Wind Power: Cape Verde - background

The Republic of Cape Verde has 17 inhabited islands. A total of 420.000 inhabitants live in Cape Verde. The GNP is 6.000 USD per capita. Cape Verde became an independent republic in 1975 after having been a Portuguese colony. Cape Verde is a developing country.

Cape Verde is a developing country with low GNP per capita. Cape Verde has little domestic energy resources – except for the wind! Due to the remote location of the islands, Cape Verde has an isolated energy system that (obviously) is not connected to a larger grid. The islands have increasing population and also increasing energy consumption per person.

In Cape Verde most of the electricity is generated by diesel generators. At the three islands - São Tiago, São Viciente and Sal - wind power was installed in 1990 and more wind power is planned. In year 2002 there was installed around 40 MW wind power capacity on the three islands.

The wind power substitute some of the diesel generation, and thereby some of the fuel. So the value of the wind power is given by the additional diesel fuel that otherwise would have been necessary for the corresponding diesel generation.

Denmark is a developed country with high GNP per capita. Denmark has several domestic energy resources including oil, gas, biomass and wind. Furthermore, Denmark has electrical power lines and gas pipelines connecting to the neighbouring countries. The electric energy consumption in Denmark is large and the energy consumption is increasing per person. The population does not increase.

Denmark usually exports electricity to neighbouring countries. Denmark has a large installed wind power capacity of 3.1 GW in year 2005. On very wind days, i.e. when the wind is above 15 m/s, the wind turbines generate more electricity than the consumers in Denmark can use. The surplus is exported through the electrical grid to other countries. On average during one year, the wind turbines produce 20% of the electricity consumption in Denmark.

The electricity consumption in Cape Verde and in Denmark is shown in figure 2. The graph also shows the expected electricity consumption up to year 2020.

The energy output from a wind turbine is a function of the wind that hits the rotating wind turbine blades. In general, the higher the wind speed, the more energy is produced. But there are limits. If it is very windy, the turbine is stopped in order to avoid destruction. If the wind is weak, the turbine cannot start.

It is not easy to calculate the produced wind power. The equation below gives a rough idea of the power production. P is power (measured in Watt, W), ρ is air density (measured in kg/m3), A is the area of the rotor plane of the wind turbine (measured in m2) and U is wind speed (measured in m/s).

It can be noted that wind speed is cubed. This means that wind speed variations have a large effect on wind energy production. It is also clear that a large wind turbine with large rotor plane (A) can produce more energy than a small wind turbine. Likewise a wind turbine in dense air will produce more energy than a wind turbine in less dense air. Wind speed and the size of wind turbine are the most important information.

Each type of wind turbine has a different power curve, i.e. the relationship between the wind speed and the produced power. Table 1 contains an example of a wind power curve for a 2MW wind turbine.

The cut-in wind speed is 4 m/s. This is when the turbine starts to rotate. The cut-out wind speed is 25 m/s. This is when the turbine is stopped to avoid damage. The rated capacity of 2MW is achieved for winds between 14 and 25 m/s. Less than full rated capacity is achieved for winds below 14 m/s.

The wind climate usually is described from a wind-rose and a histogram. As an example two contrasting wind climates are shown. One wind climate is from the trade wind belt, the other is from Denmark.

Cape Verde is located in the trade wind belt as is Fuerteventura in the Canary Islands. The example from Fuerteventura is shown in figure 3a and for Rønne in figure 3b. In both cases the wind climate is based on observations from 10 years (1972-1982).

In the wind-roses in figure 3a and 3b it is seen that the dominant wind direction at Fuerteventura is from the northeast occurring 25% of the time. In Rønne the wind direction is mainly from the west occurring 17% of the time.

The histograms in figure 3 show how much time a certain range of wind speed occurs at each site as a frequency (f) in %. For Fuerteventura the most frequent range in wind speed is between 6 and 7 m/s and it occurs 14% of the time. For Rønne the most frequent range in wind speed is between 4 and 5 m/s and it occurs 13% of the time.

In wind energy, it is usual to calculate the Weibull probability distribution function for each histogram. The Weibull function is fitted to the histogram. The Weibull fit is shown as thin curved lines in the histograms.

The Weibull function is described by two parameters: the shape parameter k and the scale parameter A.

Weibull shape k:

When the shape parameter k is small, the histogram is ‘flat’. When k is large, the histogram is more ‘peaked’. In other words, if the wind speed is occurring in a narrow range (as for Fuerteventura) k is large. In contrast, Rønne has a more variable wind speed range thus k is small.

Weibull scale A:

The scale parameter A indicates the ‘dominant’ range of wind speed. A is ‘linked’ to the mean wind speed. Thus a high value of A is good in wind energy.

The Weibull scale A parameter is higher (7.2 m/s) for Fuerteventura than for Rønne (6.6 m/s) in figure 3. This means the winds on average are higher at Fuerteventura than in Rønne.

