Computer simulation of buildings and plants is being used increasingly in energy ssessments nd design. Simulation often requires hourly weather data.In the first part of this paper the North American WYEC2, WYEC1, TMY, TMY2, and IWEC selection methods are discussed and compared with the European Test Reference Years (TRYs) and Design Reference Years (DRYs). The latter re the basis of a new proposed standard, prEN ISO15927-4. It is proposed that the robust Finkelstein-Schafer statistic, used in some of these selection methods but not mentioned in prEN ISO15927-4, should be adopted as the basic selection tool.

In the second part of the paper the lack of extensive near-extreme hourly data for plant sizing is discussed. ASHRAE 1, 3, 5, and 7 day period data and U.K.and European design summer years are considered. In conclusion,it is proposed that an extension of the Finkelstein-Schafer method be used for selecting more near-extreme hourly weather data.

Units: Dual