operating temperature, climatic and operational

factors the suitable conductor cross section is chosen on basis of the standard local cable

characteristics.

Aluminium Conductors, Steel Reinforced (ACSR) 125 mm² would be used to transmit the

electrical power from each radial feeder to the proposed switching station in the Wind

Park Ashegoda and from there the existing 230 kV overhead transmission line close to the

wind park and from there to the Mekele substation. This conductor is designed so that it

can reduce losses and reach at the highest efficiency.

The values are calculated for one overhead transmission line circuit, U = 1.05 Urated,

cos = 0.95. The cable cross section of 125 mm² has sufficient reserves for transmission

of over 10.4 MW to 19.2 MW

To be on the safe side, approximately 10% of the total internal cable lengths are assumed

to deviate for the external 33 kV overhead line due to differences in the air-line distances

and cable laying at the real site between wind turbines and the wind park substation, and

also the planned radial feeders.




WTWT

800 KW18 units

800 KW24 units

800 KW18 units

WT

GROUP A GROUP B GROUP D

WTWT WT

WTWTWTWT

WTWTWTWTWT

T1230/132/15 kV

T2230/132/15 kV

Existing overhead line 230 KvAAAC 2X180mm²

approx. 24 km

T3230/33 kV

Single-circuit ACSRoverhead line 33 kVapprox. 2,8 Km

BUS33 kV

SUBSTATION230/33 kV

BUS230 kV

MEKELE SUBSTATION

800 KW13 units

GROUP C

Single-circuit ACSRoverhead line 33 kV

approx. 2,3 Km

Single-circuit ACSRoverhead line 33 kV

approx. 1,2 Km Single-circuit ACSRoverhead line 33 kVapprox. 2,2 Km

WTWT

800 KW13 units

WT

GROUP E

Single-circuit ACSRoverhead line 33 kVapprox. 2,8 Km

T4230/132/15 kV

BUS230 kV

ALAMATA SUBSTATION

Existing overhead line 230 KvAAAC 2X180mm²approx. 110 km

BUS230 kV

Figure 9-3 Wind Park grid connection schemes

9.3 Transmission Line

The air-line distance between proposed wind park substation location and

230kV/132kV/15kV substation Mekele is approximately 24 km. The overhead transmis-

sion conductor type is the existing AAAC 180 mm² (Aldrey). The selection of the conduc-

tor type has been done by the standard overhead line used within the Ethiopian power

grid with respect to its thermal and mechanical stress, considering load factor, operating

temperature, climatic and operational factors.




9.4 Mekele Substation

The 230 /132 / 15 kV Mekele substation is the air insulated outdoor substation, located

approximately 24 km from the wind farm site. The substation supplies the region Mesobo

and is connected to the high voltage ring as shown schematically on the Figure 9-3.

The substation has to be extended with an additional feeder including 230 kV 2x40 MVA

transformer field and two circuit breakers.

Figure 9-4 Mekele substation




TEKEZE

1x10 MVARWITH CB

WU

KR

O

Pr = (ONAF)2x63/40/23 MVA

V=230±10x1%/132/15 kVTap Changer = ON-LOAD

24515

2321230 209

15

1372 132

MEKELE

CP=80

Pr = (ONAF)2x63/40/23 MVA

V = 230±10x1% /132/15 kVTap Changer = ON-LOAD

205 15

664 66

ALAMATA

1371 132

CP=77

1x15 MVARWITH CB

2320230

2x7.5 MVARWITH CB

249 15

CP=76Pr = (ONAF) 1x20/12/8 MVA V = 132±10x1% /66/15 kVTap Changer = ON-LOAD

1x15 MVARWITH

DISCONNECTOR

Prated = 1x6.3 MVAV = 66±7x1.43%/15kVTap changer=ON-LOAD

665 66

MAYCHEW 206 15



666 66

667 66

LALI

BE

LA

SE

KO

TA

2322230

10 xx.x

105

kMZ

EB

RA

/AC

SR

484.

5

Pra

ted

=1x

25M

VA

V=

66±7

x1.4

3%

/15k

VT

apch

ange

r=O

N-L

OA

D

MERLO/ACSR 65.75

141 kM

48 kM

ASH/AAAC 2x180

80kM

ME

RLO

/AC

SR

65.7

5

105

kMM

ER

LO/A

CS

R65

.75

Figure 9-5 Single line diagram of Ethiopian power grid section including MEKELE substation

According to the possibility about a grid connection of the wind parks in scaling form in

order to bridge supply gaps in the Ethiopian electrical distribution network, the consultant

doesn t see problems for its implementation. In principle the consecutively grid connection

of the turbines or turbine-groups is technically possible taking into consideration the fol-

lowing aspects:

Electrical infrastructure as switch-station, transformers, overhead transmission lines,

cable cross section and trenches for the proposed number of radial feeders had to be

performed, parameterised and defined for the total proposed installation power of the

wind park

The proposed substation for each wind park has to be previously installed with all the

necessary electrical parts and components able to transmit the whole planed installa-

tion power




10 Estimation of costs

10.1 Investment Costs Estimation

The itemised specification of investment costs are described in the table below. Further-

more a proposal of a financing cash flow broken down in foreign and local cost compo-

nents in order to increase the use of local participation and local materials in as many as-

pects of the work as possible is shown in the following investment tables.

10.1.1 Enercon E 48 Investment Costs

Table 10-1: Total Investment Cost of Wind Farm Ashegoda

Investment Cost (in ) Cost in Birr (ETB) % Cost (in USD ) Foreign invest Local invest1 86 Turbines (FOB) incl. Erection 60200000,0 624.935.118,9 75,26 72.498.723,1

Sea transport and inland transport 8600000,0 89.276.445,6 10,75 10.356.960,4Crane 300t, incl. sea transport 120000,0 1.245.717,8 0,15 144.515,7Crane worksInstallation (3 days per turbine) 602000,0 6.249.351,2 0,75 724.987,2Local installation 430000,0 4.463.822,3 0,54 517.848,0

Subtotal 69.952.000,0 726.170.455,7 87,45 84.243.034,6 84.243.034,62 Civil works

Road access 750000,0 7.785.736,5 0,94 903.223,3Crane pads 215000,0 2.231.911,1 0,27 258.924,0Foundation 2666000,0 27.675.698,1 3,33 3.210.657,7Cable trenches 450000,0 4.671.441,9 0,56 541.934,0Control building 50000,0 519.049,1 0,06 60.214,9

Subtotal 4.131.000,0 42.883.836,8 5,16 4.974.953,9 4.974.953,93 Required electrical equipment

Extension of substation 200000,0 2.595.245,5 0,31 301.074,4Civil works new substation 120000,0 1.245.717,8 0,15 144.515,7Transformer (230kV / 33kV) 2000000,0 20.761.964,1 2,50 2.408.595,5Auxiliary equipment of substation 150000,0 1.038.098,2 0,13 120.429,8230 kV components 1200000,0 12.457.178,4 1,50 1.445.157,3

OHL 1x1x125mm 2; 18 km (33kV) 225000,0 2.335.721,0 0,28 270.967,0

OHL 1x180mm 2; 1,0 km (230kV) 125000,0 1.297.622,8 0,16 150.537,2Wind park cabling, earthing, Scada (26 km) 494000,0 5.128.205,1 0,62 594.923,16 x distribution stations on site 120000,0 1.245.717,8 0,15 144.515,7Electrical equipment inside control building 120000,0 1.245.717,8 0,15 144.515,7Auxiliary transformer at control building 15000,0 155.714,7 0,02 18.064,52 x cars for maintenance team 100000,0 1.038.098,2 0,13 120.429,8

Subtotal 4.869.000,0 50.545.001,6 6,09 5.863.725,6 5.863.725,64 Engineering

International enginering 550000,0 5.709.540,1 0,69 662.363,7 662.363,7Local engineering 150000,0 1.557.147,3 0,19 180.644,7 180.644,7

Subtotal 700.000,0 7.266.687,4 0,88 843.008,4 662.363,7 180.644,75 Others

Mitigation measures 336.914,2 3.497.500,0 0,42 405.745,0 405.745,0Subtotal 336.914,2 3.497.500,0 0,42 405.745,0 405.745,0

Total 79.988.914,2 830.363.481,5 100,00 96.330.467,5 84.905.398,3 11.425.069,2

Total Investment Cost of Wind Farm Ashegoda

Wind farm Development EthiopiaPROJECT: Wind Farm at 'Ashegoda' site, Ethiopia

PHASE: Final FEASIBILITY STUDY




10.1.2 Vestas V52 Investment Costs








OHL 1x1x125mm 2; 18 km (33kV) 225000,0 2.335.721,0 0,26 270.967,0

OHL 1x180mm 2; 1.0 km (230kV) 125000,0 1.297.622,8 0,15 150.537,2Wind park cabling, earthing, Scada (26 km) 494000,0 5.128.205,1 0,58 594.923,16 x distribution stations on site 120000,0 1.245.717,8 0,14 144.515,7Electrical equipment inside control building 120000,0 1.245.717,8 0,14 144.515,7Auxiliary transformer at control building 15000,0 155.714,7 0,02 18.064,52 x cars for maintenance team 100000,0 1.038.098,2 0,12 120.429,8



Subtotal 700.000,0 7.266.687,4 0,82 843.008,4 662.363,7 180.644,75 Others


Total 85.148.914,2 883.929.348,9 100,00 102.544.643,8 91.119.574,6 11.425.069,2







10.1.3 Gamesa G58 Investment Costs








OHL 1x1x125mm 2; 18 km (33kV) 225000,0 2.335.721,0 0,27 270.967,0




Subtotal 700.000,0 7.266.687,4 0,84 843.008,4 662.363,7 180.644,75 Others


Total 83.751.414,2 869.421.926,5 100,00 100.861.637,7 89.436.568,5 11.425.069,2



PHASE: Final - FEASIBILITY STUDY




10.1.4 Estimated Enercon E-53 Investment Costs








OHL 1x1x125mm 2; 18 km (33kV) 225000,0 2.335.721,0 0,27 270.967,0




Subtotal 700.000,0 7.266.687,4 0,83 843.008,4 662.363,7 180.644,75 Others


Total 84.288.914,2 875.001.704,3 100,00 101.508.947,7 90.083.878,5 11.425.069,2







10.2 Construction Period

A period of 12 months was assumed for the construction of the Ashegoda Wind Farm. It

was estimated that construction will commence the first of January 2007 latest, implying

that commissioning will be at the End of 2007. The concession period was chosen as

20 years.

10.3 Potential for Local/Regional Input

10.3.1 Grid Connection by EEPCo

For the potential local input to the project realisation we can assume that EEPCo will be

responsible and experienced for the whole grid connection work, internal and external

cabling, of the wind park.

10.3.2 Civil Works

Besides this there are some additional tasks to be done by local staff and Ethiopian com-

panies for all other civil works like:

- road construction,

- internal and external cabling trenches,

- erection of control building at the site and building of the foundation can be done with

local materials and local staff. The work on the substation extension can also be done

by EEPCo staff. Especially the cement factory of Mekelle, located near by, will be able

to provide labor, cement, and trucks for the implementation phase. Additionally, sea-

sonal staff can be also obligated at the City of Mekelle or at EEPCo.

10.3.3 Lattice Towers for Wind Turbines

A possible additional local input could be in an extensive manner the erection of wind tur-

bines using lattice towers instead of tubular steel towers. Some manufacturers of wind

turbines are offering lattice towers as an option besides tubular towers. The production of

lattice towers by manufacturing the needed profiles in Ethiopia could increase the local

input considerably because of the large number of the planned wind turbines at the Ashe-

goda site. To take this option into consideration of the future wind park plannings, it is

necessary to negotiate with the manufacturers of lattice towers directly about their condi-

tions for local production in Ethiopia. Unfortunately, Enercon is not accepting lattice tow-

ers for its wind turbines in general.




10.4 Operation and Maintenance Costs

10.4.1 General Description

The estimates for operation and maintenance (O&M) have been modeled based on the

experience of Lahmeyer International in Due Diligence projects and calculated through a

model developed by our technical engineers to this aim.

The O&M costs include repairs, maintenance, spare parts, insurance costs, personnel

costs for wind park management and maintenance and electricity consumption.

Further, a major overhaul of all equipment has been assumed to take place between the

10th and 11th years of operation in an amount of 5% of total investment costs.

Finally, wind farm decommissioning costs in operation year 21 have been considered in

an amount of 5 % of total investment costs. For the wind parks within this report, an

analysis of the expected O&M costs has been performed. The LI-O&M cost estimation

model takes the governing influence factors of the wind park into account to predict the

expected O&M costs.

This analysis is based on the following assumptions:

- The wind speed conditions can be described using the specific energy yield per sweptrotor area. The higher this value, the higher is the operating time of the turbine whichleads to higher stresses and higher O&M costs.

- Besides the steady wind speed, also the unsteady portions of the wind will increasethe O&M costs. The situation of the wind park regarding this influence can be de-scribed using the turbulence intensity of the site as well as the park efficiency. Thepark efficiency takes into account the distance between the turbines within the park.

- The presence of a condition monitoring system (CMS) will increase the annual O&Mcosts depending on the system, it might be an online or an offline CMS. On the otherhand it is expected that the repair costs can be reduced for some parts since it is pos-sible to detect failures in an early state and reduce the required costs for the repairmeasures.

- The technical concepts of the wind turbines can be different, therefore the type of thewind turbine will have also an influence on the O&M costs which is taken into account.

- The presence of a substation within the wind park leads to higher O&M costs depend-ing on the design of the substation either as HV or MV substation.

- The length of the internal roads has an influence on the maintenance costs for theroads.




- The personnel costs depend on the country where the wind park is placed with regardto the labor productivity and average gross annual earnings in comparison to our ref-erence values in Germany.

- The prediction of the O&M costs is combined with uncertainties (e.g. lifetime of thecomponents). To take these uncertainties into account, for the influence factors uncer-tainties are defined.

- The insurance costs have been estimated as no detailed offer from an insurancecompany is currently available. This has to be specified in cooperation with the turbine manufacturer.

- Based on the experience of the Consultant and compared to other wind park projectsin established markets the P50 value of the energy yield is used as input value for themodel. For the long-term cost estimation the P50 value of the model result is used es-pecially for larger wind parks as planned in this project. In order to get a conservativeestimation, a safety factor has been applied to take into account the specific condi-tions and uncertainties for the operation of the first wind park in Ethiopia. This factoralso includes the uncertainties concerning the future inflation rates. The aim to calcu-late conservatively also leads to the use of the P75 probability value for the calcula-tions.

Within this report the results of the O&M analyses are given in the following categories as

expected average values per wind park:

1) Planned Maintenance

a) Personnel Costs

b) Consumable Costs

2) Unscheduled Repair

a) Personnel Costs

b) Replacement Part Costs




10.4.1.1 Enercon E-48 Maintenance and Repair Costs

See also the following table for the expected cost development of these components dur-

ing the operation period of the park Ashegoda with Enercon E-48 turbines.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2012

2014

2016

2018

2020

2022

2024

2026

2028

2030

Year

[k]

Repairs - personnel50%

Repairs - spare parts50%

Planned Maintenance -pesonnel 50%

Planned Maintenance -consumables 50%

Figure 10-1: Development of the O&M cost for Ashegoda




10.4.1.2 Vestas V52 Maintenance and Repair Costs


ing the operation period of the park Ashegoda with Vestas V52 turbines.


0

200

400

600

800

1000

1200

1400

1600

1800

2000

2012

2014

2016

2018

2020

2022

2024

2026

2028

2030

Year

[k]








10.4.1.3 Gamesa G58 Maintenance and Repair Costs


ing the operation period of the park Ashegoda with Gamesa G58 turbines.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2012

2014

2016

2018

2020

2022

2024

2026

2028

2030

Year

[k]









10.4.1.4 Enercon E-53 Maintenance and Repair Costs

The following table shows the estimated cost for development of these components during

the operation period of the park Ashegoda with Enercon E-53 turbines. Only estimated

costs can be taken into consideration until some turbines of this type will be erected.


10.4.1.5 Comparison of Maintenance and Repair Costs Estimations

The O&M costs for the four different turbines differ by approx. 10% between the Enercon

and the other machines. In case of Enercon the costs are generally smaller since one

main component, the gearbox, is not present. On the other hand the electrical compo-

nents are more expensive, which leads also to higher repair costs for these components.

Furthermore the Enercon machines have a higher specific energy yield at this site which

leads to slightly higher maintenance and repair costs. Overall however, it is estimated that

the Enercon machines have smaller O&M costs, but only slightly.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2012

2014

2016

2018

2020

2022

2024

2026

2028

2030

Year

[k]








10.4.2 High Technical Availability

To reach high technical availability of the turbines we suggest to negotiate with the turbine

manufacturer for sufficient EPC-contract condition. As a major aim the turbine supply con-

tract shall include a five year warranty period instead of the usual two year period.

In case of ENERCON India being the turbine manufacturer it should be possible to agree

on a special warranty agreement similar to the German EPK (Enercon partner concept)

which includes high technical availability with fixed costs in cent/kWh connected to the

energy production in kWh. It is so far not known to LI that other manufacturers offer this

for a first wind park in Ethiopia. They shall be contacted in case another manufacturer is

considered.

Additionally we propose to establish separate service teams of the manufacturer at the

site Ashegoda in order to fulfill high quality of maintenance during the warranty period and

also to shorten the reaction time in case of failures of the wind turbines. It is envisaged

that the manufacturer will be obligated to arrange a constant presence for a minimum of 2

years for training of local operational staff in maintenance and repair measures. According

to Consultant's experience, manufactures have to guarantee that in wind parks with an

approximately installed capacity of 50 MW, two experts (usually, one electrical engineer

and one mechanical engineer) will be constantly present at the wind park. Additionally, a

team of four local experts has to be established for maintenance tasks and a crane has to

be available, at least once per year, to realise operation revisions. Further, condition moni-

toring is realised by independent engineers in order to plan the repairs.

10.4.3 Local and foreign Operation and Maintenance

The concept for operation and maintenance is supposed to create the highest possible

technical availability at reasonable costs. To achieve this goal it is favourable to divide the

O&M activities between the manufacturers staff and local operation staff. The following

table shows the local and foreign O&M interfaces.




Table 10-5: Local and foreign O&M tasks

Task Manufacturer Lokal Staff1. Operation1.1 daily turbine check-up via remote monitoring responsibility execution1.2 staffing of the local monitoring room x x1.3 daily operation of the wind park responsibility2. Service and maintenance2.1 regular visual control of the turbines responsibility execution2.2 permanent service responsibility execution2.3 maintenance of the turbines responsibility execution2.4 infrastructure maintenance execution

Interfaces for operation and maintenance

The displayed interfaces and tasks allow the inclusion of the local operation staff into the

maintenance process. Thus the responsibility for the availability of the turbines remains

with the manufacturer, whereas the responsibility for the stability of the grid is fully in the

hands of the local staff. The O&M activities of the local staff on the turbines are conducted

after the briefing from the manufacturer.

Local storage of spare parts should be limited to expendable items and spare parts. To

maintain the availability on a high level, it is recommended to implement a Conditioning

Monitoring System. Thus the possible failure of components can be identified in advance

and transport and installation of spare parts can be organised ahead of time.

10.4.4 Training

The training measures have to be organised by the manufacturer to train the local experts

by beginning with basic knowledge on wind energy. Currently no well experienced wind

energy experts in the field of wind turbine technology are available in Ethiopia.




10.4.5 Overview Operation and Maintenance Costs

The expected annual operation and maintenance costs for the wind park Ashegoda are

shown in the following tables:

10.4.5.1 Scenario I 86 x Enercon E-48 Turbines

Table 10-6: Annual Operation & Maintenance Expenses

in % of turbine Years 1-20 Years 1-20 Years 1-20prize (Cost in /a) (Cost in Birr/a) (Cost in USD/a)

1. Annual Operating Expenses 2.880.570 29.903.145 3.469.0641.1 Annual cost of maintenance 52,7% 1.586.270 16.467.040 1.910.341

Maintenance of wind turbines 114.219 1.185.705 137.554

Repairing of wind turbines 131.345 1.363.485 158.178Consumables 83.100 862.660 100.077

Spare parts (including rent of crane) 1.257.607 13.055.190 1.514.533

1.2 Wind park management 20,0% 602.000 6.249.351 724.987239.756 2.488.903 288.738362.244 3.760.448 436.250

1.3 Insurance 12,0% 361.200 3.749.611 434.992Insurance of wind turbines, cables and grid connection equipment 139.219 1.445.230 167.661

Insurance of wind park staff 221.981 2.304.381 267.331

1.4 Power demand 4,0% 120.400 1.249.870 144.997Expenses of annual power demand of the turbines 120.400 1.249.870 144.997

1.5 Other Costs 7,0% 210.700 2.187.273 253.746Subscriptions to federations and associations 5.200 53.981 6.262Office costs, materials and others 205.500 2.133.292 247.483

Expenses of international technical assistance

Local technical expenses

Wind farm Development Ethiopia

PHASE: Final Draft - FEASIBILITY STUDY

Annual Operation & Maintenance Expenses of Wind Farm Ashegoda

PROJECT: Wind Farm at 'Ashegoda' site, Ethiopia




10.4.5.2 Scenario II 86 x Vestas V52 Turbines




Maintenance of wind turbines 114.219 1.185.705 137.554Repairing of wind turbines 131.345 1.363.485 158.178Consumables 83.100 862.660 100.077


1.2 Wind park management 20,0% 639.625 6.639.936 770.299239.756 2.488.903 288.738

399.869 4.151.033 481.561

1.3 Insurance 12,0% 383.775 3.983.961 462.179Insurance of wind turbines, cables and grid connection equipment 139.219 1.445.230 167.661Insurance of wind park staff 244.556 2.538.731 294.518







Expenses of international technical assistanceLocal technical expenses




10.4.5.3 Scenario III 86 Gamesa G58 Turbines







399.869 4.151.033 481.561












10.4.5.4 Scenario IV 86 Enercon E-53 Turbines


10.4.5.5 Comparison of different Scenarios

The different scenarios are based on the estimated M&R costs as described above in

chapter 10.4.1.5 in more detail. The reasons for the differences are also described there.