In addition, the Weibull k parameter is 2.78 for Fuerteventura and 1.94 for Rønne. This means that the wind speed is more ‘centered’ near the mean wind speed at Fuerteventura. This is very good for wind energy as the wind is more constant and wind power can be produced most of the time (keeping in mind the power curve in table 1).

In summary, the wind climate at Fuerteventura is better for wind energy production than in Rønne.

A simple way to test how much better the wind climate at Fuerteventura is compared to Rønne do as follows:

Use the Excel spreadsheet with a power curve for a wind turbine.

Insert the Weibull A and k for the two sites and compare the produced wind power in kW mean (for one year) and the MWh/y produced power for the two sites. The cap fact (capacity factor) shows how much wind power is produced as compared to constant full rated power during one year (for this turbine it would mean that the wind speed should be between 18 and 25 m/s always).

Once the wind climate is observed at a location from a meteorological station, the wind atlas methodology can be used to calculate the wind power production for the entire region near the mast. The wind atlas method is based on the physical principles of flow in the atmospheric boundary layer. A basic equation is the logarithmic wind profile that states that the wind speed increases with the logarithm of the height above ground. The effect of different surface conditions (mountains, coastlines, forests, etc.) and sheltering effects due to buildings and other obstacles are included.

The basic principle of the wind atlas methodology is shown in figure 4. The ‘up-arrow’ illustrates the use of the model on measured wind data to calculate a so-called regional wind climatology. The ‘down-arrow’ illustrates how the regional wind climatology is used as input to the same model to produce site-specific wind climatologies and, given the power curve of a wind turbine, power production estimates. This is wind resource assessment.

A detailed wind map for a typical day in Cape Verde is shown in figure 5. The wind map is based on a satellite SAR image from Envisat. The wind direction is from the northeast with wind speed near 10 m/s (pale blue).

From the SAR wind map it is clear that wind speed variations occur. There are lee-effects of the islands (dark blue areas). The lee-effect is due to the mountains of the islands. The larger and higher the island, the more pronounced is the lee- effect (see the height of the islands in figure 6).

There is also an example of gap-flow, i.e. accelerated wind flow (seen as yellow and red area), between two islands Santo Antão and São Vicente in the northwest in figure 5.

For more information on ocean wind from satellite SAR, please, see the running project ‘Wind’. http://galathea3.emu.dk/satelliteeye/projekter/wind/index_uk.html.

Satellites only observe wind over the ocean, not over land.

The topography of Cape Verde and Denmark is shown in figure 6. The effect of the Cape Verde islands to the atmospheric flow in figure 5 is clear. The difference in topography between the different islands in Cape Verde and Denmark can be noticed in figure 6.

The topography was observed by the space shuttle Endeavour in February 2000 during its 11-day mission. A space-borne shuttle is a space platform with a much shorter life-time than a satellite. The space shuttle mission is called the ‘Shuttle Radar Topography Mission’ (SRTM). SRTM mapped 80% of the Earth’s topography (elevation). This covers the land area of the world between 59° degrees south and 60° degrees north latitude.

The radar had two antennas. One radar antenna was located in the shuttle’s payload bay, the other at the end of a 61-m long mast that extended from the payload bay once the shuttle was in space. Radar interferometry techniques were used to infer the elevation of the terrain surface from the radar data. In interferometry, two images are taken from different vantage points of the same area. The slight difference in the two images allows scientists to determine the height of the surface. SRTM is shown in figure 7.

The shuttle obtained about 9 terabytes of raw data with a spatial resolution in the horizontal of about 25*25 m2. These data were then re-sampled to resolutions of 1 arc-sec (around 30 m) for the USA and 3 arc-sec (around 90 m) outside the USA. The accuracy in the z-dimension is around 5 to 10 m. Beautiful fly-over animations of the global topography are available at the web.

In wind energy the topography is important as the wind speed usually is higher on the mountain top than in the valley. So the local wind phenomena due to topography always have to be considered. For the island São Tiago in Cape Verde the SRTM map is shown in figure 8.

The wind climate at Cape Verde has been observed for one year on three islands. The results are shown in table 2. The observations from Mindelo are seen to be the best for wind power production (high mean wind speed, high Weibull A and high Weibull k).

Table 2. Annual mean wind speed (U), Weibull A and Weibull k for three locations in Cape Verde. The values are from 30 m above ground level. Source: Risoe National Laboratory, Jens Carsten Hansen and Niels Gylling Mortensen.

For Mindelo, the average variation in wind speed during the day for each of the four seasons is graphed in figure 9. Winds are slightly higher at noon than during night. Winds are slightly higher in winter and spring than in summer and autumn. Overall the variation both at the daily and the seasonal scale is very small.

Usually winds are observed at one location when several wind turbines are planned in the region. In order to calculate the wind resource for the region, the wind atlas method is applied. A wind resource map from an area at the island São Tiago, Cape Verde is shown in figure 10. The wind resource map covers the area outlined with a red box in figure 8.

