- Abstract We discuss some of the practical issues associated with the use of a stationary multipyranometer system for determining both the beam and diffuse components of solar radiation, and the global insolation on arbitrary inclined planes. Results from a full year of validation measurements are presented and compared with (a) the values obtained using conventional, clock-driven, instrumentation (in the case of the direct beam component); and (b) with model-derived results (in the case of total insolation on inclined planes). The yearly average rms error for the beam component is found to be 50 W/m 2 (using 10 min data) or 47 W/m 2 (using hourly totals). The corresponding rms error for the horizontal global radiation is 18 W/m 2 (for 10 min data). In the case of total insolation on arbitrary inclined planes, the rms errors are found to be smaller than 30 W/m 2 , which is comparable to the accuracy given by the best anisotropic-diffuse models currently available.