The Use of Uncertainty Quantification for the Empirical Modeling of Wind Turbine Icing

Abstract

A novel uncertainty quantification method is used to evaluate the impact of uncertainties of parameters within the icing model in the modeling chain for icing-related wind power production loss forecasts. As a first step, uncertain parameters in the icing model were identified from the literature and personal communications. These parameters are the median volume diameter of the hydrometeors, the sticking efficiency for snow and graupel, the Nusselt number, the shedding factor, and the wind erosion factor. The sensitivity of these parameters on icing-related wind power production losses is examined. An icing model ensemble representing the estimated parameter uncertainties is designed using so-called deterministic sampling and is run for two periods over a total of 29 weeks. Deterministic sampling allows an exact representation of the uncertainty and, in future applications, further calibration of these parameters. Also, the number of required ensemble members is reduced drastically relative to the commonly used random-sampling method, thus enabling faster delivery and a more flexible system. The results from random and deterministic sampling are compared and agree very well, confirming the usefulness of deterministic sampling. The ensemble mean of the nine-member icing model ensemble generated with deterministic sampling is shown to improve the forecast skill relative to one single forecast for the winter periods. In addition, the ensemble spread provides valuable information as compared with a single forecast in terms of forecasting uncertainty. However, addressing uncertainties in the icing model alone underestimates the forecast uncertainty, thus stressing the need for a fully probabilistic approach in the modeling chain for wind power forecasts in a cold climate.

Single Column Model Simulations of Icing Conditions in Northern Sweden: Sensitivity to Surface Model Land Use Representation

“…In addition, Molinder et al [24] describes the uncertainties in the formulations for ice shedding, MVD, wind erosion, α 2 , and α 3 . While these uncertainties should be considered, the details of these are beyond the scope of this study.…”

Probabilistic Forecasting of Wind Turbine Icing Related Production Losses Using Quantile Regression Forests

“…Modelling icing and related power production losses are a challenge, both being due to the initial condition uncertainty as well as uncertainties in the modelling chain, such as model formulations, and due to the representation of the exact wind energy site conditions [4][5][6][7]. A common approach for this modelling is to employ an icing model that forecasts ice load and icing intensity based on meteorological weather conditions that are forecasted by the Numerical Weather Prediction (NWP) model before estimating the power production loss (see modelling chain in Figure 1a) [5,8]. The effect of the different uncertainties in the modelling chain has previously been studied [4,6,9,10].…”

“…Adding the icing model to the modelling chain (d). Modelling chain according to Molinder et al [8] and (e). The modelling chain for the QRF probabilistic forecast.…”

“…Wu et al [17] proposed a probabilistic method that is based on observations and NWP forecasts of wind speed also improving the next-day wind power forecasts. Lately, probabilistic forecasting has also been used for icing related production losses [4,8]. In Molinder et al [8], an icing model ensemble was generated that was based on the knowledge of uncertain parts of the icing model.…”

“…Lately, probabilistic forecasting has also been used for icing related production losses [4,8]. In Molinder et al [8], an icing model ensemble was generated that was based on the knowledge of uncertain parts of the icing model. Using this ensemble resulted in a reduced forecast error and it was shown that the forecasted uncertainty in terms of ensemble spread, which was computed as the standard deviation, added valuable information.…”