Further aspects are treated in the same way in all scenarios with constant percentages of

the turbine price. This is a common estimation for these values. The quality of the mainte-

nance is assumed for all scenarios to be carried out according to good industry practice.

Therefore we do not expect significant differences at this stage. The achievement of this

quality should be secured by competent review of the works. To support EEPCo in this

field, the expenses for international technical assistance are included in point 1.2 of the

tables.






362.244 3.760.448 436.250












11 Capacity Credit

11.1 Methodology and assumptions

11.1.1 Definition of the Capacity Credit

Any kind of electrical power plant is required to answer the crucial question: To what ex-

tent is a specific energy production unit with an installed nameplate capacity available to

meet system demand?

Typically the answer is derived from a statistical analysis of this period, when a unit or

technology is not available, even if it is supposed to be working, generally termed as

forced outage. For conventional generation units this is mainly a technical question,

whereas for intermittent resources like wind, solar and hydro power, it is a question of

fuel availability: is wind, water or solar radiation available when it is needed? And if yes,

what amount is available and how does the intermittency influence the rest of the electri-

cal supply system? The answer is given by the capacity credit, which should consider

each of the posed questions.

There are various definitions of the term capacity credit (CC), and several synonyms are

used in parallel, each bearing a similar variety of definitions. The two most common syno-

nyms of CC are "firm capacity" and "capacity value".

In this feasibility study, the CC is defined as the firm capacity, which can be replaced by a

certain amount of installed wind power or any other energy source. It can be used either

as a value in MW or as a percentage of the installed wind capacity. This is the terminology

used in recent studies throughout industrialised countries, where wind energy is replacing

other forms of energy. In Ethiopia, a country rapidly developing, and with high goals for

improving the electrification rate, it is important to analyse how much capacity wind energy

can add to the current system.

Further, the influence of wind power on the Interconnected System is measured through

three different methods:

Changes on the Load Duration Curve (LDC) of the Interconnected System (ICS),

Load Following and

Spatial Smoothing Effect.




The LDC reflects how the seasonal and daily wind distribution matches to the load pattern

of the ICS. This is especially important to get an idea of the amount and capacity category

(base, intermediate or peak capacity) that have to be installed within a system with a de-

mand represented by a certain LDC. When comparing the LDC and a LDC reduced by the

wind energy distributed to the system, an illustration of the wind influences on the system

is obtained, showing when peak capacity, base load capacity is needed or when wind has

to be dumped. Wind energy is highly required at peak load hours being the CC 68 MW.

Load following indicates whether an additional effort has to be considered regarding

power balancing when introducing wind energy to the Ethiopian system. The calculations

have shown, that no further effort has to be realised.

Finally, the Spatial Smoothing Effect, which is the effect that numerous wind parks at dif-

ferent sites reduce the fluctuation of wind energy distributed to the grid, has been dealt.

Fluctuation of wind power output is reduced the more dispersed the installed capacity.

Because of this effect, the installation of more than one wind park is recommended.

11.1.2 Methodology Overview

The Capacity Credit analysis for wind power allows an evaluation of the effects of wind

power on the energy system. In the considered case of Ethiopia, it is analysed how two

wind parks located at the Harena-Mesobo and Ashegoda sites, influence the supply in the

Interconnected System (ICS).

The influence of wind power in the ICS is strongly dependent on the size of the wind

parks. Therefore, four Scenarios with different nominal capacities have been considered

in the Capacity Credit assessment. The Scenarios have been defined as:

Scenario I - the sole installation of the Mesobo-Harena Wind Park (48 MW),

Scenario II- the sole installation of the Ashegoda Wind Park (68.8 MW),

Scenario III- the simultaneous installation of the two wind parks (totalling116.8 MW) and

Scenario IV - the installation of two large wind parks at the sites of Mesobo-Harenaand Ashegoda (2 x 180 MW).




Two approaches were used to calculate the Capacity Credit, which are described in the

following:

Year-Round Capacity Credit Approach,

Peak Load Capacity Credit Approach.

Year-Round Capacity Credit Calculation Approach

The Year-Round Capacity Credit approach is based on the hypothesis, that wind energy

is distributed to the system, assuming constant power demand. An hydropower system

with seasonal storage capacities and a variation of different units can be assumed as an

ideal storage. The water is not actively stored, like in a pumped hydro power plant, where

losses of 15 % and more have to be taken into consideration7, but the wind is a water

saver, reducing the actual usage of water. In this case the efficiency loss can be dimin-

ished, since the hydro system, run at the point of optimal efficiency, will be readjusted to a

new point of optimal efficiency. The total efficiency will be the same, if the system is run

properly, even though it might be necessary to include other units and store wind energy

at various reservoirs, in order to have all units running at the point of best efficiency8.

The Year-Round Capacity Credit calculation for Ethiopia is based on the prediction of

wind power production, which can be expected to be delivered to the grid according to the

restrictions described later. This wind dispatch is compared to an equivalent of water

saved in Finchaa reservoir, which was agreed with EEPCo to be considered as the refer-

ence Hydro Power Plant (HPP). This leads to the energy based definition of the Capacity

Credit. This approach is only valid in energy systems with a high portion of hydro power

and seasonal storage potentials, where stored water energy can be distributed at any

time. Hydro power allows to balance fluctuating wind energy with little efficiency losses.

Especially when hydro power is used to provide base load and peak load, wind power can

be perfectly included into the dispatch order9.

7 Lönker, O. (2005), Zukunfsspeicher . In: Neue-Energie-das Magazin für erneuerbare Energien, Heft 4, 2005. Pag. 26-33.8 Matevosyan, J. (2004), Wind power in areas with limited export capability . Internet source.http://www.lib.kth.se/Lfulltext/2004matevoszan.pdf, last update: 2004, last access: April 26th, 2006, 15:00hours.9 An example for the interaction of wind and hydro power is the shaping service of the BonnevillePower Administration (BPA)9 active in the north west of the USA. BPA uses its hydro resources toshape wind power distributed to the system and then sells the energy as it is needed. (Source:www.bpa.gov )




Peak Load Capacity Credit Calculation Approach

Whereas the Year - Round Capacity Credit approach is annual energy driven, the Peak-

Load Capacity Credit calculation is based on power supplied in specific hours. The Peak

Load Capacity Credit approach assesses different fractions of the highest load hours for

their individual capacity factor, which has been proven in recent studies by the US Na-

tional Renewable Energy Laboratory (NREL) to be a good indicator for the CC during

times of high demand. The background of this approach is that during the hours of highest

demand the system has the highest loss of load probability. This is not necessarily the

case in countries with high reserve capacities, which can be activated for the times of

highest demand. In developing countries like Ethiopia, where capacity is still expanding, it

is important to find to what extent wind energy can add to the available system capacity

during these hours.

The method screens all 8,760 hours of the year and orders them by demand. For every

hour the corresponding wind can be analysed. Following the approach of Milligan10, differ-

ent fractions of the highest load hours are chosen to analyse the capacity factor of these

fractions. Further, a reliability calculation has been conducted for the 1 % of the highest

load hours, representing a load decrease of roughly 100 MW.

Both approaches are based on average hourly wind distribution provided by

EEPCo, measured for one year from February 2nd, 2005 until February 2nd, 2006

at the site. This wind distribution is compared to the hourly load of the ICS for the same

period of time 11. It is modelled, how the load duration curve (LDC) changes when wind

energy is introduced to the ICS and how the seasonal and daily wind distribution matches

the load pattern of the ICS. (LDC are defined in Section 11.1.1).

10 Milligan, M. (2002); Modeling utility-scale wind power plants Part 2: Capacity Credit . Internet source:http://www.nrel.gov/docs/fy02osti/29701.pdf, last access: June 2nd, 2006, 21:05 hours. Pag. 15-19.11 The provided ICS load data is from January 2005 until December 2005. The Data for January 2005 is usedfor January 2006 plus 11,7 % for every hour (average difference between January 2005 and January 2006given in the monthly reports provided by EEPCo)




11.1.3 Assumptions

Calculations are based in the following assumptions:

A hydro base load of 100 MW is assumed as must run capacity at all times, tostabilise frequency and voltage. This leads to a partial shut down of wind turbinesin hours of very low demand.

The average inflow data of Finchaa reservoir during 1960 and 1997 was used forthe calculation. The Finchaa HPP will be one of the most influenced HPP by theinstallation of wind power in Ethiopia, since it is one of the large important balanc-ing power plants. Finchaa HPP has a gross head of 596 m and an installed capac-ity of 134 MW divided into 4 evenly sized Pelton turbines. The live storage of thereservoir is 790 MM³, which equals 976 GWh, if the turbines constantly operate atrated flow. The use of Finchaa as the reference power plant, is justified due to itsimportant balancing effect. Another reason is the relative large size of the installedcapacity, which is able to balance a 100 MW wind park.

The conversion from wind power to stored hydro power is based on the conversionvalue 809 m3/MWh12, used by EEPCo in the monthly reports to calculate the dis-charge of Finchaa reservoir.

Data scarcity: the results have been evaluated for the measured period providedby EEPCo to reflect a realistic simulation. Nevertheless, extrapolations should bebased on long term measurements for wind and power loads.

The load data provided by EEPCo consists of daily data sets with load measure-ments for every hour and in parts with inter-hour measurements. The data wastransferred into a consistent time line of hourly increments using averages for thehours if more than one measurement was available. Improbable low data pointsbelow 100 MW were increased to 100 MW of minimum load. Since the load dataset did not match the wind measurements data, the load data for January 2005was increased by 11.7 % to use it as a projection for January 2006. The value11.7 % is derived from an aggregated spreadsheet provided by EEPCo and isequivalent to the increase from January 2005 to January 2006 projected for Janu-ary 2006 in December 2005. Since there was no information available regardingsuppressed demand (load shedding), neither the occurrence during the evaluatedperiod nor its amount, it has not further been considered.

12 The value for m3/MWs has been changed from the nameplate value of 0,209m3/MWs to the value used byEEPCo of 0,225m3/MWs. This leads to shifts in the following values in this Section. It has to be noted, thatdifferent Excel tables provided by EEPCo use different values for this calculation. The value used now, is themost conservative, meaning the highest water consumption per unit.




11.1.4 Load Duration Curve Calculations

Based on the load data provided by EEPCo for the year 2005 the load duration curves

(LDC) were calculated.

LDCs are constructed from a sample of observed load values measured in MW, normally

the hourly load data of one year. The load values are re-arranged in descending order

over the 8,760 hours of the year. By normalising the horizontal axis to range from 0 to 1, it

is possible to read off the probability of exceeding a certain level of load13. Given the

hourly load is a random variable L(t), the duration on the horizontal axis equals the prob-

ability for L(t) being greater or equal to the load on the vertical axis (See Figure 11-1). This

allows the determination of generation capacities with different characteristics and cost

structures. The area below the LDC represents the energy produced during the whole

year. By re-arranging the hourly load data, any information about seasonal or daily varia-

tion is lost; what counts is the distribution over the year. It indicates the duration of the

year for which a certain level of capacity has to be available to meet the demand. This is

especially important for getting a general idea of the amount and capacity category (base,

intermediate or peak capacity) that have to be installed within a system with a demand

represented by a certain LDC.

Residual load duration curves (RLDC) are LDC reduced by the hourly specific energy,

which is dispatched first within the power system. This is generally low cost, must run ca-

pacity, especially renewable energies such as hydro and wind power, but can just as well

be nuclear or coal base load14. This procedure allows creating a second LDC, e.g. by re-

ducing the LDC by the distributed wind energy. The resulting RLDC illustrates how wind

influences the system and, for example, increases the need for peak capacity. Also, it

indicates the extent to which the need for base load is reduced. To conduct a residual

load duration curve calculation, the measured (or predicted) load of each specific hour is

reduced by the measured (or predicted) energy, which is dispatched first. Formally the

residual load duration curve is represented by the following formula:

)()()( tPtLtR W

where R(t) is the residual load at a certain hour, L(t) is the total load and PW(t) is the wind

power output available at the corresponding hour. Again the hourly values have to be or-

dered to a monotonically decreasing time series to accomplish the RLDC. The area be-

tween LDC and RLDC represents the energy delivered by the reduction technology. If the

power output of the technology in question is higher than the load at certain times, the

13 Stoft, S. (2002): Power System Economics Designing Markets for Electricity . New York. Pag. 41-42.14 Billinton, R., (1996), Reliability Evaluation of Power Systems , 2nd Edition, Plenum Press, New York. Pag.73




RLDC becomes negative on the right end of the graph. The area below zero corresponds

to the energy, which cannot be used in the system and has to be either dumped or, if pos-

sible, sold to a connected system.

After calculating LDC and RLDC by the described method and depicting the two curves in

one graph as done in Figure 11-1 the capacity credit can be derived graphically. It is

specified by as the vertical-axis intercepts of the LDC and the RLDC.

Load Duration Curves

0

100

200

300

400

500

600

0 730 1460 2190 2920 3650 4380 5110 5840 6570 7300 8030 8760

Hours of one Year0 25% 50% 75% 100%

Duration (t)

Load (L(t))

Capacity Credit

Lo

ad[M

W]

Load and Wind 100 highest hours

450

500

550

600

0 10 20 30 40 50

Load

[MW

]

Capacity Credit

LDC

RLDC

Hours / Probability

Hours

Load Duration Curves

0

100

200

300

400

500

600

0 730 1460 2190 2920 3650 4380 5110 5840 6570 7300 8030 8760

Hours of one Year0 25% 50% 75% 100%

Duration (t)

Load (L(t))

Capacity Credit

Lo

ad[M

W]

Load and Wind 100 highest hours

450

500

550

600

0 10 20 30 40 50

Load

[MW

]

Capacity Credit

LDC

RLDC

Hours / Probability

Hours

Figure 11-1: Load duration curve and residual load duration curve for Ethiopia,

The inset shows the 50 highest load hours (own illustration)

The obvious advantage of the RLDC approach is its relatively low need for data and its

fast application. On the other hand, it does not give information about the risk of loss of

load during a specific hour or period of time, and leaves the question of necessary reserve

capacity unanswered. Nevertheless it illustrates how the system is influenced by the addi-

tion of a new resource.




11.1.4.1 Load Restrictions

Within the power systems of developing countries the demand during night hours often is

so low that the wind energy produced cannot be distributed to the grid completely. In

Ethiopia the lowest average demand of 154 MW is recorded between three and four

o clock a.m., with particular hours as low as 90 MW. To assure stable frequency and volt-

age a certain amount of conventional capacity (in Ethiopia hydro power) has to be running

at all times. For this reason wind energy should be totally or partially turned down at times

with too little demand. For the calculation of the must run restriction the following formula

has been applied to every hour. The distributed wind energy PWi is given with

WpiiiWi PLRLforLRLP 0,

where PWpi is the potential wind power, Li the load, and LR the load restriction, defined as

the minimum of conventional production first dispatched to the system. For all other hours,

the wind energy produced can be fully included to the system PWpi = PWi.

11.1.4.2 Load Duration Curve Analysis & Results

The load duration curve analysis is applied in two directions: On the one hand calculating

the residual load duration curve resulting in a curve below the original LDC and on the

other hand calculating the increased load duration curve (ILDC) being above the LDC.

The first calculation corresponds to the case depicted in Figure 11-2. The effective wind

output is distributed to the system while the hydro power plants work less and water is

stored.

The ILDC represents the load which could be met with the same distribution of hydro en-

ergy and the addition of the potential wind energy. Figure 11-2 shows the three load dura-

tion curves and two step graphs. The two step graphs correspond to the LDC and RLDC

representing the activity of plants other than Finchaa HPP. The hypothesis applied to the

step graphs are that 60 MW of the installed capacity is working as base load and the re-

maining 70 MW are working as balancing capacity.

This results in a maximum distance between step graph and LDC of 130 MW and a mini-

mum of 60 MW always occurring at the tip of each step. The same step graph is valid for

the ILDC as for the original LDC since the distribution of hydro power does not change, all

energy is distributed.




0

100

200

300

400

500

600

700

0 730 1460 2190 2920 3650 4380 5110 5840 6570 7300 8030 8760

Hours of one Year

Lo

ad[M

W]

Increased Load

LDC

RLDC

Base Load

Reduced Base Load

Dependable Capacity

Max Load

Max Residual Load

Figure 11-2: LDCs and power output of other plants,

dependable capacity, max. load, max. residual load (own illustration)

The change in vertical distance between the LDC and the two derived curves has to be

pointed out. During the high load hours the difference between LDC and ILDC is larger

near the vertical axis and decreases to the right, whereas the difference between LDC

and RLDC reaches its maximum towards the middle. As Figure 11-2 does not visualise

this very well, the following table shows the results for different fractions of the year in

numbers.

Table 11-1: Difference between LDC, ILDC and RLDC

Percentile of LDC ILDC - LDC LDC - RLDC

1% 68 445% 56 42

10% 51 42

20% 49 39

40% 45 38

60% 41 38

80% 39 36

100% 35 34

The difference between the two calculated curves in relation to the original LDC is due to

the method applied in this case: For the RLDC load hours corresponding with high wind

output move to the right of the curve, while for the ILDC hours of high wind output move to

the left. The result already indicates that during high load hours there is also a high wind

output available. The vertical intercept difference of LDC and ILDC is 74 MW, whereas the

average of 1 % of the peak hours is still 68 MW (see Table 11-1). It should be pointed out,




that the capacity factor of potential and effective wind energy output is reached at 100 %

in both columns.

The capacity credit for the enlarged system represented by the ILDC is 68 MW and for the

reduced system, reducing water usage, the capacity credit is 44 MW. Both values are

depicted in Table 11-1 in bold. However this value indicates a peak load capacity credit

and is not to be confused with the annual energy yield to be expected.

The reserve capacity for Ethiopia before the installation of wind is calculated using the

715 MW dependable conventional capacity and the maximum load of 586 MW. It results

in 129 MW of reserve capacity.

With regards to the expanding Ethiopian system and the large hydro capacities available,

ILDC is more representative for the situation of Ethiopia. It gives a good picture on how

wind energy will increase the system capacity.

11.1.5 Load Following

Load following, defined as the necessity to change the system power output within a cer-

tain time period15, has been calculated. The Load difference between every two consecu-

tive hours has been analysed and introduced into a frequency distribution table. The same

measure was taken, in the case of wind added to the system to increase the capacity and

for the case where wind was deducted from the load to save water.

The needed load following requirements of the capacity used can be expressed by the

ramp rate between two consecutive hours expressed by the formula:

1iii LLR

Where L is the hourly load at hour i, and R is the load following requirement in the hour i.

For adding wind to the system, the load following requirements are defined as

)()( 11 WiiWiiWi PLPLR

Where RWi is the load following requirement in the hour i for the system with wind, and PW

is the wind production at hour i, making the terms in parentheses the load net of wind in

the respective hour.

The load changes are investigated for three different occasions:

15 In Ethiopia this is hydropower.




the measured load,

the residual load, with the distributed wind energy being deducted from the load,

the increased load, with a hypothetical load being calculated, which could havebeen met during the analysed period, if wind had been added.

For each of the three cases a separate frequency distribution table is created stating the

frequency of the load changes in classes with increments of 25 MW. To visualise the dis-

tribution, the load changes are ordered and displayed in Figure 11-2.

The analysis of the change of load, and the change of residual load during two hourly val-

ues has shown, that wind power does not significantly increase the changes which have

to be followed. Figure 11-3 shows the ordered load changes of the measured load, the

residual load, the increased load and the change of distributed wind energy. This indi-

cates that no further effort has to be considered regarding the balancing activities of the

power plants.

Figure 11-3: Load following

The two load curves are so close together, that a difference can be hardly seen. To show

the changes that do occur in Table 11-2 gives the number of load changes within a certain

class, representing 25 MW increments. It is differentiated in negative and positive load

changes.




It has to be mentioned, that the load change between two hours with added wind (in-

creased load duration curve) are hypothetical, since the actual demand pattern has to be

followed which would change the dispatch of the hydro power plants and from the result-

ing fewer hours of higher load, changes in the column of the increased load are not likely

to occur.

Table 11-2: Frequency of load changes within one hour with and without wind

Class Load ResidualLoad Increased Load

-275 0 0 0-250 0 1 0-225 2 2 2-200 4 6 1-175 16 33 24-150 85 75 74-125 88 104 94-100 146 135 155

-75 186 205 217-50 329 366 425-25 952 970 1138

0 2578 2504 231325 2419 2209 209050 877 1067 99775 615 621 645

100 312 311 354125 87 93 127150 37 34 62175 11 11 25200 10 7 10225 3 3 2250 1 1 3275 1 1 1

The second important issue to be regarded when assessing the influence of wind power

on the system and especially the reserve capacity is the spatial smoothing effect.

