Layout of Wind Projects

Wind Potential Map

Based on long-term corrected data of the wind measurements on site, a wind potential map has been calculated (using WindPro and WAsP software), providing 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 1 shows an example for a wind potential map for the Ashegoda wind project in Ethiopia.

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

In this map, it can be clearly seen that ridges in the west and and in the east 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. In this example military installations north of the road (red line) are limiting the wind park area to the region south of the road until it reaches the eastern ridges. 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.

Wind Turbine Selection

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 investment cost for the tower and the foundation. In some countries available crane capacities could be 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 increasing 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 different hub heights showing the effect of higher hub heights described above: 

Figure 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 presented, 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.

Besides wind speed characteristics the choice of hub height depends on the terrain: For example the Ashegoda wind project site in Ethiopa has a quite wind speed characteristic (characterised as wind class III b according to the IEC regulations) and 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 calculations like mesoscalic models like KLIMM. 

If these advanced options are not available, it will be necessary to choose a hub height, which does not exceed the height for qualified wind regime predictions for the specific site characteristics. At the Ashegoda site meaningful prediction of the wind regime was considered possible up to hub heights of 50 to 60 m. Thus the costs related to these hub heights have been considered in the feasibility study for the Ashegoda wind park.

It has to be checked, whether the crane capacity for the construction of high turbines is available in the country the project should be realized in. If cranes of this size are not available the cost for their mobilization from another country have to be included in the financial assessment for the wind park.

Determination of the Optimal Unit Size

The selection of the wind turbine type 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 turbines in the Megawatt - range
• Local experiences with regular operation and maintenance with wind turbines
• Distances between site and turbine manufacturer who will perform the maintenance within the warranty period
• Wind turbines types yet installed in the county
• Energy Yield
• Turbine types available at the regional 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 manufacturers are not willing to offer turbines in the Megawatt class for countries which are just entering the wind energy market.

For the first projects realized in such countries, it is therefore strongly recommended to start with proven wind turbines. This will have the following advantages when compared to wind turbines with larger 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 tendering for 800 kW range wind turbines is performed, enabling the client a more detailed 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.
• It can be expected that the delivery time for turbines in the Megawatt range will be longer when compared with smaller turbines.  

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:

Figure 2: Specific investment costs for selected turbine types

Additional Assumptions for V80 and E-70:

• 50 % additional transportation costs per wind turbine
  
• 60000 € additional costs for crane
  
• 5 days for installation per turbine instead of 3
  
• 30 % additional costs per foundations
  
• 5 % additional interests due to extended construction period
  
• No electrical losses are considered
  
• prices for E-70 + V80 from 2005
  

Parameters

	Park efficiency	Availability	No. Turbines
E-48 hub height 57m	0,95	0,95	86
V52 hub height 60m	0,95	0,95	86
V80 hub height 60m	0,97	0,92	34
E-70 hub height 65m	0,97	0,92	34

The lower total number of V80 / E-70 wind turbines compared to the number of E-48 / V52 turbines is due to the effect that the distances between Multi-Megawatt turbines have to be considerable larger which may cause a reduced total installed capacity and thus a lower total energy production compared

Energy Yields [MWh/wind turbine/y]

E-48 hub height 57 m	2,652
E-70 hub height 65 m	6,462
V52 hub height 60 m	2,706
V80 hub height 60 m	5,704

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. If there is sufficient space available at the site (which is a limiting factor in most European sites today) to set up the installed capacity with smaller turbines like the 800 kW turbines in the example, a smaller turbine class could be a preferred choice, in case the transportation of large turbines to the site is difficult and connected to additional costs. Furthermore delivery time could be an important variable: Generally spoken smaller proven wind turbines have shorter delivery time as the large multi-megawatt machines. Thus, developers of a wind project have to conduct a small survey, which provider of wind turbines is able and willing to deliver the turbines to the country the project is located.

