ANALYSIS OF DAILY GLOBAL IRRADIATION DATA FOR 5 SITES IN RWANDA AND ONE IN SENEGAL

A comparative study of daily global radiation data, G, is presented for five sites in Rwanda and one in Senegal, i.e. Dakar. Analysis of hourly global values (monthly means) shows that G is more sensitive for Rwanda than for Dakar to contributions from extreme hours of the theoretical sunny day. In seasonal variations, maximum periods of G for Rwanda correspond to minimum periods for Dakar and vice versa. The G modal interval is the same for the two kinds of climate, i.e. from 2000 to 2500 J/cm2/day. Nevertheless, the mode and variability for G are higher for Dakar than for Rwanda. One can also note some days with very high G values for Rwanda, one of those values being higher than the extraterrestrial one, G(o). Studies of clearness index, K(t), sunshine duration, S, and sky cover, N(e), all converge towards the conclusion that throughout the year, the sky over Dakar is more sunny than Rwanda's. Dakar is therefore more suitable than Rwandan sites for solar plants using concentrating devices, because the seasonal requirement of a great number of days with high values of S makes them feasible. Furthermore, a research of simple relationships between clearness index, K(t), relative sunshine duration, S/S(o) and relative cloud cover, N(e)/S for the six sites shows satisfactory linear regressions in which correlation coefficients, R(i), are slightly better for Dakar. Regression coefficients, a(i) and b(i), in their turn depend on the site. It is confirmed that S/S(o) is a better estimator for G than N(e)/8 and it is shown that those correlations are more efficient when long period means are involved for the parameters in relation. It is also shown that: (i) Hay's modification to the Angstrom-Black regression doesn't reduce the scatter in the regression parameters; (ii) N(e)/8 is slightly better than (N(e)/8)2 in estimating G.