The large variation in wind resources is a function of topography with high wind speed on top of the mountains and low wind speed in the valleys. The dominant wind direction (northeast) is important to keep in mind for understanding the wind resource variations.

Winds are always observed for at least one year before a wind farm is fully planned. Usually winds are observed every hour. This is to make sure that the daily and annual variations are known, and the Weibull fit is representative for the area.

But one should be very careful! Some years are characterized by extraordinary warm summers, cold winters or rainy periods. Similarly the wind climate changes between years. The variation in winds between years is different in different parts of the world. The situation for Cape Verde is shown in table 3.

In the trade wind belt, the winds are rather constant. The wind index shows how different the wind is between years (in %). The NCAR/NCEP re-analysis data in table 3 are model results from a global model. This means that the data are not accurate for one specific island (but covers all islands as well as the ocean between). What the data can be used for is to ascertain if a one-year time-series from a mast is collected in a windy or not so windy year, and make a correction if necessary.

A wind turbine has a life-time of 20 years, so it is the 20-year average wind power production that has to be estimated before the investor/bank is ready to fund the wind farm. (also the price on electricity during the coming 20 years is important).

The average wind speed per month from 31-years for a 1 by 1 degree grid cell near Cape Verde from the NCAR/NCEP re-analysis model results is listed in table 4.

Table 4. NCEP/NCAR re-analysis data monthly mean wind speed for the period 1970-2000 for Cape Verde. Source: Risø.

Wind climates in Cape Verde and Denmark are very different. However, both countries have wind climates well-suited for wind power production.

First of all, the wind climatology in Cape Verde is governed by the trade winds. These are that nearly constantly blowing from the northeast. In contrast, the wind climate in Denmark is characterized by synoptic frontal systems passing from different directions. The wind direction is changing a lot and so is the wind speed.

Secondly, the mountains of Cape Verde gives rise to a much more varied wind resource than is found in Denmark. Denmark is a flat country.

Satellites observe the ocean winds. The SAR-based wind map from Cape Verde (figure 5) is typical for the islands. The wind map is observed from the satellite Envisat. Envisat observes Cape Verde and Denmark a few times per month. Therefore relatively few SAR wind maps are available. The advantage of the SARwind map is the high spatial resolution of 600 by 600 m2. Many small details in the horizontal domain can be seen.

In contrast, the QuikSCAT satellite always observes ocean winds twice per day with a spatial resolution of 25* 25 km2. The mean wind speed for Cape Verde and Denmark observed from QuikSCAT is shown in figures 11 and 12, respectively. The wind data time series covers seven years (around 5000 observations).

Comparing the mean wind speed maps from Cape Verde and Denmark shows winds around 7 to 9 m/s for most of the offshore area. In Cape Verde the highest winds appear in regions with gap-flow between islands in the northwest. In Denmark the highest winds appear in the North Sea.

The monthly mean wind speed for two grid cells with mean wind speed around 8 m/s is extracted and shown in figure 13 and 14 for Cape Verde and Denmark, respectively. It is very interesting to compare the differences in the two figures. For Cape Verde a modest seasonal variation is found. In Denmark, there is a strong seasonal variation with low winds in summer and very high winds in winter.

Finally, the wind-roses for the two sites observed by QuikSCAT are given in figures 15 and 16 for Cape Verde and Denmark, respectively. For Cape Verde, north-easterly flow is totally dominating (55.6% and 30.4%) of the time. Winds hardly ever occur from the south or west. For Denmark, winds from all directions occur, though most frequently from the northwest, west and southwest.

The Weibull A, Weibull k and mean winds are listed in table 5. Both Weibull A and Weibull k are higher for Cape Verde which indicates an overall better wind climate for wind power production.

Table 5. Wind climate statistics for Cape Verde (southwest of the island Santo Antão) and Denmark (Horns Rev in the North Sea) observed from QuikSCAT satellite ocean wind maps for seven years.

The wind climate in the North Sea is excellent for wind farming. This has been recognized years ago in Denmark and in year 2002 the worlds largest offshore wind farm started to operate at Horns Rev. There is a total of 80 2MW wind turbines in the grid. The wind turbines can be seen from space from satellite SAR on calm days. Then the ocean surface appears dark grey or black and the wind turbines are seen as white spots in figure 17.

The Horns Rev offshore wind farm seen from the transformer station is shown in figure 18. The transformer station is where the electricity is collected before being sent to land through the underwater grid line. At the transformer station the helicopter landing site is seen. A helicopter is typically used when servicing the wind turbines. In this case the wind turbine is stopped, and a man is lowered down with a line to the top of the wind turbine hub from where there is a door into the hub. It is also possible to sail to the wind turbines and use the elevator inside the tower to go up to the hub, but often the waves are too rough for sailing to the turbines to be safe enough.