11.1.6 Spatial Smoothing Effect

The spatial smoothing effect is the effect, that numerous wind parks at different sites re-

duce the fluctuation of the wind energy distributed to the grid. It influences the different

capacity credit fractions and with these the balancing activities, which result from wind

power. This topic is not part of the scope of this feasibility study and thus it has not been

dealt with in detail. Nevertheless, fluctuation of the wind power output is reduced, the

more dispersed the installed wind capacity. Because of this effect, the installation of more




than one wind park is recommended. The effect increases with greater dispersion of the

sites. For Ashegoda and Harena-Mesobo the impact is small since the sites are located

very close to each other.

11.2 Year-Round Capacity Credit Approach

As described previously, the Year Round Capacity Credit is based on the special condi-

tions in Ethiopia, where the wind energy can either be saved in the water reservoirs to

increase the seasonal storage or to be supplied to the grid directly and thus, increase the

capacity available. Between these two alternatives, any variation can be chosen to in-

crease energy supply and security. In the next section, it has been assumed that all wind

energy is saved throughout the year and thus, the storage level of the reference reservoir

Finchaa is increased significantly.

Figure 11-4 shows the influences on the stored energy at Finchaa HPP assuming a half

full live storage at the beginning of the period, when all wind energy is saved. As base

load energy, 100 MW hydropower has to be dispatched first into the grid. Then wind en-

ergy is dispatched, before hydropower covers the residual demand. This is important,

since hydropower is the time independent energy source compared to the fluctuating na-

ture of wind distribution. This dispatch strategy allows to preserve the unused water in the

reservoirs and the seasonal storage is conserved.

In Figure 11-4 the area between the two curves represents the effective energy, which

can be stored. It is used for the calculation of the capacity credit. In case of a full reservoir

the described restrictions apply, and the top curve would be in part flattened, as long as

no storage is available, which would reduce the effective energy yield and with it, the ca-

pacity credit.




Figure 11-4 Water reservoir levels with and without wind energy

The lower curve shows the development of the stored water without the installation of

wind power during one year based on the assumptions for Finchaa HPP. The top curve

assumes the installation of both wind parks. Evaporation can be neglected in this as-

sessment, regarding the fact that evaporation does not increase significantly, if more wa-

ter is stored in the reservoir.

Table 11-3 summarises the detailed numbers of saved water for the installation of the two

wind parks considering a water consumption of 752 m³/MWh (0.21 m³/MWs)16 at full load.

Table 11-3: Stored energy at Finchaa reservoir

Stored energy at the end of one year at Finchaa reservoir

without

windwith Harena with Ashegoda

with Harena and Ashe-

goda

MWh 281,901 379,307 513,707 784,326

MCM17 212 285 386 589

16 Source: EEPCo & ACRES (Ed.) (2001): The Ethiopian Energy System . Unpublished report.17 Million Cubic Meter




The assumption of storing all the distributed energy allows to quantify the influence of

wind energy on the hydropower system. In practice, EEPCo will regulate the use wind and

hydropower together. Figure 11-5 depicts how the available capacity is increased, if the

water saved by wind energy is dispatched continuously.

Average Load + energy based Capacity Credit

0,0

100,0

200,0

300,0

400,0

500,0

600,0

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

Hours

Lo

ad[M

W]

Load Evenly distributed Wind

Figure 11-5: Average load and energy based CC

With this approach the Year-Round Capacity Credit is 36.5 MW.

11.3 Peak Load Capacity Credit

After the previous analysis of the wind and hydropower distribution throughout the year, it

is reasonable to calculate a Peak Load Capacity Credit for the dry season only. During the

rainy season enough hydropower capacity is available to meet the load, as water avail-

ability is the restricting parameter for the HPP.

11.3.1 Seasonal Wind and Water Distribution

The plans to interconnect the ICS to the grid of neighbouring countries and to use the

large potential of hydro power to sell energy into adjacent systems require the ability to

meet system demand at all times. Because of the fluctuating water conditions in Ethiopia




the installation of wind power will increase the security of supply of the system. The com-

bination of wind and water on a seasonal basis is very good. As it can be seen in the fol-

lowing figure, during the dry season the wind energy distribution is high, while during the

rainy season the wind energy distribution drops significantly compared to the highest dis-

tribution in December.

Water Inflow and Wind Energy Potentialexemplarily for Harena

0

20

40

60

80

100

120

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

Wat

er[M

Wh

]

0

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000

18.000

Win

d[M

Wh

]Average Water Inflow Wind Energy Potential

Figure 11-6: Seasonal wind and water distribution

This seasonal difference has the effect, that competition between the two renewable en-

ergy sources, wind and water, hardly occurs. Only in hours of very low demand and high

wind speeds the potential wind energy has to be turned down to meet the 100 MW hydro

base load restriction.

11.3.2 Daily wind power distribution

The special climatic and topographic situation in Ethiopia leads especially during the dry

season to a characteristic wind distribution over the day. The wind speeds start increasing

in the early morning to reach their peak in the early evening and then to drop slowly to its

lowest point in the morning again. A similar distribution over the day can be observed with

the load.




Yearly Average of hourly Load and Wind Pattern

0

100

200

300

400

500

600

700

0 6 12 18

Hours

Lo

ad[M

W]

0

10

20

30

40

50

60

70

Win

dP

ow

er[M

W]

Load Pattern Wind Distribution

Figure 11-7: Daily load and wind distribution match

Figure 5-4 shows the base case with the two wind sites at Harena-Mesobo and Ashegoda

together. Because of the strong diurnal characteristics of the average wind speeds, the

match of the wind energy output and the load, significantly complement each other.

The Correlation Coefficient of load and wind distribution is 0.28 and increases if only the

dry season is analysed. To have a comparison, just to indicate that in studies conducted

in the US the correlation coefficient of load and wind distribution was zero.

11.3.3 Peak Load Capacity Credit Calculation

The capacity credit calculated here is not the firm capacity, if the firm capacity is defined

as the capacity which is available at all times. During the rainy season the wind distribu-

tion is very low (below 5 m/s) and also during the dry season, there are days when wind

energy output drops under the capacity credit. Thus, for these periods, balancing power

has to be supplied by HPPs or DPPs.

To calculate the peak load capacity credit a special measure is applied concentrating only

on the dry season. In this analysis, peak hours are not certain hours of the day, but the

top hours from the ordered load duration curve. In Ethiopia the capacity shortages are

primarily apparent there, when water in the reservoirs is getting scarce.




The following Table displays the frequency of supplied wind power during time of highest

demand, differentiated into the 10 %, 20 % and 30 % of highest load and classes of power

being distributed.

Table 11-4: Frequency distribution of wind power during hours of highest load

Capacity of the WP

For 30%

of highest Load

For 20%

of highest Load

for 10 %

of highest Load

MW DrySeason

RainySeason

DrySeason

RainySeason

DrySeason

RainySeason

10 15.8% 59.4% 13.2% 59.6% 6.4% 57.7%

20 4.9% 7.8% 4.1% 7.4% 2.4% 7.2%

30 7.0% 6.1% 6.2% 5.5% 5.8% 5.8%

40 8.0% 6.9% 7.0% 6.5% 4.8% 6.3%

50 7.7% 6.1% 8.0% 6.7% 9.1% 4.8%

60 9.4% 6.3% 8.7% 6.5% 9.1% 5.8%

70 3.9% 1.8% 3.5% 2.0% 3.4% 3.4%

80 6.3% 2.9% 6.5% 3.7% 8.4% 5.8%

90 16.7% 2.3% 19.8% 2.0% 23.1% 2.9%

100 8.5% 0.2% 9.6% 0.2% 10.0% 0.5%

110 11.7% 0.0% 13.6% 0.0% 17.4% 0.0%

The variance of the demand during the top 30 % of the hours is 277 MW, whereas for the

20 % of the hours it is 254 MW and for the highest 10 % of the hours it is 197 MW (start-

ing with the highest load during 2005 of 586 MW).Table 11-4 shows how the frequency of

low wind distribution decreases when approaching the highest loads of the year. This

supports the conclusion drawn earlier from Figure 11-7, i.e. the higher the demand the

higher the wind distribution.

Table 11-5 shows the percentage of hours when the available wind power drops under the

Year-Round Capacity Credit.




Table 11-5: Frequency of hours with available wind below the Year-Round CC

Frequency of hours with available wind below CC

highest load hours Dry season Rainy season

30% 35.8% 80.3%

20% 30.4% 78.9%

10% 19.5% 76.9%

5% 15.3% 75.5%

During 5 % of the highest demand hours, only for 4.9 % of the time does the available

capacity drop under the capacity credit. The following graph illustrates the distribution of

the CC in correspondence to different load levels. 100% equals the highest load of

586 MW. The values given along the curve are the number of highest load hours for which

the capacity factor has been calculated.

35%

40%

45%

50%

55%

60%

65%

40% 50% 60% 70% 80% 90% 100%

Percent of maximum Load

Cap

acit

yC

red

it

CC at different Load Levels

9 h87 h

219 h

438 h876 h

1314 h

Figure 11-8: CC at different Load Levels




11.4 Results: Capacity Credit

Year-Round Capacity Credit Approach

Using Finchaa HPP station as a reference and as it was calculated in the merit order dis-

patch section, the additional saved water allows a permanent increase of available capac-

ity of 36.5 MW.

Peak Load Capacity Credit Approach

When the correlation between wind and demand distribution is considered, the average

capacity factor of the highest load hours has to be evaluated. By using different fractions

of the highest load hours, it is analysed to what extent wind energy matches the demand

and is available during peak demand hours.

Table 11-2 shows the Capacity Credit results for Mesobo - Harena and Ashegoda and the

two parks together. Additionally, a hypothetical scenario with 2x180 MW wind park sizes

with similar wind behaviour than Mesobo-Harena and Ashegoda has been calculated and

presented in the following Table.




Table 11-6: Capacity Credit Results

Capacity Credit Comparison

ResultsMesobo-

HarenaAshegoda

Mesobo-

Harena &

Ashegoda

360 MW

Scenario

Installed capacity [MW] 48.00 68.8 116.8 2 x 180

Year-Round CC Aproach

Capacity Credit [MW] 11.12 26.46 36.51 83.90

Capacity Credit [%] 23 % 38 % 31.69 % 23.31 %

Peak-Load CC Approach

CC for the 30%

highest load hours30.0 % 47.2 % 40.3 % 36.7 %

CC for the 20%


CC for the 10%


CC for the 5%


CC for the 1%

highest load hours [%]47.1 % 68.4 % 59.8 % 57.8 %

Generated wind energy

[MWh]18 97,406 231,806 319,831 996,510

Rejected wind energy

[MWh/a]461 3,969 13,811 261,560

Rejected wind 0.47 % 1.68 % 4.14 % 26.25 %

18 Does not correspond with the windpro results, because of the explained assumptions




11.5 Conclusion: Capacity Credit

Comparatively high Capacity Credit (CC) values for wind energy have been obtained for

all the sites. The Mesobo-Harena Wind Park is expected to have a CC of 23 %, whereas

the CC in the Ashegoda Wind Park is even higher at 38 %. Further, the CC of both wind

parks has been calculated at 31.69 %. Finally, a hypothetical scenario with the installation

of 180 MW at the Mesobo-Harena site and 180 MW at the Ashegoda site has been calcu-

lated resulting in a CC of 26.25 %.

The CC has also been calculated for the 30 %, 20 %, 10 %, 5 % and 1 % of the highest

load hours. When calculated with the 1 % of the highest load hours, the CC reaches val-

ues of 47.1 % and 68.4 % for the Mesobo-Harena and Ashegoda sites, respectively.

The results in Table 11-2 also show that assuming constant load and an isolated grid,

hydro power and wind power start to compete with each other with increasing installed

capacity of wind energy and an increasing amount of wind energy is rejected starting from

0.47 % when only considering Mesobo-Harena only, and up to 4.14 % when considering

both wind parks (Mesobo-Harena with 48 MW and Ashegoda with 68.8 MW). When con-

sidering two wind parks of 180 MW each, located at similar sites as the Mesobo-Harena

and Ashegoda sites, the amount of wind energy rejected because of the load restriction is

as high as 26.25 %. As soon as energy can be exported to neighbouring countries or de-

mand increases, the amount of rejected wind decreases again or is zero.

Among all the results obtained, the wind park planned at Ashegoda presents the highest

capacity factor values and also the highest output during peak load hours. The implemen-

tation of both wind parks would have the following advantages:

Wind energy contributes to improve the security of supply in Ethiopia. Consideringthat Ethiopia suffers on constant low water dam levels, which reduces the ability touse the installed dominating hydropower capacities, specially during the peak de-mand hours of the dry season, wind would add available capacity.

The positive correlation between load and demand occurs precisely during the dryseason, whereas during the rainy season the peak load Capacity Credit cannot bereached in the majority of the hours. So that wind and hydropower complementeach other.

The grid expansion plan to interconnect the ICS to the grid of neighbouring coun-tries and a growing demand as indicated in the PSEMP to raise the electrificationratio will be dependent on a secured energy supply. Wind energy reduces the in-fluence of low water levels caused by drought in years of low rainfalls. The com-paratively high Capacity Credit for wind power in Ethiopia is a reliable alternative to support the hydro power system.Moreover, hydropower has lower efficiency




losses, when running in partial load, which makes it the perfect shaping capacityfor wind power19.

Another advantage is that unused hydro power capacity can be fully accounted foras reserve capacity20. (Obviously, this is only the case, if the water is stored toproduce energy at other times).

19 Leonhard, W.; Grobe, M. (2004): Realistisches Langzeitkonzept oder Utopie? Nachhaltige elektrischeEnergieversorgung mit Windenergie, Biomasse and Pumpspeicher . In: EW Energiewirtschaftliche Tages-fragen, Jg. 103, Heft 5 2004, pages 26 31.20 Stoll, H.G. (1989): Least-Cost Electric Utility Planning. New York, 1989.




12 Economic Analysis

12.1 Methodology & Main Assumptions

Following the World Bank Handbook for Economic Analysis of Investment Operations 21,

the main purpose of an economic analysis is to help to design and select projects that

contribute to the welfare of a country. Whereas the financial analysis evaluates the project

from the point of view of the operating company or Independent Power Producer (IPP),

the economic analysis evaluates the project from the point of view of the whole economy

of the country.

The purpose of the investigation is to compare from a macroeconomic standpoint the

benefits of the project with the costs it incurs, as is customary in any cost-benefit analysis.

The standard of evaluation for costs and benefits is a monetary quantification. To the

greatest possible extent, the project impacts are evaluated in terms of economic market

prices. Shadow prices are employed, i.e., internal accounting prices that free the day-to-

day (market) prices from multifarious biases. In other words, shadow prices represent an

attempt to illuminate the actual costs of a product or service for the economy as a whole.

In comparison with micro- and macro-economic prices, shadow prices are devoid of taxes

and charges, duties and subsidies.

The calculations are made for the years 2007 through 2028, i.e., the year in which the

wind turbines are expected to be commissioned plus 20 years of operation.

As in the technical analysis, four Scenarios have been considered in the economic analy-

sis of the Ashegoda Wind Park:

Scenario I - 86 Enercon E48 turbines

Scenario II 86 Enercon E53 turbines

Scenario III 86 Vestas V52 turbines

Scenario IV 86 Gamesa G58 turbines

The economic analysis is conducted in the form of equalising the value of getting wind

farms introduced to the power system to the induced savings in the power system in terms

of avoided costs of thermal power generation. The incremental economic cost of the wind

21 Handbook for Economic Analysis of Investment Operations , World Bank, May 1996, p. 169




farm output is defined as the difference between the economic costs of the wind farm and

the avoided economic cost (economic benefits) of a diesel power plant.

The economic benefits or the costs of power generation with a diesel power plant com-

prise:

capital costs of the plant;

fuel costs;

variable and fixed operating costs;

external costs of diesel power generation.

As per recommendation of the GTZ TOR s (Item 1.4 Estimation of Benefits, footnote

no. 2), the Consultant based the technical data and derived calculations for the capital

costs of the diesel power plant on one existing 40 MW DPP located in the northern part of

Ethiopia. The financial data was adjusted where ever applicable to arrive at the estimated

current market value of the existing plant.

The economic costs for power generation with the alternative project (wind power) are

accounted for:

capital costs of the wind park,

fixed operating costs of the wind power installation,

external costs of wind power generation - leakage costs-.

From an economic point of view, the project is profitable, if during the period of time in

question the cost of generating electricity with the wind park is lower than the cost of gen-

erating electricity with the diesel power plant. In other words: the costs incurred for build-

ing and operating the discussed wind park must be lower than the utility value, or eco-

nomic benefits, which it provides. The economic benefits are measured here in terms of

avoided costs (savings). If the wind park is built, the operating costs, and the external

costs, of diesel-based power generation will be avoided.

Pertinent to the cost categories for the diesel systems, differentiation can be made for the

following economising effects:

Capital effects: These account for savings on capital costs and fixed operatingcosts, because the result is, thanks to the wind power project, less money willhave to be spent on new equipment and spare parts for the diesel power plantsthrough to the year 2028.




Fuel and lubrication oil substitution. The wind turbines avoid fuel consumption ofthe diesel system. The difference between non-fuel operating costs of the dieselpower plant and O&M costs of the wind park is also accounted for here.

External effects. These stand for the reduced level of harmful emissions.

Since costs and benefits arise at different points in time, the time factor must be ac-counted for in the form of cash-flow discounting, hence bringing costs and benefits in linewith a uniform initial date. A so-called standard discount rate (SDR) is used for discount-ing. The SDR is defined as the interest rate at which the company discounts a marginalfuture increase in consumption to its present value.This makes it possible to summarise and compare costs and benefits, each as a single

factor.

The following profitability criteria are postulated for the purposes of this analysis:

Benefits-cost ratio (B/C). The present values of the benefits are divided by the pre-sent values of the costs, and the project is profitable if the resultant benefits-costratio is greater than one.

Economic Internal Rate of Return (EIRR). The internal interest is the social dis-count rate at which the present values of costs and benefits are equal.

Economic Net Present Value (ENPV). Present value is the financial-mathematicalexpression used for the sum of the discounted values of a time series. The netpresent value is the difference between the present value of the benefits and thepresent value of the costs. The project is profitable, if the net present value is posi-tive.

The discount rate (the opportunity cost of capital) applied in the calculation of the ENPV

has been set at 10 %, in accordance with conversations held with EEPCO s Management.

Also, this rate is considered by the Ministry of Economic Development and Co-operation

as appropriate for Ethiopia at present. The indicated discount rate has also been applied

in other recent feasibility studies carried out for EEPCO, particularly in hydropower gen-

eration.




12.2 Economic Benefits

In the economic analysis the identified benefits are (i) the avoided capital costs, (ii) the

avoided fuel costs, (iii) avoided O&M costs and (iv) the avoided emissions.

12.2.1 Basic Diesel Power Plant Data

As already mentioned, the basic data for this study was based on one existing Diesel

Power Plant located in the northern part of Ethiopia.

This heavy fuel operated DPP commenced commercial operation in July 200422. The plant

consists of four state-of-the-art 18 cylinder, V-type, 4-stroke, medium speed Wärtsilä die-

sel engines type 38 coupled to ABB alternators: each genset is rated at 9,991 kW (site

capacity at alternator terminals). The net plant capacity exported to the grid at full load is

38 MW.

The production and the sale of Wärtsilä type 38 diesel gensets was stopped by Wärtsilä

within the course of strengthening Wärtsilä s engine and alternator portfolio. The type 38

was replaced by the type 46. The Wärtsilä 12V46 genset has a similar capacity as the

18V38 genset. The engine speed of the type 46 is 500 rpm.

12.2.2 Avoided Capital Costs

The economic feasibility is determined comparing the wind park with an equivalent diesel

power plant alternative. The wind parks relative to the equivalent diesel based generation

has been considered in terms of energy production. If the wind park would not be in-

stalled, additional energy would have to be provided by new diesel generators at higher

costs.

The chosen DPP is especially suited for the analysis, since its annual generation is similar

to the estimated energy production of the wind park.

The capacities of the diesel units at this power plant are derated because of the prevailing

site conditions. Due to the importance of this fact, for the correct interpretation of the cal-

culated capital costs and operation costs, the influence of the site conditions on the en-

gine s capacity will be further analysed.

22 Source: EEPCo, Project Completion Report for Dire Dawa 38 MW Heavy Fuel Oil Fired Power Plant , No-vember 2004.




Diesel engines are internal combustion engines whose capacity is mainly influenced by

the following site conditions: altitude of the site, combustion and cooling air temperature.

ISO 3046 Part 1 specifies the standard reference conditions, declarations of power, con-

sumption and test methods for diesel engines. The ISO standard reference conditions are

used as normative conditions for all diesel engines. The ISO conditions are as follows:

Barometric pressure: 100 kPa (corresponds to 100 m altitude)

Air temperature: 25°C

Relative humidity: 30%

Charge air coolant temperature: 25°C

Except for the humidity value, an increase of these values results in re-duced capacity of the diesel engine. ISO 3046 Part 1 specifies also the al-gorithms for the derating calculations.

