Wind at Northern Senja: Use of numerical weather prediction models for wind resource assessment

Abstract

Wind resource assessment for establishing decentralized power production in remote and complex areas, is challenging due to a lack of observational data. To overcome this problem, Numerical Weather Prediction (NWP) models can be used to provide data where no observations exist. This thesis uses NWP models to do a wind resource assessment at northern Senja. The power grid in the area is prone to voltage drops, and decentralized power production would release some transmission capacity. The study is twofold: The first part investigated the model performance at three observation sites Hekkingen, Innhesten, and Senjahopen. The second part is a wind resource assessment based on model data.

Calculated statistical scores and wind roses showed that Arome 500 with a horizontal resolution of 0.5 km was the best model for predicting wind speed and direction in the fjords at northern Senja. Furthermore, no evident correlation between Arome 500, AROME-Arctic and NORA3 could be concluded from the statistical scores and wind roses, mainly due to short overlapping periods and datasets. Production estimates at Mykjeneset, Giskhaugen, and Sporra were computed from three Enercon wind turbines' power curves. Two levels of turbulence corrections were applied. Before any turbulence corrections; Sporra showed the highest annual energy production of 2587 MWh and a capacity factor of 36.8% for the Enercon E-53 turbine. After the turbulence corrections, the same turbine had the highest annual energy output at Giskhaugen of 2180 Mwh and a capacity factor of 31%.