WIND SPEED DYNAMICS IN COASTAL AREAS VS. MOUNTAIN RANGES - IMPLICATIONS OF DIURNAL VARIABILITY IN URBAN WIND POWER

This paper presents the application of a new approach for taking into consideration the variability of an intermittent energy resource , the wind in this case, at different temporal scales (hourly, daily, monthly, annual) in scenario construction for energy systems modelling. It is argued that commonly used one-step, two-step Markov models, and AR(MA) models do not contain enough low frequency information meaning that eventual stronger winds during a certain time of day, e.g. afternoon winds, are neglected. A new methodology that evaluates the occurence frequency of different wind day types, such as afternoon winds or morning winds, is applied on one hand to wind data from locations with high average wind speeds and little diurnal effects and on the other hand to locations with lower average wind speeds but a clear diurnal cycle. The latter locations are close to or in urban areas and the penetration wind power in these locations is compared to that of the on average windier locations. It is concluded that including diurnal has a slight impact on evaluating excess wind power in an energy system in the assessed locations. A more significant impact is expected in locations where the diurnal effect is not as well aligned with the daily demand pattern, such as locations with night winds.