Abstract
The energy industry uses weather forecasts for determining future electricity demand variations due to the impact of weather, e.g., temperature and precipitation. However, as a greater component of electricity generation comes from intermittent renewable sources such as wind and solar, weather forecasting techniques need to now also focus on predicting renewable energy supply, which means adapting our prediction models to these site specific resources. This work assesses the performance of The Air Pollution Model (TAPM), and demonstrates that significant improvements can be made to only wind speed forecasts from a mesoscale Numerical Weather Prediction (NWP) model. For this study, a wind farm site situated in North-west Tasmania, Australia was investigated. I present an analysis of the accuracy of hourly NWP and bias corrected wind speed forecasts over 12 months spanning 2005. This extensive time frame allows an in-depth analysis of various wind speed regimes of importance for wind-farm operation, as well as extreme weather risk scenarios. A further correction is made to the basic bias correction to improve the forecast accuracy further, that makes use of real-time wind-turbine data and a smoothing function to correct for timing-related issues. With full correction applied, a reduction in the error in the magnitude of the wind speed by as much as 50% for “hour ahead” forecasts specific to the wind-farm site has been obtained.
Highlights
Weather forecasting plays an important role within the electricity industry [1], with forecasts mainly focusing on estimating future demand, which is primarily temperature dependent
Forecasting does not require a large amount of computational time compared to other Numerical Weather Prediction (NWP) models, and it allows
Weather forecasting within the power sector to date has largely concentrated on how the weather, temperature, affects consumers’ power usage and overall system demand
Summary
Weather forecasting plays an important role within the electricity industry [1], with forecasts mainly focusing on estimating future demand, which is primarily temperature dependent. The wind energy sector has grown substantially over the past decade and is emerging as an important contributor to electricity generation around the world [2,3,4]. Useful predictions of future wind output can play a valuable role in appropriately committing (starting) and dispatching other generation to meet expected demand. The number of research groups investigating how to better predict wind power due to the increased uptake of wind energy is growing, as is the emergence of some commercial wind power forecast providers ([5,6,7,8,9,10], to name a few). Providing wind farm specific forecasts has different challenges to conventional weather forecasts
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have