Search

Wind energy is one of the potential options to fill the gap in renewable energy production in Switzerland during the winter season when the energy demand exceeds local production capacities. With likely further rising energy consumption in the future, the winter energy deficit may further increase. However, a reliable assessment of wind energy potential in complex terrain remains challenging. To obtain such information, numerical simulations are performed using a combination of the "Consortium for Small-scale Modeling" and "Weather Research and Forecasting" (COSMO-WRF) models initialized and driven by COSMO-1E model, which allows us to simulate the influence of topography at a horizontal resolution of 300 m. Two LiDAR measurement campaigns were conducted in the regions of Lukmanier Pass and Les Diablerets, Switzerland. Observational LiDAR data and measurements from nearby wind sensor networks are used to validate the COSMO-WRF simulations. The simulations show an improved representation of wind speed and direction near the ground compared to COSMO-1E. However, with increasing height and less effect of the terrain, COSMO-WRF tends to overestimate the wind speeds, following the bias that is already present in COSMO-1E. We investigate two characteristic mountain-terrain flow features, namely waves and Foehn. The effect of mountain-induced waves of the flow is investigated through an event that occurred in the area of Diablerets. One-year analysis for the frequency of conditions that are favorable for mountain wave formation is estimated. The Foehn impact on wind was observed in the Lukmanier domain. We attempt quantification of the probability of occurrence using the Foehnix model. The result shows a high probability of Foehn occurrence during the winter and early spring seasons. Our study highlights the importance of incorporating complex terrain-related meteorological events into the wind energy assessment. Furthermore, for an accurate assessment of wind speed in complex terrain, our study suggests the necessity to have a better representation of the topography compared to COSMO-1E. [ABSTRACT FROM AUTHOR]

In order to assess the level of power reserves during down-regulation, the available power of a wind turbine needs to be estimated. The current practice in available power estimation is heavily dependent on the pre-defined performance parameters of the turbine and the curtailment strategy followed. This paper proposes a single-input model-free approach dynamic estimation of the available power using recurrent neural networks. Accordingly, it combines wind turbine control considerations and modern forecasting methodologies for a model-free, single-input estimation of available power. It enables a robust real-time implementation of dynamic delta control, as well as higher-accuracy provision of the reserves to the system operators. The model-free approach requires only 1 Hz wind speed measurements as input and estimates 1 Hz available power as output. The neural network is trained, tested and validated using the DTU 10MW reference wind turbine HAWC2 model under realistic atmospheric conditions. The unsteady patterns in the turbulent flow are represented via long short-term memory (LSTM) neurons which are trained during a period of normal operation. The adaptability of the network to changing inflow conditions is ensured via transfer learning, where the last LSTM layer is updated using new measurements. It is seen that the sensitivity of the networks to changing wind speed is much higher than that of turbulence, and the updates are to be implemented solely based on the altering inflow velocity. The validation of the trained LSTM networks on time series with 7, 9 and 11ms-1 mean wind speeds demonstrates high accuracy (less than 1% bias) and capability of transfer-learning online. Including highly turbulent inflow cases, the networks have shown to comply with the most recent grid codes, which require the quality of the available power estimations to be evaluated with high accuracy (less than 3.3% standard deviation of the error around zero bias) at 1 min intervals. [ABSTRACT FROM AUTHOR]