The site reference conditions for the regarded DPP are as follows:

Site altitude: 1,200 m

Air temperature: 30°C

Relative humidity: 30% (estimated)

Charge air coolant temperature: 40°C

The ISO capacity at alternator terminals of each Wärtsilä unit is 11,058 kW. The site ca-

pacity is 9,991 kW. This comparison shows, that each diesel genset is derated by 9.65%

due to the site conditions, mainly due to the site altitude. This diesel power plant would

deliver approximately 9.65% more capacity, if it was installed between 0 and 300 meters

above sea level.

This comparison is important for the interpretation of the study s results, since the derated

capacity of the plant increases the specific costs of the plant compared to a non-derated

plant by approximately 9-10%.

The gross plant capacity (= at alternator terminals) is 4 x 9,991 kW = 39,964 kW.

The net plant capacity (= capacity exported to the grid) is 4 x 9,500 kW = 38,000 kW. The

difference between both values (= 1,964 kW = 4.9%) is the auxiliary power consumption

of the plant.

The technical and economic data of the reference DPP considered in the economic analy-

sis are presented in Table 12-1 and in Table 12-2.




Table 12-1: Technical data of the reference DPP

Engine Data Value

Engine model Wärtsilä 18V38

Engine Type 4 stroke

Number of cylinder 18

Scenario V

Cylinder bore 380 mm

Stroke 475 mm

Swept volume 970.2 dm³

Mean piston speed 9.5 m/s

Number of valves per cylinder 2 inlet, 2 outlet

Engine speed for 50 Hz 600 rpm

Aspiration Turbo charged & charge air cooled

Mean effective pressure (pme) 23.4 bar

Mechanical ISO capacity 11,340 kW

Mechanical site (derated) capacity 10,246 kW

Derating factor, calculated acc. ISO 3046-1 9.65% = 0.9035

Alternator Data Value

Alternator model ABB AMG 1250RR10

Phase Three phase

Insulation class F

Protection class IP 23

Frequency 50HZ

Power Factor (P.F.) 0.8

Output Voltage 15 kV

Efficiency 97.51 at 0.8 P.F. at 100 % load

Genset Data Value

ISO capacity at alternator s terminals 11,058 kW= 13,823 kVA (@ P.F. 0,8)

Site capacity at alternator s terminals9,991 kW = 12,489 kVA (@ P.F. 0,8)

4 x 9,991 kW = 39,964 kW

Net plant capacity (measured at the HV sideof the step-up transformer, according toISO 8528-1)

4 x 9,500 kW = 38,000 kW

Lifetime 15-20 years




Table 12-2: Economic data of the reference DPP

Value23

Economic DataValue in USD Value in ETB24

DPP Investment Requirements 31,069,272.98 USD 267,815,472 ETB

Capital Cost of the DPP @ ISO electri-cal gross capacity of 4 x 11,058 kW =44,232 kW

702.42 USD/kW 6,054.79 ETB/kW

Adjusted Cost for Derating (9.65 %) 770.20 USD/kW 6,639.08 ETB/kW

Capital Cost of the DPP @ site electri-cal net plant capacity of 4 x 9,500 kW = 38,000 kW

817.61 USD/kW 7,047.78 ETB/kW

The avoided capacity costs refer to the investment costs that would occur when installing

a DPP like the investigated 40 MW plant. The avoided capacity costs were calculated at

USD 817.61/kW.

Due to strong continued demand for diesel power plants over the last few years, and lim-

ited production capacity of the diesel engine & genset manufacturers, the specific costs of

diesel power plants (EPC contracts) are slightly, but nevertheless continuously increasing.

12.2.3 Avoided Fuel Costs

In the current economic analysis, avoided fuel costs are determined by the fuel prices and

the fuel consumption of the diesel gensets. The fuel prices are calculated as delivered to

the site .

The considered DPP is laid out for continuous operation on cheap heavy fuel oil (HFO).

Expensive light fuel oil (LFO) is only needed as back-up fuel: used during start-up and

shut down of the diesel engines, and when HFO is not available due to technical problems

of the transfer and/or fuel treatment systems.

HFO is a residual oil produced during the refinery process of crude oil. It has a lower qual-

ity and higher viscosity than LFO and consequently a lower price than LFO.

Depending on its viscosity, HFO must be kept heated during transport in order to avoid

problems during loading and unloading due to high viscosity.

23 Source: EEPCo, Project Completion Report for Dire Dawa 38 MW Heavy Fuel Oil Fired Power Plant , No-vember 2004 and exchange rates as per June 2nd 2006.24 Rate 1:8.61




The light fuel (LFO) used at the considered DPP varies between 45 and 80 centi Stokes

(cSt) at 50°C, and the HFO used, has a maximum viscosity of 180 cSt at 50°C.

The annual fuel consumption recorded in the operation log books of the regarded DPP

can be divided as follows:

HFO consumption: 82%

LFO consumption: 18%

Considering the same lower heating value (LHV) for HFO and LFO, the above mentioned

percentages do not need to be corrected and can be directly used for the fuel consump-

tion cost calculations.

The fuel consumption of diesel units of a state-of-the-art DPP is measured by means of

volumetric fuel flow meters with thermal correction and automatic data transfer to the con-

trol system of the plant (history/data record). Together with the produced energy (kWh;

GWh) a specific fuel consumption can be calculated, which is normally given in g/kWh,

referred to the alternator s terminal.

The stated HFO to LFO consumption ratio are values of the 38 MW Dire Dawa diesel

power plant (DPP). Based on the log sheet figures of this DPP, it can be derived that Dire

Dawa DPP was not operated in continuous mode, thus resulting in numbers of starts

above the average of comparable DPP´s.

Fuel losses occurring during the fuel treatment (water & sediment drainage, separation

and filtering of fuel) should be added to the above mentioned specific consumption, since

the lost fuel/water/sediment volume was also purchased by the Owner. These losses were

estimated in this case to 3% of the specific consumption.

The fuel consumption is normally based on a LHV of fuel of 42.7 MJ/kg. Other LHV values

can be also applied by calculating the specific consumption by means of a linear relation

between the two LHV s.

The fuel consumption of a diesel genset is also subject to derating based on the prevail-

ing site conditions like the genset s capacity. The genset s consumption is normally given

as ISO based value and as site based value. In case of the Wärtsilä 18V38 genset, the

ISO fuel consumption is 182.1 g/kWh (@ LHV 42.7 MJ) at the alternator s terminals, ±5%

tolerance. This represents an electrical efficiency of the gensets of 46.3% at ISO condi-

tions.

Considering the site conditions, the site fuel consumption of each unit is 200.3 g/kWh (@

LHV 42.7 MJ) at the alternator s terminals, ±5% tolerance. This represents an electrical

efficiency at site of the genset of 42.1 %.




Including 3% losses inside the plant, the net plant fuel consumption of each genset is

206.3 g/kWh (@ LHV 42.7 MJ). Including 5% tolerance the value is 216.6 g/kWh.

The following table summarises the relevant fuel consumption values:

Denomination Value Comments

Share of HFO consumption 82%

Share of LFO consumption 18%

ISO fuel consumption at the

alternator s terminals

182.1

g/kWh

As per manufacturer s handbook,

±5% tolerance

Site fuel consumption at the

alternator s terminals

200.3

g/kWh

Losses inside the plant

boundary3% Estimated value

Site fuel consumption at the

alternator s terminals including

losses

206.3

g/kWh= 8,809 kJ/kWh

For the economic analysis a fuel consumption of 207.0 g per generated kilo watt hour at

the alternator s terminals was considered. This corresponds to a net plant heat rate of

8,839 kJ/kWh.

Fuel prices vary strongly, proportional to the volatile price development of crude oil prices.

Table 12-3 shows stock market fuel prices at different international harbours in metric tons

(t) for different categories of fuel oil.

The fuel oil prices consists normally of the following components:

Fuel price at port of loading

Cost of sea transport and insurance

Cost at port of unloading: pumping, taxes, transport to fuel depot

Cost of land transport (tank truck) and insurance




Table 12-3: Fuel prices at different international ports in USD/metric ton

[USD/t] HFO 38025 HFO 18026 MDO27 MGO28

Singapore 333.50 343.50 624.00 631.50

Fujairah 334.50 353.50 648.50 648.00

Rotterdam 314.50 335.00 568.50 620.00

Houston 330.00 345.00 579.00 *

Source: www.bunkerworld.com; date: June 1st, 2006

May2006

HFO 380Price (USD/t)

Figure 12-1: HFO380 price development29

Figure 12-1 shows the HFO 380 price development in Singapore (yellow), Fujairah (red),

Houston (blue) and Rotterdam (green) during May 2006. The y-axis shows the fuel price

in USD/t. In the last months, prices tended to decrease.

Despite in the month of May of 2006 there has been a decreasing oil price trend, the brent

crude oil price development (see Figure 12-2) has had an increasing trend when analys-

ing the years 1997 till December 2005.

25 Heavy fuel oil (HFO) viscosity 380 cSt at 50°C 26 Heavy fuel oil (HFO) viscosity 180 cSt at 50°C27 Marine diesel oil28 Marine gas oil29 Source: www.bunkerworld.com, June 1st 2006




Figure 12-2: Development of the brent crude oil price from 1997 to 200530

The heavy fuel oil price for HFO 180 at Fujairah port is 353.50 USD/Mt. With an

estimated HFO density of 0.96, the metric ton is equivalent to 1,041.7 litres. The HFO 180

litre price at Fujairah port is USD 0.3393.

According to EEPCO, the HFO 180 price at Djibouti harbour amounts to

3.6949 ETB per litre as per May 2006. Considering the above mentioned density

and an exchange rate of 1 USD = 8.61 ETB, the price of HFO 180 is

446.51 USD/Mt.

The price for LFO at Djibouti port is 3.7754 ETB per litre corresponding to

USD 0.4379. With a maximum density of 0.90, the metric ton is equivalent to

1,111.1litres. Consequently, the LFO price is 486.65 USD/Mt.

According to EEPCO s data, the land based transport from Djibouti port up to the

dealer s depot in the region of the considered DPP is routed via 105 km paved

roads (= transport cost of 45.40 ETB/Mt) and 239 km gravel roads (= transport

cost of 125.10 ETB/Mt).

The fuel finally delivered by dealers from the depot to the considered DPP has the follow-

ing price increment covering the transport, service charges and profit of the dealers:

HFO: 0.4151 ETB/liter = USD 50.16/Mt

30 Source: OECD




LFO: 0.4146 ETB/liter = USD 53.44/Mt

The final fuel prices at the regarded DPP are as follows:

HFO: USD 516.45 /Mt = USD 0.4958/litre = 4.2737 ETB/liter

LFO: USD 559.87 /Mt = USD 0.5039/litre = 4.3435 ETB/liter

With a specific HFO price of 11.88 USD/GJ and a net plant heat rate of 8,839 kJ/kWh, the specific fuel cost is 0.106 USD/kWh.

The following table summarises the relevant fuel price data:Table 12-4: Relevant fuel price data

Denomination Value Comments

HFO IF 180 estimated density 0.96 1 Mt = 1,041.7 litres

LFO estimated density Max. 0.90 1 Mt = 1,111.1 litres

USD/ETB exchange rate 1:8.61

Distance Djibouti-depot 105 km paved road239 km gravel road

HFO (IF 180) price at Fujairah 353.50 USD/Mt

HFO price at Djibouti port 446.51 USD/Mt

LFO price at Djibouti port 486.65 USD/Mt

Transport cost Djibouti - depot 19.78 USD/Mt 344 km paved and gravelroad

HFO price at local depot 466.29 USD/Mt

LFO price at local depot 506.43 USD/Mt

Dealer s fixed charge for HFO

delivery depot - DPP0.4151 ETB/litre = USD 50.16.30/Mt

Dealer s fixed charge for LFO

delivery depot - DPP0.4146 ETB/litre = USD 53.44/Mt

HFO price at the DPP USD 516.45 /Mt = USD 0.4958/litre= 4.2737 ETB/litre

LFO price at the DPP USD 559.87 /Mt = USD 0.5039/litre= 4.3435 ETB/litre

Specific HFO cost 0.10690 USD/kWhnet = 0.92146 ETB/kWhnet

Specific LFO cost 0.11589 USD/kWhnet = 0.99896 ETB/kWhnet




A HFO price of USD 516.45/Mt (0.10690 USD/kWhnet) and a LFO price of 559.87/Mt

(0.11589 USD/kWhnet) were considered for the economic analysis.

12.2.4 Avoided non-Fuel O&M Costs

Depending on the plant load factor, operating hours and fuel price, the fuel costs repre-

sents more than 80% of the total operation costs of a DPP. The remaining costs, the

(non-) fuel O&M costs consist of fixed and variable costs.

The fixed O&M costs include all those cost items which will be incurred irrespective of an

operation of the plant s operation status. These fixed costs include costs for personnel,

insurance, management and administration, as well as general maintenance costs. The

general maintenance cost component includes costs of administration for services, con-

sumables, materials, supplies procured, costs of postage, telephone, facsimile, reproduc-

tions and travel expenses.

The variable O&M costs include such cost components which are only incurred if the plant

is operating. These costs comprise of lubrication oil and other consumables like chemi-

cals, etc. Variable costs for the power plant also include the cost for overhauls including

spare parts.

Each diesel engine has to undergo service and maintenance every

1500/3000/6000/12000/24000 and 36000/48000 operation hours. After 12000/24000 and

36000/48000 operation hours the diesel engines undergo major maintenance works,

which are very cost intensive. This means that the below mentioned specific variable O&M

costs (6 USD/MWh) for HFO operated DPP s of similar design and configuration is an

average value calculated within one whole operation cycle until major overhaul at

36000/48000 operation hours.

EEPCo s figures on variable and fixed O&M costs provided for the reference years 2004

to June 2006 are very low, even considering that the power plant has recently started

commercial operation. Among other reasons, this is due to a low plant load factor. Since

the annual fix non-fuel O&M costs provided by EEPCo could not be considered as repre-

sentative for the whole period of analysis31, moderate international standard cost esti-

mates settled at 100.000 USD for fix non-fuel O&M costs per year were applied in the

economic analysis.

31 Fix non-fuel O&M costs have been budgeted by EEPCo as per May 2006 as low as USD 48,525(ETB 426,791) for the period June 2005 to June 2006.




The estimation of avoided non-fuel variable O&M costs per year for the Diesel Power

Plant has been considered according to the Consultant s experience from similar African

Diesel Power Plants and according the following data:

Table 12-5: Non-fuel variable O&M for a DPP

Net plant capacity 38,000 kW

Assumed plant load factor, DPP as base load plant 75 %

Net plant energy production at HV side of step-up transformers 249,660 MWh/year

Specific variable O&M costs for HFO operated DPP s of similar

design and configuration all over the world (source Evaluation of

Institution of Diesel and Gas Turbine Engineers (IDGTE) Working

Cost and Operational Report 1997

6 USD/MWh

Estimated annual expenditure for variable O&M costs USD 1,497,960

12.2.5 Avoided Emissions

The avoided CO2 emissions are calculated considering that a DPP with an efficiency of

43 % emits 670 gr. of CO2 per kWh. The calculation follows the formula:

The results of applying the above formula to each Scenario are summarised in Table

12-6.




Table 12-6: Avoided emissions of the DPP

Avoided Emissions of the DPP(CO2 tonnes)

Scenario IEnercon E48

Scenario IIEnercon E53

Scenario IIIVestas V52

Scenario IVGamesa G58

Annual emissions avoided 135,355 155,848 136,300 164,437

Total avoided emissions (CO2 tonnes) 2,707,090 3,116,955 2,726,002 3,288,741

ASHEGODA SITE

In the economic analysis the wind park is compared with a diesel power plant. Thus, theavoided emissions refer to the DPP. The economic monetary quantification of the avoidedemissions has been based on the Mitigation Cost Approach. In the Mitigation Cost Ap-proach, the use of USD20/CO2 tonnes is considered as a reasonable estimate for theshadow price of carbon emissions and it is consistent with the existing work by many ex-perts:

Anderson et al. (1990, 1993) estimated a present-day USD25/CO2 tonnes shadow price using a carbon accumulation-backstop technology model based on the Hotel-ling rule;

Fankhauser (1995, 1996) estimates a global damage function for climate change,and derives a range of USD6-45/CO2 tonnes shadow price, with a best estimate of USD20/CO2 tonnes;

The Federation of American Scientists arrived at a shadow price of USD10-20/CO2

tonnes based on a Delphi-type assessment;

And simulations of the global carbon offset market performed by the Norwegianresearch group, ECON, indicate a future market price for carbon of USD10-30/tC;

whereas in ADB's Economic Evaluation of Environmental Projects (March 1996),Appendix H, Average Annual Global Climate Change Damages for Carbon Emis-sions are estimated at USD7.85-USD17.66/CO2 tonnes for 1991 to 2000, increas-ing to 8.90 USD/CO2 tonnes - 20.03 USD/CO2 tonnes for 2011 to 2020, and de-creasing thereafter.

The value has been also compared to the prices of CO2 emissions in the Euro-pean Emissions Trading System at the European Energy Exchange (EEX) basedin Leipzig (Germany). As showed in the figure below, prices have been over 20




/ CO2 tonnes (24 USD/ CO2 tonnes) almost all the time since October 200532, sothat a price of 20 USD/ CO2 tonnes has been applied in the economic analysis33.

CO2 certificate price development at the EEX

Figure 12-3: CO2 price development at EEX

32 VIK Mitgliedsrundschreiben 15/2006, March 3rd, 200633 In the financial analysis, current Certified Emission Reduction (CER) credit prices for Clean DevelopmentMechanism (CDM) activities have been considered at 6 USD/CER.




12.2.6 Diesel Summary Assumptions

The basic parameters of the DPP used in the economic evaluation are summarised in the

following table.

Table 12-7: Summary of basic assumptions of the reference DPP

Item Data Comment

Project Implementation Start Date 2006

Construction Period 18 months

Commercial Operation Date 2007

Tax and Duties Tax-free status

Exchange Rate ETB/USD 8.6199 : 1 Rate as per March 8th, 2006

Net Plant Capacity 38,000 kW

Plant Load Factor 75 %

Average Saleable Capacity 249,660 MWh/year This data has been adapted to each windfarm s output

Capital Costs 817.61 USD/kW @site electrical net plant ca-pacity of 4 x 9,500 kW

Fixed non-fuel O&M Cost USD 100,000 p.a. As per EEPCo information

Variable non-fuel O&M Cost USD 1,497,960 p.a. = 6 USD/MWh

Heat rate 8,839 kJ/kWh @ LHV

Specific HFO180 Fuel Cost 0.10690USD/kWhnet

share: 82 %

Specific LFO Fuel Cost 0.11589USD/kWhnet

share: 18 %




12.2.7 Indirect benefits

Main indirect benefits that have not been quantified, but are to be considered are:

the fact that the injected wind energy reduces the absolute consumption of dieselfuel, which is relatively expensive,

the generation of power will become more diversified,

the dependence on imported diesel fuel will decline,

the project confirms the energy-policy objectives of the Government of Ethiopia.

12.3 Economic Costs

In the economic analysis, the identified economic costs are (i) capital costs - investment

costs - of the wind park, (ii) operating costs of the wind power installation, and (iii) leakage

costs.

12.3.1 Investment Costs of the Wind Park

An itemised specification of investment costs (wind turbines, foundation, civil works, elec-

trical work, consulting services, physical and price contingencies, etc.) in actual prices

broken down in foreign and local cost components has been included in Section 9 Part I.

Internal prices in Ethiopia are considered to reflect an open market economy and do not

require further correction for distortions created by constraints of supply and demand in

the market. Based primarily on the conditions of foreign currency acquisition as well as to

take into account of the national allocation system of foreign currency, the Ministry of

Economic Development and Co-operation recommends that a shadow exchange rate

factor of 1.11 would be applied. This leads to a standard conversion factor (SCF) of 0.9 is

obtained for local currency expenditure, effectively reducing local costs accordingly when

expressed in foreign currency units. Similar conversion factors were also applied in other

recent studies carried out for EEPCo34.

In the economic analysis the SCF of 0.9 has been applied to the expenditures in local

currency, resulting in total investment costs for the different wind park scenarios as de-

tailed in the following tables.

34 Feasibility Study of Weles, Zhemoga-Yeda and Halele-Werabesa Hydropower Project , Lahmeyer Interna-tional Gmh in association Mid-day Consulting Engineers and Tropic Consulting Engineers, June 2005.




Table 12-8: Investment costs considering shadow prices

Enercon E48 WP Investment Costs Costs [USD] [%] SCF Economic Value [USD]

Investment in Foreign Currency 84,905,398 88.14% - 84,905,398

Investment in Local Currency 11,425,069 11.86% 0.9 10,282,562

Total Investment 96,330,468 100.00% 95,187,961

Enercon E53 WP Investment Costs Costs [USD] [%] SCF Economic Value [USD]




Vestas V52 WP Investment Costs Costs [USD] [%] SCF Economic Value [USD]




Gamesa G58 WP Investment Costs Costs [USD] [%] SCF Economic Value [USD]




12.3.2 Economic O&M Costs of the Wind Park

The standard conversion factor (SCF) of 0.9 for local currency expenditure, effectively

reducing local costs accordingly when expressed in foreign currency units, has also been

applied to the O&M costs. To this end, O&M costs have been divided into foreign and lo-

cal costs. For maintenance and repair of the wind turbines it is assumed that after suffi-

cient education of the local operation team, the work for these two activities can be exe-

cuted to a significant extend by the local personnel. For maintenance this portion is higher

than for repair, since for the repair procedures more specialised know-how is required and

it has thus, to be carried out by experienced wind energy foreign experts. (A further expla-




nation is included in Section 9.4 Part I). The following Tables include a detail of the annual

O&M costs for the Ashegoda Wind Park considered in economic values.

