TANZANIAN DATA APPLIED

7b. The above formulae were applied to 59 Tanzanian locations for which monthly average indicated cca wind velocity data was available.

These average annual indicated windspeeds were from 0.63 metre/second to 3.28 metre/second, with an average value for the 59 sites of 1.88 m/s.

These gave optimum cut-in windspeeds of from 2.69 to 4.76 m/s, with an average value of 3.67 m/s.

The lowest and highest energy yield months were noted, and the ratios of the values of vciopt for these lowest and highest months were found to lie in the range of 0.47 through 0.94, with an average of 0.78; similarly, the ratios of vciopt for the lowest month to the vciopt for the annual situation were found to lie in the range of 0.55 through 0.97, with an average of 0.87; only 3 locations out of 59 were outside the 90% energy band ratio of 0.78. This indicates that an optimisation strategy to select a single value of vci for year-round operation will result in an annual energy yield of more than 90% of absolute maximum for 56 of the 59 locations - i.e. such a strategy of basing vcibest on the vciopt for the worst month of the year seems good.

Using this strategy of setting the vciopt at that for the worst month (as long as that worst month vciopt was greater than or equal to the 90% energy yield band of 0.78 * vciopt annual), then a value of vcibest was determined for each of the 59 locations. These varied from 2.53 through 4.13 m/s, with an average value of 3.22 m/s.

These vcibest values were graphed against vcca (average indicated annual windspeed), and gave a good correlation with a best-fit formula of :

vcibest = 2.2 + (0.52 * vcca)

Note that this is similar to the vciopt formula as in (5) above.

Note also that if monthly vcca figures are available then really vcibest should be worked out from the vciopt figures for each month, and NOT by using this 0.52 formula....

7c. The seasonal variability of wind energy for the 59 Tanzanian locations was examined graphically - it was found that for most locations there was a clear trend with a peak August-October and a trough December-January; by comparison, to consider rainfall, for Dar es Salaam, rainfall is usually close to zero June-September, with rain starting October and building up steadily to a peak April-May.

8. On a practical note, vci is often not the actual physical cut-in speed (since the pump tends to stop at the difficult part of the cycle) - more on this later and how to assess the performance of existing windpump installations.

9. Summarising so far, we see that if we have cca data for a site, then we can calculate the energy yields for all types of WEC. We can also calculate the best cut-in speeds for impulse type WEC's so as to maximise energy yields and to minimise seasonal energy yield variations or otherwise to optimise seasonal behaviour.