Turbine Distances / micro-siting

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 turbines. The minimum distance of three times or less the rotor diameter in cross wind direction is only feasible in case the wind direction is strictly perpendicular to the row of wind turbines. The smallest distances in cross wind direction are determined by a layout development iteration process, carried out under the condition that the wake losses do not fall below a chosen average level for the individual turbines, which is considered as necessary for the economical operation of the wind farm (e.g. the wake losses do not cause decrease in energy output below 85 % for the following wind turbines). 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 distance between two adjacent turbines can be larger.

For the Ashegoda wind project the minimal distances are summarized in the following table. As the wind turbines at this site should be erected in a single row, the distance of the turbines in the prevailing wind direction is not considered. If space is not a limiting factor for a wind park, the placement in a single row is very benificial because an important part of the wake effects can be avoided.

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

The wind potential map (see Figure 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, resulting 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 distance between the individual rows are more preferable than clusters of turbines, leading to the layout-example shown in Figure 3.

In the example layout the orientation of the central part of the eastern ridge nearly parallel to the main wind direction (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.

Figure 4: Layout example for the Ashegoda wind park in Ethiopia

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. The turbulence of the wind flow is a factor which causes stress and fatigue 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 can be calculated and analysed by means of the WindPro software package using for example the Frandsen Turbulence Model (as recommended in IEC 61400 1, Rev 3) for GFK-Materials (rotor).

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 calculated 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.

New studies show that the Frandsen-Model overestimates the turbulence applied to the turbines. The Empirical Turbulence Dutch TNO Laboratory 1993 Model could be chosen as alternatively. A consultation of the manufacturers of the wind turbines for checking the turbulence impact is generally required.

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 distances 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 software 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 can be considered depending on the utilisation of the area:

Noise standards according to German standards

Utilisation	Noise emission [dB(A)]
	Day time 6:00-22:00	Night time 22:00 - 6:00
Regimen and hospital areas	45	35
Exclusive residential areas	50	35
General residential areas	55	40
Village centres, mixed utilisation with small trades	60	45
Working areas	65	50
Industrial areas	70	70