Table 12-9: Economic values of Enercon E48 annual O&M costs

WT-Type: Enercon E48 Years 1-20 SCF Economic Value

Item (USD/year) % (USD/year) % (USD/year) 0.9 O&M (USD)

Maintenance of wind turbines 137,554 35.0% 48,144 65.0% 89,410 80,469 128,613

Repairing of wind turbines 158,178 50.0% 79,089 50.0% 79,089 71,180 150,269

Consumables 100,077 70.0% 70,054 30.0% 30,023 27,021 97,075

Spare parts 1,514,533 100.0% 1,514,533 0.0% 0 0 1,514,533

International technical assistance 288,738 100.0% 288,738 0.0% 0 0 288,738

Local technical expenses 436,250 100.0% 436,250 392,625 392,625

Insurance 167,661 100.0% 167,661 150,895 150,895

Insurance of wind park personnel 267,331 100.0% 267,331 240,598 240,598

Power demand of the turbines 144,997 100.0% 144,997 130,497 130,497

Subscriptions 6,262 100.0% 6,262 5,636 5,636

Office costs, materials & others 247,483 100.0% 247,483 222,735 222,735TOTAL (USD/year) 3,469,064 2,000,558 1,468,506 1,321,656 3,322,213TOTAL (%) 100.0% 57.7% 42.3% 38.1% 95.8%

Foreign Portion Local Portion

Table 12-10: Economic values of Enercon E53 annual O&M costs

WT-Type: Enercon E53 Years 1-20 SCF Economic Value

Item (USD/year) % in USD % in USD 0.9 O&M (USD)



Consumables 100,077 70.0% 70,054 30.0% 30,023 27,021 97,075

Spare parts 1,532,657 100.0% 1,532,657 0.0% 0 0 1,532,657



Insurance 167,661 100.0% 167,661 150,895 150,895



Subscriptions 6,262 100.0% 6,262 5,636 5,636






Table 12-11: Economic values of Vestas V52 annual O&M costs

WT-Type: Vestas V52 Years 1-20 SCF Economic Value




Consumables 100,077 70.0% 70,054 30.0% 30,023 27,021 97,075

Spare parts 1,718,662 100.0% 1,718,662 0.0% 0 0 1,718,662



Insurance 167,661 100.0% 167,661 150,895 150,895



Subscriptions 6,262 100.0% 6,262 5,636 5,636



Table 12-12: Economic values of Gamesa G58 annual O&M costs

WT-Type: Gamesa G58 Years 1-20 SCF Economic Value




Consumables 100,077 70.0% 70,054 30.0% 30,023 27,021 97,075

Spare parts 1,734,068 100.0% 1,734,068 0.0% 0 0 1,734,068



Insurance 167,661 100.0% 167,661 150,895 150,895



Subscriptions 6,262 100.0% 6,262 5,636 5,636



Further, a major overhaul of all equipment has been assumed to take place between the

10th and 11th years of operation in an amount of 5 % of total investment costs.

Also wind farm decommissioning costs in operational year 21 have been considered with

an amount of 1 % of total investment costs.




12.3.3 Leakage Costs of the Wind Park

No leakage costs (or other external costs) could be identified for the wind park activity.

Leakage is defined by the United Nations Framework Convention on Climate Change

(UNFCCC) in its Guidelines for Completing CDM Project Design Documents, Version 02,

as the net change of antropogenic emissions by sources of GHG which occurs outside

the project boundary, and which is measurable and attributable to the project activity .

The project activity essentially involves the generation of electricity from wind, the em-

ployed wind turbines can only convert wind energy into electrical energy and cannot use

any other input fuel for electricity generation. Thus, no fuel leakage costs occur from the

wind park project.

12.4 Results: Economic Analysis

The economic appraisal of the Ashegoda Wind Park has been carried out by comparing

the cash flow associated with construction and operation the wind power scheme with the

cash flow for the construction and operation of the equivalent least cost thermal alterna-

tive plant (diesel power plant). In the appraisal, the avoided costs of thermal generation

are regarded as benefits attributable to the Wind Power Project. The difference between

the costs of the wind power project and the benefits of the avoided thermal power and

energy has been determined over a 20 year operational period. With regard to implemen-

tation of the wind power plant, a fast-track schedule has been adopted. Only a fast-

track schedule will come close to meeting EEPCo s short and long-term strategic in-

stalled capacity target. Under the fast-track implementation schedule, construction

probably will start at the end of 2006, with the first energy feeding into the ICS in 2007.

The results of the comparison of the proposed Ashegoda Wind Power Project develop-

ment with an equivalent DPP are shown in Table 12-13.

Three main economic parameters have been used to evaluate the economic feasibility of

the wind park: the EIRR, the Benefit/Cost Ratio, and the ENPV calculated at a 10 % dis-

count rate.

12.4.1 Economic Cash-flow Projections

Cash flow projections associated with construction and operation of the wind parks have

been compared with the cash flow projections of construction and operation of the equiva-




lent least cost thermal alternative plant, in this case, an emergency diesel power plant with

nominal capacities ranging from 14 MW to 16 MW depending on the estimated power

output of the wind park.

In the study, the avoided costs of thermal generation are regarded as benefits attributable

to the wind park project. The difference between the costs of the wind park project and the

benefits of the avoided thermal power and energy has been determined over a 20 year

operational period. Further the economic benefits of avoided emissions have been quanti-

fied.

12.4.2 EIRR and NPV

In this study, the EIRR is defined as the discount rate that causes the present value of the

project costs to be equal to the present value of the benefits. The EIRR indicates the ac-

tual profit rate of the total investment outlay. The project is feasible if the EIRR is greater

than the agreed economic discount rate. It is given by the following equation:

i=1

nnet flow i

(1 + R )i-1= 0

where n denotes calculation period (years) and R denotes discount rate.

As indicated in the assumptions, the discount rates for the basic scenarios are 10 %.

The ENPV of an investment is the present (discounted) value of future cash in-

flows minus the present value of the investment and any associated future cash

outflows. The ENPV of the Ashegoda Wind Park has been calculated at different

discount rates (8 %, 10 % and 12 %). Results are indicated in the table below.

12.4.3 B/C Ratio

In the Benefit/Cost (B/C) Ratio, the total discounted benefits are divided by the total dis-

counted costs. Projects with a benefit-cost ratio greater than 1 have greater benefits than

costs as well as positive net benefits. The higher the ratio, the greater the benefits relative

to the costs.




Table 12-13: Results economic analysis Ashegoda wind park

Key Economic ParametersDiscount

Rate

Wind Turbine Type

(Scenario) EIRR (%) B/C Ratio ENPV (Mio USD)

8% 1.97 107.78

10% 1.81 91.59

12%

Enercon E48 30.23 %

1.66 64.21

8% 2.16 134.83

10% 1.98 115.86

12%

Enercon E53 33.44 %

1.81 82.37

8% 1.89 104.04

10% 1.73 87.42

12%

Vestas V52 27.76 %

1.59 60.39

8% 2.22 144.45

10% 2.04 124.78

12%

Gamesa G58 35.35 %

1.87 89.30

The results of the economic analysis are highly positive, showing that the wind park in all

four Scenarios and at a discount rate of 10 % is highly economically feasible. The highest

result is produced by the Scenario with the Gamesa G58 followed by the Scenario with

Enercon E53 wind turbines, Scenario I with Enercon E48 and Scenario III with Vestas

V52.




12.5 Scenario Analysis

A scenario analysis has been carried out for the wind park scenario with the highest EIRR, highest B/C Ratio and highest net benefits, i.e., the Scenario IV (57 Gamesa G58 tur-bines). Changes in (i) avoided capacity costs, (ii) diesel fuel prices, (iii) CO2 penalties and (iv) electricity generation and their impact on the EIRR have been evaluated.

Change in avoided Capacity Costs

As indicated previously, avoided capacity costs are calculated as the difference between

capacity costs of installing a DPP and the capacity costs of implementing a wind park.

These avoided capacity costs are negative since the investment costs of the wind park

are higher than the costs of the DPP.

The effect of increasing and reducing the wind park investment costs has been studied in

two cases:

o Best Case: Investment costs 10 % lower than in the Base Case.

o Worst Case: Investment costs 10 % higher than in the Base Case.

Change in Fuel Prices

Oil prices oscillate along the time. (Figure 12-4 for the development of the crude price in

USD per barrel (bbl) from April 2004 to March 200635).

35 Source: www.tecson.de




Figure 12-4: Crude price (USD/bbl) from April 2004 to March 2006

The impact of oil price variations (i.e., oscillations in the HFO crude oil price at DPP

516.45 USD/Mt and in the LFO oil price at DPP 559.87 USD/Mt-) , have been analysed

in the scenario analyses by modelling two cases:

Best Case: with an annual increase of 2 % on HFO & LFO prices at the DPP;

Worst Case: with an annual decrease of 2 % HFO & LFO prices at the DPP.

Figure 12-5 reflects the three scenarios used in the economic analysis, where in year

2028 HFO prices are expected to increase until 798.42 USD/Mt and LFO until 865.55

USD/Mt in the Best Case and to decrease until 331.13 USD/Mt for HFO and 358.97

USD/Mt for LFO in the Worst Case.




Figure 12-5: Scenario analysis: HFO and LFO price development

Change in Emission (CO2) Penalties

Two cases have been tested in the scenario analysis in the costs of mitigating CO2 emis-sions (penalties for emitting CO2), which were set at 20 USD/t in the base case:

Best Case: emissions penalty is set at 25 USD/t and

Worst Case: emissions penalty is set at 15 USD/t.

Change in Electricity Generation

The base case has been calculated assuming a Probability of Exceedance of 75 % (P75).For the scenario analysis two further cases have been considered:

Best Case: Probability of Exceedance of 50 % (P50);

Worst Case: Probability of Exceedance of 90 % (P90).

A definition of the Probability of Exceedance can be found in Section 6 of Part I).

12.5.1 Results: Economic Scenario Analysis

The scenario analysis shows that the variable with the highest impact on the EIRR is the

investment cost of the wind park followed by the electricity generation estimates. The best

results are obtained when decreasing investment costs by 10 %, whereas the impact on

EIRR of increasing emission penalties is from an economic point of view very low. The

following table summarises the results obtained in the scenario analysis.

Scenario Analysis - Oil Price Development -

20.00

120.00

220.00

320.00

420.00

520.00

620.00

720.00

820.00

920.00

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028

Year

USD/Mt

Base Case (HFO) Best Case (HFO) Worst Case (HFO)Base Case (LFO) Best Case (LFO) Worst Case (LFO)




Table 12-14: Summary results of scenario analysis

Scenario AnalysisVariable

Best Case Worst Case

Avoided capacity costs EIRRInv(-10%) = 41.46 % EIRRInv(+10%) = 30.77 %

Change in fuel prices EIRRFuel Price(+2%)

..= 38.99 % EIRRFuel Price(-2%) = 31.72 %

Emission penalties EIRR (25USD/t) = 36.57 % EIRR (15USD/t) = 34.14 %

Electricity Generation (P) EIRR (P50) = 39.37 % EIRR (P90) = 31.72 %

12.6 Conclusions: Economic Analysis

The economic appraisal of the Ashegoda Wind Park scheme has been carried out by

comparing the cash flow associated with construction and operation of the wind park with

the cash flow of construction and operation of the equivalent least cost thermal alternative

plant36. In the appraisal, the avoided costs of thermal generation are regarded as benefits

attributable to the Ashegoda Wind Park Project. The difference between the costs of the

Ashegoda project and the benefits of the avoided thermal power and energy has been

determined over a 20 year operational period37. With regard to implementation of the wind

park, a fast track schedule has been adopted. Only a fast track schedule will come close

to meet EEPCO s short and long-term strategic installed capacity target. Under the fast

track implementation, construction probably will start in 2007, with the first energy feeding

into the ICS in 2007.

The comparison of the proposed Wind Power Project with an equivalent thermal plant has

been made for 4 different Scenarios (Enercon E48, Enercon E53, Vestas V52 and

Gamesa G58). The results (Table 12-13) show that all scenarios are economically feasi-

ble, being the best Scenario the wind park with Gamesa G58 wind turbines followed by

Enercon turbines type E53 and E48. Since all the scenarios produce an EIRR higher than

36 The least costs thermal alternative plant has been defined by EEPCo as a Diesel Power Plant.37 Cash-flows are presented for 20 year operational period plus decommissioning in year 21.




the discount rate of 10 % settled by the Ministry of Economic Development and Co-

operation38 for Ethiopia, all the scenarios can be considered as economically feasible.

For the Scenario with the highest produced results, (Scenario IV with Gamesa wind tur-

bines) a sensitivity analysis has been carried out. Four variables have been subject to the

sensitivity analysis: (i) changes in avoided capital costs through an increase /decrease on

the investment costs of the wind park in +10 %/-10 %; (ii) changes of fuel prices, i.e., an

annual increase/decrease on fuel prices of +2 %/-2 %; (iii) an increase/decrease of the

emission penalties from 20 USD/t considered in the base case scenario to 25 USD/t and

15 USD/t considered in the best and worst cases, respectively; and finally, (iv) an in-

crease/decrease in electricity output.

The results of the economic sensitivity analysis have shown that changes on the invest-

ment costs of the wind park have the major influence on the economic results. If the in-

vestment costs could be negotiated and reduced by 10 %, the EIRR would increase from

35.35 % to 41.46 %.

38 Information provided by EEPCo during first mission.




13 CDM Assessment

A detailed assessment of the Clean Development Mechanism (CDM) has been carried out

in order to evaluate the institutional and regulatory framework for CDM in Ethiopia as well

as the opportunities to register the project as a CDM activity. Furthermore, the potential

amount of annual Certified Emission Reduction (CER) credits to be generated by the wind

park has been calculated and considered in the financial analysis.

13.1 Introduction

The Clean Development Mechanism (CDM) as established under Article 12 of the Kyoto

Protocol (KP) represents a collaborative policy approach between industrialised and de-

veloping countries, aimed at promoting sustainable infrastructure projects in developing

countries whilst simultaneously reducing greenhouse gas (GHG) emissions such as CO2.

From the perspective of industrialised countries, CDM represents a cost-effective option to

comply with their national emission reduction obligations under the KP. Developing coun-

tries benefit from CDM as an additional co-financing source for sustainable energy pro-

jects. Its applicability to particular projects, however, depends on how these cope with

country-specific and project-specific eligibility requirements. This section provides an as-

sessment of the applicability of the CDM to the envisaged wind park.

13.2 Institutional Framework for CDM Projects in Ethiopia

In order to be eligible for CDM, aspiring host countries like Ethiopia need to be Parties of

the United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto

Protocol. Furthermore, each host country needs to have established a Designated Na-

tional Authority (DNA) for the CDM, which co-ordinates the CDM approval process on the

national level on behalf of the host country government.

Ethiopia registered with the Kyoto Protocol on greenhouse gas emissions in April 2005 as

a "non-Annex I state", which means it has not undertaken measures to meet specific tar-

gets; its future plans should emphasise the development of power plants to reduce growth

in carbon emissions. Hence, Ethiopia fulfils the first main requirement to become a CDM

host country.

Furthermore, Ethiopia has established a Designated National Authority (DNA) for the

CDM, namely the Ethiopian Environmental Protection Authority (EPA). EPA s objective is

to formulate policies, strategies, laws and standards, which foster social and economic

development in a manner that enhances the welfare of humans and the safety of the envi-




ronment sustainable, and to spearhead in ensuring the effectiveness of the process of

their implementation. 39

The contact address is mentioned below:

Environmental Protection Authority (EPA)

Contact: Yeka Kifleketema

P.O. Box 12760

Gurd Shola, Addis Ababa

The Federal Democratic Republic of Ethiopia

Phone: 251 -1 46 46 07 / 46 48 80 / 46 38 43

Email: [email protected]

http: www.epa.gov.et/index.html

39 http://www.epa.gov.et/AboutEPA.htm; April 4th 2006




13.3 CDM Project Cycle

Developing countries and transition economies may host a project and benefit from CDM

as an additional co-financing source. Project sponsor can be either the host country or

usually a project sponsor of an Annex I country. In the case of a wind energy project, for

instance, the owner of the wind park in a non-Annex I country generates revenues from

electricity sales in the national market and also from the sale of Certified Emission Reduc-

tion (CER) credits.

Project Sponsor in Host Country

EPPCo

Host CountryETHIOPIA

Kyoto Protocol

Compliance

(e.g. NDA)

Carbon Fund /CER Buyer

Letter of Approval

Emission Reduction

Purchase Agreement

(ERPA)ERs

LendersLoan

Debt Service

Customers

electricitysales

revenues

CERssales

revenues

monetizationof ERPA

Figure1: CDM structure

CERs can be sold to carbon exchanges or to a CER acquisition program such as a car-

bon fund. Carbon funds offer the opportunity to project owners to close an Emission Re-

duction Purchase Agreement (ERPA) for the whole CER crediting period, usually 10

years, which avoids risks of oscillating CER market prices.

A potential buyer for the CERs generated by a wind park project could be found, for ex-

ample, through the Community Development Carbon Fund of the World Bank, which links

small scale projects seeking carbon finance with companies, governments, foundations,

and through NGOs looking for improvement of livelihoods in local communities and obtain




verified Emission Reductions. The advantages of collaborating with an established carbon

fund include, for example:

the signing of a long term ERPA (likely to be characterised by a low price, but with theguarantee of having an assured buyer); and

possibility to have a buyer/investor which covers the associated up-front payments.

Under the second point, the target would be to find an investor willing to overtake the up-front costs, which relate to, for example:

the preparation of a Project Design Document (PDD);

the project validation by a Designated Operational Entity (DOE); and

costs incurred for registration at the UNFCCC CDM Executive Board.

13.4 Emission Reductions attributable to the Ethiopian WPs

CDM project activities must reduce emissions below those emissions that would have

occurred in the absence of the CDM project activity. Due to this claim and due to the in-

creasing demand of electricity in Ethiopia it is necessary to compare the emissions

avoided by realising the CDM project according to a reasonable alternative of electric en-

ergy generation such as wind energy. To do so, the approved consolidated baseline

methodology ACM0002, Version 05, as of 3rd March 2006, Consolidated baseline meth-

odology for grid-connected electricity generation from renewable sources , which is appli-

cable to grid connected power generation project activities including wind farms has been

applied by the Consultant.

The ACM002 methodology is applicable because the proposed project activity:

implies electricity capacity additions from wind sources;

does not imply switching from fossil fuels to renewable energy at the site of the project activity;

and its geographic and system boundaries can be clearly identified and information onthe characteristics of the grid is available.

The baseline scenario assumed for the proposed project activity consists of electricity

supplies, which in the absence of the project would have been generated by the op-

eration of grid-connected power plants in the Ethiopian supply system and by the addition

of new generation sources.




The baseline emission factor (EFy) is calculated as a combined margin (CM), consisting of

the combination of operating margin (OM) and build margin (BM) factors according to the

following three steps:

Step 1: Calculation of the Operating Margin

For the derivation of the Operating Margin emission factor (EFOM,y), the baseline method-

ology ACM0002 / VERSION 05, 03 MARCH 2006 provides four different procedures:

Dispatch Data Analysis OM , or

Simple OM, or

Simple adjusted OM, or

Average OM.

Dispatch data analysis should be the first methodological choice. However, in the case of

the proposed project activity, the data collection, processing and analysis would cause

very high transaction costs, which cannot be justified by the minor gain of accuracy that it

might provide. Even with regard to the Ethiopian power plant structure shown in the table

below and granted that hydropower jointly comprised more than 50 % in 2004 of the

Ethiopian power generation, the Simple adjusted OM calculation is the method which

should be used. For the calculation of the Simple adjusted OM, the power sources of the

Ethiopian power supply system are separated into low cost/must run power sources (k) -

in Ethiopia singly hydropower -and other sources (j), i.e. thermal power plants. For both

groups, the average emission factor needs to be derived. In the case of hydropower, this

is obviously zero. (See Table 13-1 for a list of the low cost/must run power sources in

Ethiopia and other sources, i.e. diesel power plants).