Table 6-5: Noise standards according to German standards Utilisation Noise emission [dB(A)] Day time 06:00 22:00 Night time 22:00 06:00 Regimen and hospital areas 45 35 Exclusive residential areas 50 35 General residential areas 55 40 Village centres, mixed utilisation with small trades 60 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 boundary 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 proposed wind farm. The detailed results can be found in the Annex B 2. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 85 LI / GE6 25 0477 final report ashegoda 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 happens 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 commonly 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 probability . 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 adequately 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 - 3.6 Technical Layout of Ashegoda Wind Park Lahmeyer International was requested by the client to develop a wind park layout for approximately 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), providing 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. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 68 LI / GE6 25 0477 final report ashegoda 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 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 69 LI / GE6 25 0477 final report ashegoda 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 limiting 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 investment 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 increasing 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 different hub heights showing the effect of higher hub heights described above: Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 70 LI / GE6 25 0477 final report ashegoda 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 /kwh 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 presented, 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 calculations like mesoscalic models like KLIMM. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 71 LI / GE6 25 0477 final report ashegoda 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 normal 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 turbines in the Megawatt - range - Local experiences with regular operation and maintenance with wind turbines - Distances between site and turbine manufacturer who will perform the maintenance 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 manufacturers are not willing to offer turbines in the Megawatt class for countries which are just entering the wind energy market. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 72 LI / GE6 25 0477 final report ashegoda 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 larger 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 tendering for 800 kW range wind turbines is performed, enabling the Client a more detailed 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: Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 73 LI / GE6 25 0477 final report ashegoda Total Invest- Cost [EUR] Total Invest- Cost [EUR] plus Construction Interests Single Cross Energy Yield P50 [MWh/y] Wind Farm Net Energy Yield P50[MWh/y] Specific Investment Costs [EUR/MWh/y] V52 85,148,914 85,148,914 2,706 210,026 405 E-48 79,988,914 79,988,914 2,652 205,835 389 V80 HH 60m 84,498,914 88,723,860 5,704 173,068 513 E-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 that the distances between Multi-Megawatt turbines have to be considerable larger which may cause a reduced total installed capacity 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 installed 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 Ashegoda 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-70 1) 50%/ WTG additional transportation Costs 2) 60.000 EUR additional Costs for Crane 3) 5 days for installation per Turbine instead of 3 4) 30% additional costs per Foundations 5) 5% additional interrests due to extended construction periode 6) No Electrical Losses are considered 7) Prices for E-70 + V80 from 2005 Parameters Park efficiency Availibility No. Turbines E-48 hub height 57m 0.95 0.95 86 V52 hub height 60m 0.95 0.95 86 V80 hub height 60m 0.97 0.92 34 E-70 hub height 65m 0.97 0.92 34 Energy Yields [MWh/WTG/y] E-48 hub height 57m 2,652 E-70 hub height 65m 6,462 V52 hub height 60m 2,706 V80 hub height 60m 5,704 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 74 LI / GE6 25 0477 final report ashegoda Table 6-2: Requests for expressions of interest Manufacturer (contact person) Contact(s) Remarks Nordex: Mr. Dwenger March 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. Fricke March 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 statement 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. Hoch March, 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 Enercon 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 answered until now, smaller turbines are currently not available Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 75 LI / GE6 25 0477 final report ashegoda Suzlon Energy: Mr. Patel May and June 06 An enquiry for 600kW and 950kW turbines has not been answered until now Siemens: Mr. Kruse May 06 New markets like Ethiopia are not priority markets for Siemens 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 Energiesysteme: Mrs. Lefevre March and April 06 SeeBA is a manufacturer of lattice towers for several tower heights and turbine types, e.g. Nordex-, Fuhrländer-, REpower- and Vestas turbines. This means that SeeBA s answer 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 envisaged 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 by Enercon-India) Note: According to the latest information from Enercon, the company is currently developing 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. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 76 LI / GE6 25 0477 final report ashegoda 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 turbines. The minimum distance of three times or less the rotor diameter in cross wind direction 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 considerable 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 distance 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 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 77 LI / GE6 25 0477 final report ashegoda 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 chapters 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, resulting 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 distance 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 direction (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 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 78 LI / GE6 25 0477 final report ashegoda 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 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 79 LI / GE6 25 0477 final report ashegoda 6.4.2 Wind Park Layout Vestas V52 Figure 6-5: Ashegoda Wind Park, Layout Vestas V52 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 80 LI / GE6 25 0477 final report ashegoda 6.4.3 Wind Park Layout Gamesa G-58 Figure 6-6: Ashegoda Wind Park, Layout Gamesa G-58 Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 81 LI / GE6 25 0477 final report ashegoda 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 efficiency and energy production. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 82 LI / GE6 25 0477 final report ashegoda 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 fatigue 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 analysed 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 calculated 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 ambient turbulence and wakes Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 83 LI / GE6 25 0477 final report ashegoda 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 impact 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 distances 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 software 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 depending on the utilisation of the area: Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 84 LI / GE6 25 0477 final report ashegoda Table 6-5: Noise standards according to German standards Utilisation Noise emission [dB(A)] Day time 06:00 22:00 Night time 22:00 06:00 Regimen and hospital areas 45 35 Exclusive residential areas 50 35 General residential areas 55 40 Village centres, mixed utilisation with small trades 60 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 boundary 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 proposed wind farm. The detailed results can be found in the Annex B 2. Feasibility Study for Windpark Development in Ethiopia and Capacity Building August 2006, Final Report - page 85 LI / GE6 25 0477 final report ashegoda 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 happens 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 commonly 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 probability . 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 adequately 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.