Table 13-1: Ethiopian power plant structure40

Installed capacity

Plant [MW]Gilgel Gibe 184.00Finchaa 134.00Tis Abay II 75.00MELKA WAKANA 153.00Awash III 32.00Awash II 32.00Tis Abay I 11.40Koka 43.20Total Hydro 664.60Kaliti I 9.00Awash 22.40Dire Dawa 40.00Adwa 3.00Adigrat 1.10Shire 0.80Mekele 1.30Axum 0.55Nekempt 1.70Ghimbi 0.30Jimma 0.10

Total Diesel 80.25Total all 744.85

DIESEL

Source

Hydro

EF OM, simple adjusted, y =

kyk

kikiyki

y

jyj

jijiyji

y GEN

COEFF

GEN

COEFF

,

,,,,

,

,,,,

1

40 Source: EEPCo, www.eepco.gov.et/brief.html; April 6th, 2006 12:10 and EEPCo Power Sector ExtensionPlan, April 2004

= 1.05 t CO2

EF OM, 2004 /GWh




where:

yjiF ,, is the amount of fuel i (in a mass or volume unit) consumed by relevant

sources j in year(s) y;

yjiCOEF ,, is the CO2 emission coefficient of fuel i (t CO2 / mass or volume unit

of fuel), taking into account the carbon content of the fuels used by relevant powersources j and the percent oxidation of the fuel in year(y) ;

yjGEN , is the electricity (MWh) delivered to the grid by source j; and

%y Hours per year for which low cost/must run sources are on the mar-

gin 8,760 hours per year

where lambda ( y) was assumed to be almost 1, because the Ethiopian power plant port-folio consists on 95% of hydro.41

The CO2 emission coefficient COEFi is obtained as

iiCOii OXIDEFNCVCOEF ,2

COEFi= 0.000011 GWh/l x 809.58 CO2t/GWh x 0.955

COEFi= 0.008505 CO2t/l

where:

NCVi is the net calorific value (energy content) per mass or volume unit of a fueli,

OXIDi is the oxidation factor of the fuel,

EFCO2,i is the CO2 emission factor per unit of energy of the fuel i.

Since local values of NCVi and EFCO2,i were not available, country-specific values (e.g.IPCC Good Practice Guidance) were used as preferable to IPCC world-wide default val-ues which can be seen in the Annex , using a 3-year average based on the most recentstatistics available.

41 Note that even a slight decrease in -value advances the CDM results as well as the project cash flow significantly. It isrecommended to pay attention on this value beyond this report.




13.4.1 Step 2. Calculation of the Build Margin

The Build Margin emission factor (EFBM,y) is calculated as the generation-weighted aver-age emission factor (t CO2 / MWh) of a sample m of power plants, as follows:

mym

mimiymi

yBMGEN

COEFF

EF,

,,,,

,

(268,961.81 liters/GWh) x 0.008505 tonnes CO2/l

1,142.91 GWh2004BMEF

16.39 GWh x

where

Fi,m,y, COEFi,m and GENm,y are analogous to the variables described for the simple OMmethod above for m plants.

The sample group m consists of either

the five power plants that have been built most recently, or

the power plants capacity additions in the electricity system that comprise 20 % of thesystem generation (in MWh) and that have been built most recently.

The sample group that comprises the larger annual generation should be used for thiscalculation (noting that power plant capacity additions registered as CDM project activitiesshall be excluded from the sample group m).

With regard to the calculation of the Build Margin emission factor (EFBM,y), the ACM0002 /VERSION 05, 03 MARCH 2006 allows to choose between one of the following two op-tions:

Option 1

Calculate the Build Margin emission factor EFBM,y ex-ante based on the most recentinformation available on plants already built for sample group m at the time of PDDsubmission. The sample group m consists of either the five power plants that havebeen built most recently, or the power plant capacity additions in the electricity systemthat comprise 20% of the system generation (in MWh) and that have been built mostrecently.6 Project participants should use from these two options that sample groupthat comprises the larger annual generation.

Option 2

For the first crediting period, the Build Margin emission factor EFBM, y must be up-dated annually ex-post for the year in which actual project generation and associatedemissions reductions occur. For subsequent crediting periods, EFBM, y should be cal-

2004BMEF = 2/GWh32.8 t CO




culated ex-ante, as described in option 1 above. The sample group m consists of ei-ther the five power plants that have been built most recently, or the power plant capac-ity additions in the electricity system that comprise 20% of the system generation (inMWh) and that have been built most recently.7 Project participants should use fromthese two options that sample group that comprises the larger annual generation.

Taking into account the available information, in the analysis the Consultant has madeuse of the first option.

13.4.2 Step 3. Calculation of the Baseline Emission factor

The baseline emission factor EFy is calculated as the weighted average of the OperatingMargin emission factor (EFOM,y) and the Build Margin emission factor (EFBM,y):

50% x 1.05 + 50% x 32.802004EF =

16.92 tonnes CO2/GWh2004EF =

BMBMOMOMy EFwEFwEF ×+= ×

where:

the weights wOM and wBM, by default, are 50% (i.e., wOM = wBM = 0.5), and EFOM,y and EFBM,y are calculated as described in Steps 1 and 2 above and are expressed

in t CO2/MWh.

According to ACM0002 / VERSION 05, 03 MARCH 2006, alternative weights can be

used, as long as wOM + wBM = 1. The annual GHG emission reduction is calculated as the

product between the annual net electricity production fed into the grid, and the average

emission intensity of the existing power plant fleet (the grid emission factor usually re-

ferred to as the baseline emission factor, BEF).

Hydropower plants are free of any emission during the electricity generation process and

go into the equitation with a value of zero. The average emission factor of the Ethiopian

thermal units has been calculated as 2,287.48 tCO2 per GWh of electricity produced. It

follows from above that the EF operation margin, EF OM 2004; derived was equivalent to

1.05 tCO2 / GWh.

At next, the Build Margin emission factor was calculated. It was derived from a batch (m)

of power plants already built. Three Scenarios were considered for the Build Margin.




As stated before, the sample group m consists of either

(Scenario A) the five power plants that have been built most recently, which accountsfor 1,142.91 GWh of the total capacity generation as can be seen in the table below,or

Table 13-2 Baseline Scenario A

Scenario AInstalled capacity Commissioning Gerneration Generation Share

Plant [MW] Year [GWh] %Kaliti I 9.00 2004 2.07 1.21%Awash 22.40 2004 5.14 3.01%Dire Dawa 40.00 2004 9.18 5.37%Subtotal DPP 71.40 16.39Finchaa 134.00 2003 750.76 17.99%Tis Abay II 75.00 2001 375.77 10.07%Subtotal HPP 209.00 1126.53Subttal A 280.40 1142.91Adwa 3.00 1998 0.40%Shire 0.80 1995 0.11%Adigrat 1.10 1995 0.15%Mekele 1.30 1993 0.17%Axum 0.55 1992 0.07%MELKA WAKANA 153.00 1988 20.54%Ghimbi 0.30 1984 0.04%Nekempt 1.70 1984 0.23%Awash III 32.00 1971 4.30%Awash II 32.00 1966 4.30%Tis Abay I 11.40 1964 1.53%Koka 43.20 1960 5.80%

(Scenario B) the power plant capacity additions in the electricity system that comprise20 % of the system generation (in MWh) and that have been built most recently con-sist on the Gilgel Gibe HPP with a share of 24.70 % in the total electricity.

In Scenario B, a comparison of HPP and wind power makes no sense in this context,

since both technologies are free of any emission during the electricity generation process.

Computing a Build Margin emission factor in this context would not increase emission

savings. Therefore, the Consultant decided to distinguish two cases in Scenario B.

Build Margin: Case B1 and Case B2

The batch of scenario (B) consists of the power plant capacity additions in the electricity

system that comprises 20 % of the system generation in GWh. As the CDM aims to help

non-Annex I Parties such as Ethiopia achieve sustainable development i.e. by reducing

GHG emissions, the target of this section was to analyse whether current and future GHG




emissions can be avoided per substitution of polluting technologies, i.e. current diesel

power plants. As mentioned above Scenario B is divided into two cases.

Case B1 consists on three recently build DPP Kaliti I, Awash and Dire Dawa as well as

the latest HPP of Gilgel Gibe. Together covering about 34.29% of ICS annual generation

capacity.

Table 13-3: Baseline Scenario B1

Scenario B1Installed capacity Commissioning Gerneration Generation Cummulative Generation

Plant [MW] Year [GWh] Share [%] Share [%]Kaliti I 9.00 2004 2.07 1.21% 1.21%Awash 22.40 2004 5.14 3.01% 4.22%Dire Dawa 40.00 2004 9.18 5.37% 9.59%Subtotal DPP 71.40 16.39Gilgel Gibe 184.00 2004 378.71 24.70% 34.29%Subtotal HPP 184.00 378.71Subtotal B 255.40 395.10

Case B2 consist only on the Gilgel Gibe DPP since it is even covering more than 20 % ofthe ICS generation capacity namely 24.7 % at 378.71 GWh (as is deployed in the follow-ing table).

Table 13-4: Baseline Scenario B2

Scenario B2Installed capacity Com. Year Gerneration Generation Cummulative Generation

Plant ? [MW] [GWh] Share [%] Share [%]Gilgel Gibe 184.00 2004 378.71 24.70% 24.70%Subtotal B2 184.00 378.71Dire Dawa 40.00 2004 5.37% 30.07%Awash 22.40 2004 3.01% 33.08%Kaliti I 9.00 2004 1.21% 34.29%Finchaa 134.00 2003 17.99% 52.28%Tis Abay II 75.00 2001 10.07% 62.35%Adwa 3.00 1998 0.40% 62.75%Adigrat 1.10 1995 0.15% 62.90%Shire 0.80 1995 0.11% 63.01%Mekele 1.30 1993 0.17% 63.18%Axum 0.55 1992 0.07% 63.25%MELKA WAKANA 153.00 1988 20.54% 83.80%Nekempt 1.70 1984 0.23% 84.02%Ghimbi 0.30 1984 0.04% 84.06%Awash III 32.00 1971 4.30% 88.36%Awash II 32.00 1966 4.30% 92.66%Tis Abay I 11.40 1964 1.53% 94.19%Koka 43.20 1960 5.80% 99.99%

Comparing these three options and according to the UNFCCC (methodology AMC0002)

the sample group that comprises the larger annual generation has to be used. The result

of the calculation Scenario (A) the five power plants that have been built most recently ,

was 1,494.25 GWh which is comprised of 58.15 % of the total system generation (ap-

proximately 2,570 GWh based on EEPCo s 2004 data). While, the result of the calculation

for (B1) the power plants capacity additions in the electricity system that comprise 20 %




of the system generation which has been built most recently including the three diesel

power plants was 764.44 GWh and (B2) excluding DPP was 378.71 GWh. The result of

scenario A was selected because the higher generation has to be chosen comparing one

to another. As mentioned above, the grid emission factor is calculated as the weighted

average of the Operating Margin emission factor and the Build Margin emission factor,

which results in a value of approximately 16.92 t CO2 per GWh.

The average emission factor of the Ethiopian grid is very low when compared with the

emission factor of other African countries like Egypt, where Lahmeyer International GmbH

has estimated it to be 520 t CO2 per GWh. Nevertheless, and since a 100 MW coal power

plant (Yayu coal-fired power station42) is currently under planning, the average emission

factor of the Ethiopian grid is expected to increase 10 % instead of decreasing in the sec-

ond and third crediting period.

Taking into account the expected annual net output of the foreseen Ashegoda Wind Park

Project (in P75, P50, P90 and in the four different wind park scenarios) the expected an-

nual and total emission reductions considering a crediting period of 7 x 3 years, are shown

in the table below:

Table 13-5: Emission reductions & CER revenues for P75, P50 and P90

PoEAshegoda SiteWP Configuration

EstimatedGeneration

[GWh]

Emission Reduction Potential [t CO2/yr]

Annual CER Revenues

[USD]

TOTAL[t CO2/yr]

TOTAL CER Revenues

[USD]

Enercon E48 197.392 3,340.64 20,043.84 60,666.03 363,996.2Enercon E53 227.278 3,846.43 23,078.57 69,851.13 419,106.8Vestas V52 198.771 3,363.98 20,183.87 61,089.85 366,539.1Gamesa G58 239.804 4,058.42 24,350.50 73,700.84 442,205.1



P75

P50

P90

In order to evaluate the financial impact of the emission reductions, the following assump-

tions as stated in the next Table were considered. An explanation of these assumptions is

entitled in Section 14.1.

42 EEPCo builds 52 milion birr power distribution network , Capital: Local News, May 21st, 2006. Vol. 8 No.388.




Table 13-6: CDM Assumptions

Parameter Unit Value

Crediting Period (Chosen option: 3 x 7 ) Years 7 + 7 + 6 = 20

CDM Up-front Costs (PDD / Validation / Registration) USD0 (assumed to be

granted)

CDM Periodic Costs USD/year 5,000

Average emissions factor of Ethiopian grid (1st to 7th year) tCO2/GWh 16.92

Average emissions factor of Ethiopian grid (8st to 14th year) tCO2/GWh 16.75

Average emissions factor of Ethiopian grid (15th to 21st year) tCO2/GWh 16.58

Table 13-7: CER generation and CER revenues with Enercon E48 turbines

Annual generation of CERs (1st to 7th year) tCO2/year 3,340


Annual generation of CERs (15th to 21st year) tCO2/year 2,706

Assumed average CER price USD/ tCO2 6

Annual net revenues from CER sales (1st to 7th year) USD/year 20,040

Annual net revenues from CER sales (2nd to 14th year) USD/year 18,036

Annual net revenues from CER sales (15th to 21st year) USD/year 16,236

Table 13-8: CER generation and CER revenues with Enercon E53 turbines











Table 13-9: CER generation and CER revenues with Vestas V52 turbines








Table 13-10: CER generation and CER revenues with Gamesa G58 turbines











13.5 Conclusions: CDM Assessment

Due to the high portion of hydropower resources in the electricity grid of Ethiopia, which

results in a low emissions factor, the annual emission reductions generated by the wind

park is comparatively low43. When calculating the Net Present Value (NPV) of the total

emission reduction potential, it is below 0.5% of the total investment costs of the wind

park. Thus, CDM does not indicate significant improvement in project performance

through the effect of the relevant CDM cash flows.

Regarding the assumption of considering CER prices constant at 6 USD/CER, it should

be noted that a certain amount of uncertainty surrounds the development of the market in

the post-Kyoto period. In this regard, remuneration from the sale of CO2 post-2012 is diffi-

cult to estimate with certainty. The positive impact of the CER on project cash flows is

evaluated in the financial analysis in Section 14. Under current CDM prices a registration

of the project is financially hardly feasible but it would establish Ethiopia as an example for

CDM activities in Africa.

43 Whilst the grid emission factor in Ethiopia is 16.92 tCO2/GWh, the grid emission factor of other African coun-tries like Egypt is estimated at 502 tCO2/GWh (Source: Project Design Document Final Draft for the ZafaranaIV Windfarm elaborated by Lahmeyer International GmbH of October, 2005).




14 Financial Analysis

14.1 Methodology & Main Assumptions

This financial analysis is intended to clarify whether or not the wind farm project is finan-

cially feasible for the described parameters. And, as the case may be, whether or not the

financial efficiency threshold can be reached and surpassed and, if so, how by which

means of technical and non-technical modifications.

The main difference between a financial analysis and an economic analysis is that, in the

financial analysis, the wind farm is viewed as an enterprise and in the economic analysis

the windfarm is evaluated from the point of view of the national economy of the country. In

the financial analysis, the windfarm operating company has to earn enough money, by

feeding electricity into the existing grid, to cover its operating costs, interest payments,

loan payments and distribution of dividends to equity investors. The object of considera-

tion is the commercial or microeconomic.

As in the economic analysis, four Scenarios have been considered in the financial analy-

sis of the Ashegoda Wind Park:

Scenario I - 86 Enercon E48 turbines

Scenario II 86 Enercon E53 turbines

Scenario III 86 Vestas V52 turbines

Scenario IV 86 Gamesa G58 turbines

The parameters and assumptions, upon which the financial assessments are calculated,

are elaborated below. These parameter and assumption values are used to calculate the

Base Case scenario. In addition to the base scenario, other scenarios are elaborated

within the scope of the sensitivity analysis.

Project s financial feasibility has been evaluated with the financial internal rate of return

(FIRR), the Debt Service Coverage Ratio (DSCR) and the Return on Equity (ROE). Fur-

ther, the levelized generation costs of the wind park in its four Scenarios have been com-

pared to the levelized generation costs of the Dire Dawa Diesel Power Plant and two

Ethiopian large hydropower plants (Halele-Werabesa and Finchaa).




14.1.1 Inflation Rate

According to EEPCO, local currency price inflation is assumed to increase 3.6 % annually

whereas foreign currency price inflation is assumed to increase 2.50 % annually. Starting

from year 2009, the mentioned inflation rates are assumed to decrease to 2 % and 3 % in

2015 for foreign and local currency, respectively.

As agreed with EEPCo, no inflation has been considered for the O&M costs, whereas a

2 % annual escalation has been considered on the current power tariff.

The O&M model of Lahmeyer International comprised for this feasibility study 50% addi-

tional charge on general O&M costs, thus to ensure a realistic calculation of arising ex-

penses.

14.1.2 Rate of Exchange

An exchange rate of ETB 8.61995 equal to USD 1.00 has been adopted as applicable for

the year 2006. The financial (as well as the economic model) have assumed the following

average interbank exchange rates for the period July 1st, 2005 to December 31st, 2005.

Table 14-1: Average exchange rates

Commodity currency Rate Average value for the period:

USD 1 July 1st, 2005 to December 31st, 2005

ETB 8.61995 July 1st, 2005 to December 31st, 2005

Euro ( ) 0.83036 July 1st, 2005 to December 31st, 2005

The future development of the USD /local currency exchange rate has been modelled on

the principle of Purchasing Power Parity (PPP).

The USD/ETB exchange rate was analysed as some quotes within the investment costs

were received in USD. As it is foreseen that the debt capital be raised in ETB as well as in

USD and that routine O&M costs are mainly to be paid in USD whereas operating reve-

nues are to be obtained in ETB. This item presents the project with exposure to foreign

exchange (FX) risk if the ETB devaluates prior to the contract prices being fixed. The de-

velopment of the USD/ETB exchange rate since the beginning of the year is presented

graphically below. The standard deviation of the ETB/USD exchange rate is 0.1317 in

relation to the median exchange rate of 8.7645.




Figure 14-1 shows the exchange rate USD/ETB development for last year 2005. The FX

has been relatively stable except for the month June, where a strong oscillation took

place.

Exchange Rate USD/ ETB

8.4

8.5

8.6

8.7

8.8

8.9

9

9.1

9.2

1/1

/20

05

1/3

1/2

00

5

3/2

/20

05

4/1

/20

05

5/1

/20

05

5/3

1/2

00

5

6/3

0/2

00

5

7/3

0/2

00

5

8/2

9/2

00

5

9/2

8/2

00

5

10

/28

/20

05

11

/27

/20

05

12

/27

/20

05

Date

FXR

ate

[US

D/E

TB

Figure 14-1 USD / ETB exchange rate development

14.1.3 Depreciation Rates

Depreciation is calculated using the straight-line method for all investment cost compo-nents.

14.1.4 Dividend Distribution

Annual dividends were calculated as the minimum of:

internal funds available for distribution (cash after debt service, taxes, and reserveaccount payments); and

net income.Dividends in any period were only distributed if the Debt Service Coverage Ratio (DSCR)

was greater than or equal to 1.20x, and if net income in that period was positive.




14.1.5 Applicable Taxes

Corporate Taxes

In accordance with information provided by EEPCo, the company in charge of operatingthe wind park, i.e. EEPCo, is not subject to corporate taxes. Consequently, the projectionsdo not foresee any related tax estimations.

Import Taxes

In accordance with present legislation on imports of material and equipment for HPPs, thecompany is not liable for related import duties.

Although the legislation is not expressively referring to wind parks, as these did not existat the time relevant legislation was approved, it can be assumed that the legal norms gov-erning HPPs will be applicable also for imports of material and equipment for wind parks.

Consequently, and for the purpose of the calculations presented in this feasibility study,no import taxes are considered.

14.1.6 Discount Rate (WACC)

14.1.6.1 Definition

The weighted average cost of capital (WACC) serves as representative for the financial

opportunity cost of capital (FOCC) to assess the financial viability of projects. The net

cash flows during the lifetime of the project are discounted at FOCC to show the project s

worth.

Although it is an accepted benchmark, it is important to understand that the WACC may

not fully reflect the FOCC in the market. A project may generate sufficient returns to allow

full recovery of all investment and operations and maintenance costs while still yielding a

small return on investment, this return may not be sufficient incentive for the owner to

make the original investment or to maintain the investment. Private foreign investors will

be looking for returns on equity that also include an allowance for risks, such as political

and economic. Private domestic investors will also have alternative investments, whether

they are in financial assets, other productive activities or areas such as real estate. Gov-

ernmental investment may be guided by whether the funds are fungible, by the real cost of

investment funds and the economic benefits of the project. If funds are fungible, they may

be more interested in investing in projects with higher returns, economic and/or financial.

Finally, projects with low returns are riskier to implement and strain the financial sustain-

ability of the corporate entity (public or private) charged with its operation and mainte-

nance. Consequently, it is important to keep these issues in mind when comparing the

FIRR of a project against a benchmark such as the WACC. These issues become particu-




larly important as the role of government in the supply and operation and maintenance of

infrastructure services changes and private sector participation becomes more prevalent.

14.1.6.2 Methodology

The discount rate to be normally used in financial benefit-cost analyses is the WACC. The

WACC represents the cost incurred by the entity in raising the capital necessary to im-

plement the project. Since most projects use several sources to raise capital and each of

these sources may seek a different return, the WACC represents a weighted average of

the different returns paid to these sources.

To avoid highly optimistic results and after clearance with the Ethiopian Ministry of Eco-

nomic Development and Co-operation, a discount rate of 10% instead of the WACC has

been used in all wind park calculations of this study.

The methodology used to calculate the WACC is as follows:

Step 1: A categorisation of financing components according to the Project Financing Plan

has been done. These components are domestic (local loan at an interest rate of 7 % p.a.

and local equity share) and foreign components (foreign loan at an interest rate of 5 %

and international equity).

Step 2: The Consultant estimated the cost of funds, since:

Government funds are not costless they might be applied to purposes other than the project, such as debt repayment or to alternative investments. For simplicity, the aver-age cost of government funds can be calculated by dividing total government debtservicing by total public debt. For the Ethiopian Wind Park Project governmental fundswere not expected.

In order to estimate the cost of equity capital, the capital asset pricing model (CAPM)has been used. CAPM describes the relationship between risk and expected re-turn and that is used in the pricing of risky securities.




Capital Asset Pricing Model (CAPM)

Rf = 4.00%= 1.50

iE =r expected return rate on a security Km = 10.00%

Rf rate of a "risk-free" investment, i.e. cash

Km return rate of the appropriate asset class iE = 13.00%

r = Rf + *( Km - Rf )

risk factor, according to the projects sensitiveness to market fluctuation

Figure 14-2: Capital Asset Pricing Model

The general idea of CAPM is that investors need to be compensated in two ways: time

value of money and risk. The time value of money is represented by the risk-free (rf)

rate in the formula and compensates the investors for placing money in any investment

over a period of time. The other half of the formula represents risk and calculates the

amount of compensation the investor needs for taking on additional risk. This is calculated

by taking a risk measure (beta) that compares the returns of the asset to the market over

a period of time and to the market premium (Km-rf).

Figure 14-3: Meaning of CAPM

Step 3: Adjustment for Taxation. If interest payments are deductible for taxation, applica-ble tax rates have to be adjusted to each component. In the Ethiopian case and asstated by EEPCo - no tax purposes have to be considered.

Step 4: Adjustment for Domestic Inflation. The estimated costs of borrowing and equitycapital have been adjusted for inflation to obtain the WACC in real terms. Domestic infla-tion rate (of 2 %) has been used for domestic loans and equity. Shadow Price Adjustment




does not apply in the financial analysis and hence, the standard conversion factor (SCF)of 1 has been considered for local currency expenditure, without reducing local costs.

Step 5: Application of the Minimum Rate Test, i.e. to review the real cost of capital foreach component. The rate for each component should be at least 4 %.

Step 6: The determination of the WACC has been done by applying the weighting per-centage to each component.

14.1.6.3 WACC Results

In this project, an equity share of 20 % has been assumed, that arise additional capital

from local and foreign banks. No thought has been given to additional funding options as

well as governmental grants. Differing nominal returns on each source of capital are as-

sumed, including an expected return of 10 % on equity to the shareholders. In this calcula-

tion the corporate tax rate for EEPCo was set to zero.

Loan 1 Loan 2 FundEquityshare Loan 1 Loan 2 Fund

20%

A [USD x10^ 11.425 0 0 19.266 65.639 0 0 96.330

B [%] 11.86% 0.00% 0.00% 20.00% 68.14% 0.00% 0.00% 100%

C [%] 7.00% 0.00% 0.00% 13.00% 5.00% 0.00% 0.00%

D 0.00% [%] 0.00% 0.00% - - 0.00% 0.00% -

E [%] 7.00% 0.00% 0.00% 13.00% 5.00% 0.00% 0.00%

F 1.00 7.00% 0.00% 0.00% 13.00% 5.00% 0.00% 0.00%

G 2.00% [%] 2.00% 2.00% 2.00% 2.00% - - -

H [%] 4.90% -1.96% -1.96% 10.78% 5.00% 0.00% 0.00%

I 4% [%] 4.90% -1.96% -1.96% 10.78% 5.00% 0.00% 0.00%

J [%] 0.58% 0.00% 0.00% 2.16% 3.41% 0.00% 0.00% 6.15%

K 6.15%

Total

Min. rate test

Weighted component of WACC

Domestic Inflation rate

Real cost

Tax-adjusted nominal costs

SCF

Foreign

Computing WACC [specific model]

Amount

Weighting

Nominal cost

Taxation (corporate tax rate)

WACC

Domenstic

Table 14-2: WACC for the Scenario I (Enercon E48)

The WACC in real terms amounts to 6.15 %. This is the discount rate to be normally used

in the financial benefit-cost analysis of this particular project as representative for the

FOCC. (A list of the WACCs for the different wind park scenarios is presented in Table

14-11).

The resulting WACC differs from the discount rate of 10 % considered by the Ethiopian

Ministry of Economic Development and Co-operation as appropriate for the coun-




try at present. This is due to the relative low interest rates considered for the debt

service of the wind park.

To avoid highly optimistic results, the discount rate of 10% instead of the WACC (6.15%)

has been used in all wind park calculations of this study.

14.1.7 Project s Milestones

Main project s milestones are specified in Table 14-3. The project construction timeframe

is planned to comprise 12 months. It includes the construction of infrastructure, erection,

installation, and commissioning of the wind turbines. Using the financial model all potential

financial fees and interests during construction incurred prior to commissioning have been

calculated.

Since more than 10 wind turbines are planned in each wind park, the construction will take

place at least in 3 sets (shipments)44.

Table 14-3: Project s milestones

Milestone Ashegoda Wind Park

Financial Close 01.10.2006

Down Payment 01.10.2006

First Shipment 01.01.2007

Last Taking over Certificate 01.10.2007

Starting of wind park operation 01.10.2007

Number of Wind Turbines 86

Major Overhaul / Decommissioning Year 11 and year 21, respectively

Period of Analysis 21 years

44 The financial model considers that the first sets installed will be operating previous to Taking over Certifi-cate, so that they start generating revenues. Nevertheless and previous to Financial Close at the latest, theenergy production of the turbines during the first months of operation needs to be calculated into more detail.




14.1.8 Investment Costs

Table 14-4 shows the total investment costs used for each wind park Scenario in the fi-

nancial analysis. A detail of the different investment cost components is detailed in Sec-

tion 10.

Table 14-4: Investment costs (financial value)

Enercon E48 WP Investment Costs Financial Value [USD] [%]

Investment in Foreign Currency 84,905,398 88.14%

Investment in Local Currency 11,425,069 11.86%

Total Investment 96,330,468 100.00%

Enercon E53 WP Investment Costs Financial Value [USD] [%]




Vestas V52 WP Investment Costs Financial Value [USD] [%]




Gamesa G58 WP Investment Costs Financial Value [USD] [%]







14.1.9 Operation and Maintenance Costs

The estimates for operation and maintenance (O&M) have been created based on the

experience of Lahmeyer International in Due Diligence projects, in the country (through its

hydro power projects) and calculated through a model developed by our engineers.

The O&M costs include repairs, maintenance, spare parts, insurance costs, personnel

costs for wind park management and maintenance and electricity consumption.

The range of the costs is determined by the changing lifetime of the components and their

prices which can not be predicted exactly. Furthermore the way of operating the wind

parks can have an influence on the expected costs which can also be only estimated for

future times. In addition, the size of the machines and the operating time under full load is

expected to have an influence on the maintenance costs. The LI-O&M cost estimation

model takes these uncertainties into account, giving an expected range of maintenance

costs (expected average, expected upper bound and expected lower bound). For the cal-

culation the expected average has been considered. The results of applying the LI-O&M

Model to the two wind parks are shown in detail in Section 10.

Table 14-5 gives an overview of the O&M costs considered in the different wind park sce-

narios.

Table 14-5: O&M costs

WT TypeEnercon E48Scenario I

Enercon E53Scenario II

Vestas V52ScenarioIII

Gamesa G58ScenarioIV

O&M Annual Costs (ETB)

29,903,145 30,059,379 32,502,485 32,635,2842

O&M Annual Costs (USD)

3,469,064 3,770,613 3,786,0193,469,064

During the elaboration of this study, the questions about which tasks should stay with the

manufacturer and whether constant presence of the manufacturer in Ethiopia is neces-

sary, has been raised. According to Consultant s experience, manufactures have to guar-

antee that in wind parks with an installed capacity of approximately 50 MW, two experts

(usually, one electrical engineer and one mechanical engineer) will be constantly present

at the wind park. Additionally, a team of four local experts has to be established for main-

tenance tasks and a crane has to be available, at least once per year, to realise operation

revisions. Further, condition monitoring should be realised by independent engineers in

order to plan the repairs.




The O&M costs have taken insurance costs into consideration in the form of an insurance

following commissioning (Business Interruption Insurance). This has been calculated as

an annual 12.0 % of the wind turbine price.

Additionally, a major overhaul of all equipment has been assumed to take place between

the 10th and 11th years of operation in an amount of 5 % of total investment costs.

Finally, wind farm decommissioning costs in operation year 21 have been considered in

an amount of 1 % of total investment costs.

14.1.10 Land Lease Costs

As indicated by EEPCo, no annual land lease charge was used in the analysis.

14.1.11 Costs for Mitigation Measures

The costs for measures of mitigation required by environmental regulations in Ethiopia is

estimated to be zero, because the environmental impacts of wind farms concern mainly

the visual impact on the landscape and the impact on birds. In general, these impacts are

minor. There have been problems at a few sites in the world with rather significant killings

of birds and bats. But even examples for local resistance to the setting up of wind farms

are known. But, overall, surveys tend to show that the population living in the vicinity of

wind farms has a more positive attitude to wind farms than persons who do not, and sur-

veys do not show a negative impact on real estate prices in the local area. Consequently,

the cost of environmental damage from wind farms is set to zero in this feasibility study.

An amount of USD 244,870.1 for mitigation measures (non-environmental) has been

considered in the investment costs. A description of this cost is contained in Section 9.4 of

the Part I.

14.1.12 Project Financing Structure

Based on EEPCO s information, a possible financing structure with which the required

funds could be raised, is under development. As such, the financing structure is to be in-

terpreted as preliminary only and subject to optimisation.

A debt / equity ratio of 80 % has been assumed, and it has furthermore been assumed

that the debt will be raised in US Dollar and local currency. The total equity shares will be

held by EEPCOo and (an-)other international organisation(s), whereas donor and local

loans will provide debt financing. The local loan shall cover local investment costs,

whereas the international/donor s loan is calculated to cover international investment

costs. The proposed project financing structure is summarised numerically in Table 14-6.




Table 14-6: Financing structure and disbursement

Ashegoda Wind Park Scenarios

Item Scenario I: En-

ercon E48

Scenario II:

Enercon E53

Scenario III:

Vestas V52

Scenario IV:

Gamesa G58

Debt / Equity

Ratio80 /20 80 /20 80 /20 80 /20

Investment

Require-

ments (*)

96.33 Mio.USD 101.51Mio. USD 102.54Mio. USD 100.86Mio. USD

International

Donor Loan65.64 Mio. USD 69.78 Mio. USD 70.61 Mio. USD 69.26 Mio. USD

Local Loan 11.43 Mio. USD 11.43 Mio. USD 11.43 Mio. USD 11.43 Mio. USD

Equity 19.27 Mio. USD 20.30 Mio. USD 20.51 Mio. USD 20.17 Mio. USD

(*) Interest during Construction (IDC) has been considered to be zero, whereas commissioning

fees are considered to be fix at 100.000 USD.

14.1.12.1 Equity Finance

Total equity required for both wind parks has been estimated at 20 % of total funding re-

quirements. Equity funds will be used to finance both foreign and local components.

Furthermore, additional equity will also cover the initial working capital requirements. Eq-

uity investments will be finalised prior to Financial Close or similar legal procedure. Once

the equity is drawn into the Projects, the debt facilities will be utilised.

14.1.12.2 Debt Finance

The assumed terms and conditions for the debt finance are based on the terms and condi-

tions obtained in Ethiopia during the first mission of this assignment, the Consultant s ex-

perience in financial analysis of wind parks in developing countries and the current local

loan conditions.

The ultimate financing conditions at which debt can be raised will depend upon several

factors, including for example the general environment in the debt markets, local political

conditions as well as the Lender s interpretation of the project s risk profile.




The debt will be provided by foreign or domestic financial institutions, while local activities

will be covered by local loans. The debt will be utilised to finance the project costs includ-

ing the development costs, financing fees, operation cost during the construction period,

IDC and contingency for the Project.

In the financial analysis, the debt structure for both wind parks is as follows:

Tranche 1:

Foreign currency investment portion supported by an international donor loan

and/or an Export Credit Guarantee

Availability periodfrom financial closing and/or similar

financial legal procedure

Final maturityapproximately 15 years from finan-

cial close

Repayment profilethirty sculpted semi-annual instal-

ments

Base ratea fixed rate CIRR of 4.750 % p.a.

spread or margin of 0.250 % p.a.

Grace period 2 semi-years (1 year)

Tranche 2:

Local currency investment portion supported by a domestic financial institution

Availability periodfrom financial closing and/or similar

financial legal procedure

Final maturityapprox. 10 years from financial close

and/or similar legal procedure

Repayment profilefourteen sculpted semi-annual in-

stalments;

Base rate fix rate of 7 %

Grace period 2 semi-years (1 year)

14.1.13 CDM Up-Front & Administrative Costs

As specified in Section 13, in order to register a project as a CDM activity, the project par-

ticipants have to incur mainly in CDM up-front and CDM monitoring costs. CDM Up-front




Costs for large scale project are usually estimated at USD 100,000. In our analysis, we

have assumed that these costs will be subsidised or supported by international donors,

and hence, to be zero.

An amount of USD 5,000 / per annum has been assumed for the ongoing CDM Adminis-

trative Costs. CDM Administrative Costs are applicable only in years where CDM has

been taken into consideration, i.e. up to and including 2028 (since the considered credit-

ing period has been 7x3 years45).

14.1.14 Financial Benefits

Financial benefits have been defined as (i) electricity sales and (ii) sales of Certified

Emission Reduction Credits (CERs).

14.1.14.1 Electricity Sales

Financial revenues will be mainly generated through the sell of electricity. According to

EEPCo, sales tariffs are as low as 0.06 USD/kWh on average and a 10 % tariff increase

each 5 years is settled by law. Thus and as indicated by EEPCo, the Consultant has as-

sumed an annual tariff increase of 2 %.

Electricity sales per annum are defined as the net amount of electricity generated and

connected to the grid multiplied by the electricity price (as per USDct/kWh) as indicated in

the following formula:

Wind Power generation (Base Case: P75) x power tariff

[kWh/year] x [6 USDct/kWh]

In the base case, the estimated wind power generation (P75) for each wind park scenariois as follows:

45 2x7+1x6 years when considering an operational period of 20 years.




Table 14-7: Annual energy generation

WT Type Enercon E48 Enercon E53 Vestas V52 Gamesa G58

Wind park output(P75) in KWh

197,392,000

227,278,000 198,771,000 239,804,000

ASHEGODA SITE

The energy potential as measured by gross production is given in the Base Case

using the 75 % probability of exceedance.

14.1.14.2 CDM Revenues

Additional to electricity sales, the potential to generate financial benefits through the sell of

Certified Emission Reduction (CER) credits of the Clean Development Mechanism has

been analysed. The influence of the Clean Development Mechanism (CDM) has been

isolated for the with CDM and without CDM scenarios. The with CDM scenario in-

cludes the additional Certified Emission Reduction (CER) revenues as cash inflows, as

well as the CDM Up-Front Costs and Administrative Costs as cash outflows.

CER PriceGiven the projects specific boundary conditions, and taking into consideration the rela-

tively small size of the project, the estimated sales price for the CER has been assumed

to be USD 6 / t CO2. Note, that this is the price to sell certificates. In the economic analy-

sis we estimate the penalties to polute which include costs for environmental impact al-

leviations. Details on CDM assessment are included in Section 13.

14.2 Results: Financial Analysis

LI has exemplary assessed the future financial performance of Ashegoda Wind Farm with

the aim to analyse their financial viability. To reach this objective, a comprehensive finan-

cial model has been developed. The model determines, on the one hand, the required

tariff to achieve a 10 % Return on Equity (ROE). On the other hand, it calculates the IRR

when applying for 20 years the existing average electricity tariff of 0.06 USD/kWh with

an escalation of 2 % per annum. Furthermore, the model has been structured in semi-

periods, so that in case the wind conditions would have been strongly different from sum-

mer than in winter months, a different tariff for winter and summer could be calculated.




In the model, all the conventional financial statements have been developed, giving an

overview of costs and revenues, assets, and liquidity. The values are based upon all rele-

vant financial aspects, such as investment costs, capital structure, fees accrued during

construction, operational costs and distribution of dividends.

14.2.1 Major Financial Indicators

This Section deals with financial key parameters characterising the Ashegoda Wind Farm.

Short introductions to the indicators provide a better understanding of the subsequently

presented financial results. Results are depicted in the summary Table 14-11. A final sen-

sitivity analysis on the best wind park scenario illustrates the impact of chosen parameters

to the financial feasibility of the project and identifies leverage effects (See Section 14.3).

14.2.1.1 Net Present Value

The net present value (NPV) of an investment has been defined as the present

(discounted) value of future cash inflows minus the present value of the invest-

ment and any associated future cash outflows. The NPV of the Ashegoda Wind

Park has been calculated using the WACC. Results for all wind park scenarios are

indicated in Table 14-11.

A positive NPV indicates that the projects are justified in an economic sense and vice

verse for a negative NPV. The NPV of the cash flow from operations (for both cases: Base

Case with CDM and Base Case without CDM) is not higher than zero. Thus, the project is

not financially feasible.

For each Scenario the produced NPV is:

Scenario I: -8.19 Mio. USD

Scenario II: +3.99 Mio. USD

Scenario III: -15.95 Mio. USD

Scenario IV: +9.37 Mio. USD




14.2.1.2 Financial IRR (FIRR) and Return on Equity (ROE)

The Financial Internal Rate of Return (FIRR) is an indicator to measure the financial return

on investment of an income generation project and is used to make the investment deci-

sion.

The FIRR has been obtained by equating the present value of investment costs (as cash

out-flows) and the present value of net incomes (as cash in-flows) as shows below.

I0= Initial Investment

B= Benefits

r= Discount Rate

m= Period

The FIRR of Ashegoda wind farm has been calculated with the impact of CDM andwithout CDM. (See results in Table 14-11.).

The FIRR with and without CDM for each Scenario is as follows:

Scenario I 8.73 % (without CDM) 8.75 % (with CDM)

Scenario II 10.52 % (without CDM) 10.54 % (with CDM)

Scenario III 7.67 % (without CDM) 7.69 % (with CDM)

Scenario IV 11.26 % (without CDM) 11.28 % (with CDM)

14.2.1.3 Return on Equity (ROE)

The ROE measures the profitability of a project, calculated as net income divided by

Shareholders Equity. Essentially, ROE reveals how much profit a project generates for the

capital shareholders, which have invested in it. At least it turns out that a project revenue

cannot grow faster than its current ROE without raising additional cash. The ROE of the

Ashegoda Project is illustrated in Table 14-11.




ROE is very high, concretely:

Scenario I 13.02 % (with CDM)

Scenario II 18.45 % (with CDM)

Scenario III 10.17 % (with CDM)

Scenario IV 20.79 % (with CDM)

Due to a very favorable energy generation in the location of the wind park, the absence of

taxes on revenues and a favorable tariff for wind energy, the produced ROE is compara-

tively high.

14.2.1.4 Debt Service Coverage Ratio (DSCR)

The long-term debt-service coverage has been examined to ensure that all long-term

loans and the related financial expenses can be paid in the yearly instalments without

depriving the firm of needed funds. The Debt Service Coverage Ratio (DSCR) defines the

capability of the Ashegoda Wind Farm operating company to repay principal and interest

payments on debt. It is stated as the ratio between cash available to service debt and total

debt service.

Conventionally, the lender will require that a minimum DSCR of 1.20x be upheld during

the amortization period, to ensure that the project can meet its upcoming debt commit-

ments with sufficient buffer. Table 14-11 shows the minimum DSCR of the Ashegoda

Farm, which in the Scenario IV (Gamesa G58) is high enough at 1.20x.

For the remaining scenarios below 1.20x, further improvements in the results could be

achieved through, for example:

a slightly higher sales tariff;

a further revision of the debt conditions;

a further revision of the local and foreign investment costs. (Investment costs arebased on initial supplier offers and/or on the Consultants experience in similar pro-jects, so that they do not include any discount as a result of price negotiations).

The sensitivity analysis carried out in Section 14.3 reveals the extent to which a variation

in certain project parameters would improve the economics of the project.




14.2.1.5 Levelized Costs

Similar to the internal rate of return calculation, the levelized costs of energy are calcu-

lated by searching for a tariff for electricity, with which the net present value turns out as

zero and the internal rate of return equals the applied discount rate (World Bank 2006,

Minister of Natural Resources Canada, 2005, p. 65). In this calculation only the annual

project costs such as operation and maintenance and debt service are included, whereas

the inflation rate is set to zero. The results of the application of the base case assump-

tions for the levelized cost calculation of wind energy production is displayed in Table

14-8:

Table 14-8: Levelized costs for wind energy


Levelized Generation Costs (USDc/kWh)

6.44 5.82 6.84 5.60

ASHEGODA Wind Power Project

Comparing the levelized costs of the different Scenarios to the current power tariff

(6 USDc/kWh), it can be seen that results are congruent with the positive NPV obtained in

Scenario II (Enercon E53) and Scenario IV (Gamesa G58). In both cases, levelized costs

are below the current tariff. Therefore, they have been compared to the current levelized

costs of generating power with a DPP and with HPPs in Ethiopia.

For the comparison the following plant types are used:

DPPs: Dire Dawa DPP with an annual production of 249,660 MWh,

HPPs: The Halele Werabese HPP and the Finchaa HPP with average annualproductions of 2,030,000 MWh and of 615,572 MWh, respectively.

The assumptions regarding the Dire Dawa DPP have been described in detail in Section

12.2.6 The main assumptions are recapitulated in Table 14-9 together with the resulting

levelized electricity generation unit cost for the DPP.

The assumptions concerning the Halele-Werabesa HPP are based on two feasibility stud-

ies, namely the Feasibility Study of Weles, Zhemoga-Yeda and Halele-Werabesa Hydro-

power Project46 and the Feasiblity Study of Halele-Werabesa Stage II Hydropower Pro-

46 "Feasibility Study of Weles, Zhemoga-Yeda and Halele-Werabesa Hydropower Project , Lahmeyer Interna-tional GmbH in association with Mid-day Consulting Engineers and Tropic Consulting Engineers, June 2005.




ject47 . The latter correlates to the first study and states as financing terms the following

parameters:

Loan repayment period: 20 years from commencement of operation

Interest rate for foreign currency: 10,0 %

Interest rate for local currency: 8,0 %Share of local currency on total investment: 26 %

Based on these assumptions and the technical and economic plant data shown in Table

14-9, a levelized electricity generation unit cost of 3.0 USc/KWh for the complete project

(Phase I & II) is calculated. There is a probability that the current costs are higher than 3.0

USc/KWh, due to a low capacity factor of the HPP and decreased capacity of the reservoir

due to siltation. Note that the estimation of the levelized unit cost can be problematic in

HPP projects and is highly dependent on the assumptions made, since the cost-intensive

components belonging to the civil engineering lot have a lifetime much higher (40 years or

longer) than the usual depreciation period. Further, HPP projects often are realised in

various phases or extended during the operation period.

Concerning the Finchaa HPP, the assumptions for the calculation are based on data pro-

vided by EEPCO. A summary of the main assumptions and results are provided in Table

14-9. Since for the investment cost of the plant only inconsistent data was available, a

typical specific investment cost of 1,300 USD/KWh for a medium sized HPP was used as

a basis for the calculations (figure marked with *). Note that the specific investment cost

for HPP can vary within a wide range depending on factors such as length and height of

dam, geological conditions etc. For the financing terms, the same conditions as for the

Halele-Werabesa HPP were assumed.

47 Feasibility Study of Halele-Werabesa Stage II Hydropower Project , Lahmeyer International GmbH,(2004)




Table 14-9: Assumptions for the DPP & HPP levelized cost calculation

Dire Dawa Halele-Werabesa

Finchaa

Plant Type DPP HPP HPP

Installed Power [MW] 38 422 134

Annual Production [GWh] 250 2030 616

Total Investment Cost [$] 31,100,000 508,800,000 174,200,000

specific Investment Cost [$/kW] 818 1,206 1,300*

Depreciation Period 15 20 20

Annual Capital Costs [$/a] 4,087,964 53,893,508 20,461,467

Annual O&M (non-fuel) [$/a] 1,597,960 7,632,000 534,050

Annual Fuel Cost [$/a] 26,463,960

Total Cost per year [$/a] 32,149,884 61,525,508 20,995,516

Electricity Gen. Cost [$c/kWh] 12.9 3.0 3.4

As seen in the above table, the HPP plants supply relatively cheap electricity in spite of

the high specific investment costs. The Dire Dawa DPP has much higher operating costs,

mainly consisting of the expenses for fuel. As such, the electricity generation cost for the

DPP considerably exceeds all other options at current fuel prices, as can be seen in Table

14-9.

Table 14-10: Wind, diesel, hydropower levelized costs

DPP HPP HPP

WT TypeEnercon

E48Enercon

E53Vestas

V52Gamesa

G58Dire Dawa

HaleleWerabese

Finchaa

Levelized Generation Costs (USDc/kWh)

6.44 5.82 6.84 5.60 12.90 3.00 3.40

Ashegoda Wind Power Project




14.2.2 Summary of Key Financial Parameters

A summary of the key financial parameters is included in the following table:

Table 14-11: Summary of key financial parameters

ASHEGODA SITE


WACC (%) 6.15% 6.15% 6.15% 6.15%

FIRRw/o CDM (%) 8.73% 10.52% 7.67% 11.26%

FIRRwith CDM (%) 8.75% 10.54% 7.69% 11.28%

Financial NPV(mill. USD)

-8.19 +3.99 -15.95 +9.37

ROE (%) 13.02 % 18.45 % 10.17 % 20.79 %

Min. DSCR 0.81x 1.08x 0.69x 1.22x

Financial Feasible (YES/NO)

NO YES NO YES

Power tariff to reach financial feasibility* (USDc/kWh)

6.44 5.82 6.84 5.60

(*) Financial feasibility has been defined with a moderate 10 % FIRR.

14.2.3 Conclusions: Financial Analysis

The summary of the key financial parameters in Table 14-11 indicates that the project

generates a high ROE in all Scenarios. So that the project is very attractive for equity in-

vestors. This is mainly due to high energy generation of the wind park, absence of taxes,

favorable loan conditions and relative high power tariffs (6USDc/kWh with an escalation of

2 % per annum).

In two Scenarios (Scenario II Enercon E53 - and Scenario IV Gamesa G58 -) the NPV

of the cash flow from operations is clearly higher than zero, so that the project is finan-

cially viable. On the contrary, the project struggles to generate sufficient cash to service

its debt commitments in Scenario I (Enercon E48) and Scenario III (Vestas V52).




Similarly, the FIRR (with and without CDM) is above their target value of 10 % in two Sce-

narios (Scenario II Enercon E53 - and Scenario IV Gamesa G58 -) and it is below

their target values in Scenario I (Enercon E48) and Scenario III (Vestas V52).

Due to a relatively low emission grid factor (as assessed in Section 13), the positive im-

pact of CDM is rather moderate. CDM-up front costs have not been included in the in-

vestment costs, which could have made the results of including CDM even worse.

Furthermore, the DSCR stays well above the desired mark of 1.20x on one Scenario

(Gamesa G58), whereas in the Scenario II (Enercon E53) it approaches the target with

1.08x. In the other Scenarios the DSCR is below the desired mark.

Possible approaches to improve the results could include, for example:

to settle higher sales tariff; Tariffs in order to reach a 10 % FIRR have beencalculated, resulting in tariffs higher than 8.5 USDc/kWh;

local investment costs could be further revised.

The sensitivity analysis (in Section 14.3) reveals the extent to which a variation in certain

project parameters would improve or worsen the financial viability of the project. The sen-

sitivity analysis has been done for the wind park scenario that presents the best financial

results, i.e., the Gamesa G58 wind turbines.




14.2.4 Financial Statements

The results of the financial model have been presented in pro-forma financial statements.

Annual asset and liquidity positions are reflected in the Balance Sheet, Profit and Loss

Accounts and Cash Flow Statement, giving an overview of the profitability and capital

structure of the project. These basic financial statements are described briefly in turn.

14.2.4.1 Cash-flow projections

To assist with the financial planning and the assurance of liquidity of the wind farm operat-

ing company, a cash flow schedule, showing all sources and applications of funds, has

been prepared. The operating cash-flow for the Ashegoda wind park in all four scenarios

are shown in Annex E.

14.2.4.2 Profit and Loss Accounts

The Profit and Loss Account is used to compute the net earnings or deficit of the projectarising each year. The Profit and Loss Account shows the revenues, operational ex-penses, financial activities and taxes and dividend distribution of the project, shown inAnnex F for the different scenarios.

14.2.4.3 Balance

The Balance Sheets show the accumulated assets the wealth of the project and howthis wealth is financed And are present in Annex F.

14.3 Sensitivity Analysis

The purpose of the sensitivity testing is to establish which project parameters have the

potential to alter the financial feasibility of the project by the greatest amount. By system-

atically altering individual parameters and recording the influence on the project evaluation

criteria, those parameters with the greatest potential to cause the economics of the project

to deviate from its expected value can be isolated. This provides a transparent overview of

the project s risk profile, assists project planning, and helps prepare risk mitigation strate-

gies.




14.3.1 Methodology

Where applicable, parameters listed in Section 14.1 had their expected values (as used in

the Base Case with CDM calculations) increased and decreased by +5 %, -5 %, and

+10 %, -10 % respectively. This has been done for one parameter at a time, so as to iso-

late the impact of that variable on project feasibility. The resulting project IRR has been

recorded in table form, and a so-called Spider diagram was derived to provide a graphical

representation.

Complimentary to this, the respective Sensitivity Indicators (SI) and Switching Values (SV)

have been derived and presented.

The definitions of the SI and SV are as follows:

SI: represents the percentage change in project NPV as a result of a 1 % in-crease in the parameter; and

SV: represents the percentage change in the parameter value necessary todrive project NPV down to zero.

14.3.2 Sensitivity Variables

Sensitivity testing was conducted on the wind park scenario of Gamesa Wind Turbinesand for the following variables:

Investment costs

Energy generation

Routine O&M

Sales Tariff

CO2 Contract Price

14.3.3 Results: Sensitivity Testing

Results of the sensitivity analysis on the financial model with the Gamesa wind turbinesare presented on the following table:




Table 14-12: Financial sensitivity testing results

Change -10% -5% Base +5% +10%

Sensitivity Items

Investment Costs 12.88% 12.05% 11.28% 10.58% 9.93%

Generation (GWh) 9.33% 10.32% 11.28% 12.23% 13.15%

Routine O&M 11.72% 11.50% 11.28% 11.07% 10.85%

Sales Tariff 9.34% 10.32% 11.28% 12.23% 13.15%

CO2 Contract Price 11.28% 11.28% 11.28% 11.29% 11.29%

Project IRR (%)

Each sensitivity variable was altered independently, so as to ensure that its potential im-

pact on the project s IRR is isolated. The results are provided graphically below for clarifi-

cation.

Spider Diagram

1.00%

3.00%

5.00%

7.00%

9.00%

11.00%

13.00%

15.00%

-10% -5% Base +5% +10%

FIR

R

Investment Costs Generation (GWh) Routine O&M Sales Tariff CO2 Contract Price

Figure 14-4: Spider diagram showing results of sensitivity testing




The spider diagram is useful to provide a different perspective of the sensitivity testing

results. The x-axis represents the change in the value of the sensitivity variable, while the

y-axis represents the subsequent change in project IRR.

The diagram can best be interpreted by analysing the slope of the individual functions.

Where a small change in the value of the sensitivity variable causes a large change in

project IRR, the gradient of the curve will be steep. Where the change in project IRR, fol-

lowing a change in the value of the variable, is small or negligible, the gradient of the func-

tion will be flat. Hence, it is concluded that those variables having the greatest potential

effect on project IRR are those which have the steepest slopes, positive or negative.

Figure 14-4 indicates that this includes the parameters sales tariff, the energy generation

and investment costs. This is confirmed by the sensitivity indicators and switching values,

as it is elaborated below.

It can also be seen in the spider diagram that the sales tariff and wind power generation

have an identical potential influence on the project IRR. This underlines the fact that both

increase equally the volume of sales.




14.3.3.1 Sensitivity Measurement

As Table 14-12 shows, the parameters with the greatest potential influence on project IRR

are the energy generation, sales tariff and the investment costs, whereas project IRR is

the most immune against fluctuations in the CO2 price. These findings are confirmed by

the Sensitivity Indicators and Switching Values, as illustrated in Table 14-13.

Table 14-13: Sensitivity indicators & switching values

SV SI

Investment Costs 9.56% -10.46%

Generation 6.74% 14.83%

Sales Tariff 6.74% 14.83%

CO2 Contract Price 100% 0.00%

Sensitivity MeasureSensitivity Item

The following remarks should be taken into consideration when referring to Table 14-13:

Since the project NPV for the Base Case with CDM is higher than zero an increasein the basic costs by 1 % decreases project NPV. Accordingly, this results in anegative SI.

For the SV figures and since the NPV is being driven down to zero, the SV has apositive value since items have to be reduced to reach the zero NPV.

The indicators in Table 14-13 again highlight the significance of the investment costs,

sales tariff and energy generation on project IRR, as expected.

Unfavorable fluctuations in the CO2 price or routine O&M would have a less dramatic im-

pact on the project. The variable having the least impact on project returns is the CO2

price, as evidenced by its low Sensitivity Indicator. This means that even large deviation

from its expected value will not bring the project NPV to zero. This is due to, as explained

in Section 13, a low emissions grid factor.




15 Framework Analysis for Wind Energy in Ethiopia

15.1 Financing Options

As presented in Section 14, the project will be financed 80 % by debt and 20 % by equity.

EEPCo will hold 100 % of the equity shares while the debt will be raised in USD and local

currency.

Various sources of financing can be explored for the project, and this includes the follow-

ing:

Multilateral institutions. Many of these institutions have introduced programs andinitiatives that provide various forms of financing especially in the seed stage.Lending organisation such as the World Bank, AfDB, etc have traditionally beensource of capital for large renewable energy projects. The involvement of thesemultilateral lenders makes the participation of commercial lenders and other pri-vate investors more likely.

Available ODA (Overseas Development Assistance) instruments could be sur-veyed in order to optimise financing of the project. These instruments includegrants, soft loans, promotional credits and other financial supporting instrumentsas guarantees. This seems to be the most likely opportunity to finance the project.

Private equity funds. A number of private equity funds have been founded over the last 10 years in co-operation with multilateral institutions that focus on the devel-opment of renewable energy. Many of these funds primarily invest in companiesthat are engaged in the development of new technologies but they are also inter-ested in stand alone projects such as wind power projects that meet their respec-tive investment criteria. Private equity funds have been specified by EEPCo, al-though EEPCo already informed that these are still subject to approval.

Export credit agencies (ECAs). ECAs promote the exports of countries throughcredits and political risk insurance to the buyers of goods. The participation ofECAs is often crucial for commercial lenders to loan money to energy projects incountries with high degree of economic and political instability. With the growingcommercialisation of wind power, many producers of wind turbines use ECA topromote the exports of machinery. This alternative to finance the project depend-ent on the choosed turbine manufacturer.

Similarly, international promotional schemes such as carbon finance (Clean Development

Mechanism, EU-Emissions Trading Scheme) and GEF could also be explored.




15.2 Regulatory and Legal Framework

The government should establish enabling frameworks to reduce the transaction costs inthe preparation, implementation and operation of the wind park.

These frameworks could be categorised into:

Authorisation procedures. The authorisation process involves different agencies at the central and local levels, and is often time consuming. The wind park developerneeds permits and licenses related to environmental assessment, construction,project site use, use of protected areas, etc. The government can increase theproductivity of the administrative process by establishing a one-stop agency for the approval process of the wind park project. This agency would co-ordinate the ap-proval activities of all involved public agencies.

Access conditions to the grid. The existing grid codes may pose obstacles to theintegration of the wind park. It may be necessary to introduce technical standardsthat facilitate intermittent energy producers access to the grid.

Access conditions to the power market. The current power market rules and regu-lations could also impose high entry costs to the wind park project developer. This could be reduced by introducing regulations that standardise connection chargesand fees for market access, use of system charges (payments for balancingpower, back-up power, wheeling charge), and metering services.




16 Conclusion and Recommendations

Due to the promising wind conditions at the project site at Ashegoda and the available

open space at the large proposed area, a realisation of the project in general is feasible.

Considering the need of Ethiopia to diversify power generation currently highly reliable on

hydropower, the prevailing energy crisis due to decreasing rainfalls and the increasing

power demand, a short term supply solution has to be implemented.

In the short-run it is necessary to increase current power generation mix, to cover increas-

ing and unmet power demand and to avoid dependence on fuel imports. Wind and diesel

power generation are the two fast-track implementation alternatives considered by

EEPCo.

Between both, the results show that the implementation of the wind park project is the

most economic and financial feasible power generation alternative to be implemented in

the short run in Ethiopia

The calculated net energy output at the P75 level (as a common value for financing wind

parks by international banks) for the wind park Ashegoda is in range of 197,392 MWh/y to

239,804 MWh/y at hub heights between 57 m and 60 m. These values can be considered

as conservative. The capacity factors of 31.0% to 37.7 % surpass average values in com-

parison with other international projects, even when taking into account the reduced per-

formance of the wind turbines due to the low air density (at the altitude of 2,000 m), be-

cause of the high average wind speed on site.Depending on the wind turbine type consid-

ered, the Internal Rate of Return oscillates between 16.64 % and 11.00 %. Additional

benefits can be generated through the avoidance of CO2 emissions, estimated at

1,359,387 CO2 tones for the diesel power plant.

The Ashegoda Wind Park implementation is also recommendable from the point of view of

the Capacity Credit. For Ashegoda a comparatively high Capacity Credit of 38 %

(26.46 MW of hydropower capacity) can be generated.

Even higher are the Capacity Credit results when combining the Ashegoda Wind Park

with a large scale implementation of wind energy in Ethiopia.

The realisation of the project could have other secondary positive effects for Ethiopia.

Since the Ashegoda Wind Park fulfils the conditions to be registered as a CDM-activity,

potential benefits of a CDM registration could be generated, such as the strengthening of

the institutional framework regarding CDM and the DNA, the facilitation of know-how

transfer, the constitution of a path for the registration of further CDM activities in Ethiopia,

and finally, since the project is the first of this kind in the region, it would position Ethiopia

as an example for other CDM activities in Africa.




Concerning the technical part of Ashegoda wind park, the selected turbine types of the

manufacurers VESTAS, GAMESA and ENERCON are suitable for the planned project

and can be considered as well engineered and proven technologies. For delivery, installa-

tion and commissioning of the turbines, a first Expression of Interest from the manufac-

turer ENERCON-India is available, which shows the interest in general for wind power

projects in Ethiopia, by a foreign turbine manufacturer.

As stated in this report, some barriers were identified concerning the project implementa-

tion:

Transportation of the large number of turbines will take several months. We rec-ommend the development of a detailed transportation concept in advance.

Several manufacturers have been requested for an EoI by the consultant, exceptof Enercon India most answers are still pending or under clarification of details.

One of the main risks lies in the time frame for project construction (twelvemonths) in 2007. An extension of the construction phase seems difficult due to theextreme time pressure from the Ethiopian Authorities. However, the timely realisa-tion of the construction works is possible in the event that a turbine supply contract is signed as soon as possible, and supervision of the construction works is ap-plied.

The grid connection to the 230 kV level is feasible.




17 Annex A

17.1 Annex A - 1: Aviation corridors at Mekelle airport




17.2 Annex A - 2: Enviromental Report of Ashegoda




17.3 Annex A - 3: Soil investigation report




17.4 Annex A - 4: Road Authority Report, Weights and Wheel base




17.5 Annex A - 5: Description of available crane by MIDROC (Addis Abeba)




17.6 Annex A - 6: Road Map Djibouti - Mekelle




17.7 Annex A - 7: Road Survey Report




17.8 Annex A - 8: Terms of reference for consultant's work




18 Annex B

18.1 Annex B 1: Correlation Diagrams




18.2 Annex B 2: Noise Impact




18.3 Annex B 3: Shadow Impact




19 Annex C

19.1 Annex C 1: Map of Ashegoda Wind Park Layout




19.2 Annex C 2: Energy Calculations

Energy yield

Coordinates

Production analysis

Wind data

Map

Power curve

Wind turbine distances




19.3 Annex C 3: Turbulence Calculations




19.4 Annex C 4: two-dimensional view of the digital terrain model




19.5 Annex C 5: three-dimensional view of the digital terrain model




20 Annex D

20.1 Annex D 1: Technical Information of Enercon E-48




20.2 Annex D 2: Technical Information of Vestas V52




20.3 Annex D 3: Technical Information of Gamesa G58




20.4 Annex D 4: Preliminary Technical Information of Enercon E-53




21 ANNEX E :

21.1 Annex E-1 ENERCON E-48 Cash-Flow Economic Analysis




21.2 Annex E-2 ENERCON E-53 Cash-Flow Economic Analysis




21.3 Annex E-3 VESTAS V52 Cash-Flow Economic Analysis




21.4 Annex E-4 GAMESA G58 Cash-Flow Economic Analysis




22 ANNEX F :




22.1 Annex F-1 ENERCON E-48 Operating Results




22.2 Annex F-2 ENERCON E-48 Profit & Loss Account




22.3 Annex F-3 ENERCON E-48 Balance Sheet




22.4 Annex F-4 ENERCON E-53 Operating Results




22.5 Annex F-5 ENERCON E-53 Profit & Loss Account




22.6 Annex F-6 ENERCON E-53 Balance Sheet




22.7 Annex F-7 VESTAS V52 Operating Results




22.8 Annex F-8 VESTAS V52 Profit & Loss Account




22.9 Annex F-9 VESTAS V52 Balance Sheet




22.10 Annex F-10 GAMESA G58 Operating Results




22.11 Annex F-11 GAMESA G58 Profit & Loss Account




22.12 Annex F-12 GAMESA G58 Balance Sheet




23 ANNEX G :

23.1 Annex G-1 Fuel Factors used in the CDM Assessment

Copy of en Feasibility Study Wind Park Ashegoda 2006

Documents

4315kv tap

federal democratic

designated

modestly shaped

tropic consulting

red arrow

ashegoda wind

extensive