Wind Speed Return Period Calculation via Mesoscale Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for calculating extreme wind speeds at wind turbine sites rely on short-term local measurements, leading to inaccurate estimates for long-term return periods, which can result in wind turbines being under or over-designed, wasting resources and potentially failing to withstand actual wind conditions.
Innovation Solution
A method that combines local wind speed measurements with mesoscale data to calculate a correction factor, allowing for more accurate estimation of wind speeds over extended periods by selecting and characterizing storms in both measured and modeled data sets, and applying these corrections to modelled wind speeds to determine the wind speed associated with a return period.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If short-term local wind speed measurements are used to calculate extreme wind speeds, then the measurement process is simple and quick, but the accuracy of the return period wind speed estimation is poor
Solution Approach 1:
The patent applies preliminary action by pre-processing wind speed measurements to identify and characterize individual storms within the measurement period. Storms are selected based on exceeding threshold wind speeds, and characteristic wind speeds are determined for each storm before combining them with mesoscale model data. This pre-characterization of storms enables accurate extrapolation to return periods without requiring measurements spanning the entire return period.
Solution Approach 2:
The patent uses mesoscale wind speed models as an intermediary to bridge the gap between short-term local measurements and long-term return period estimates. The modelled wind speeds provide additional storm data that can be corrected using the correction factor derived from local measurements, effectively extending the measurement record without actually measuring for the full return period duration.
2Reliability
If short-term local measurements are used for return period calculation, then data collection is straightforward, but the reliability of the wind speed estimation is low
Solution Approach 1:
The patent merges local measured wind speed data with mesoscale modelled wind speed data to create a combined data set for return period calculation. Local measurements provide accurate characterization of recent storms, while mesoscale models provide additional historical storm data. The correction factor method combines these two data sources, allowing reliable return period estimates with limited local measurement data.
Solution Approach 2:
The patent changes the parameter representation by determining characteristic wind speeds for individual storms rather than using continuous or averaged wind speed data. This transformation allows selective combination of storms from different data sources (local measurements and modelled data) while maintaining the physical meaning and statistical properties needed for reliable return period calculation.
3Manufacturing precision
If conventional Gumbel calculation is used with limited data, then the calculation process is simple, but the manufacturing precision of wind turbine design is compromised
Solution Approach 1:
The patent segments the wind speed data into individual storm events, each characterized by its own characteristic wind speed. This segmentation allows selective processing and combination of storms from different sources (local measurements and mesoscale models). The correction factor is applied to modelled storm data rather than to the entire data set, providing precise control over how measured and modelled data are integrated for design purposes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of calculating a wind speed associated with a return period at a proposed wind turbine site is provided. The method comprises selecting a first set of storms from wind speed measurements measured at the proposed site, and selecting a second set of storms from modelled wind speeds, the modelled wind speeds being estimates of wind speeds at the proposed wind turbine site during the measurement period from a mesoscale data set. A comparison is made of characteristic wind speeds of the first and second sets of storms to determine a correction factor. A third set of storms is selected from an extended set of modelled wind speeds from an extended mesoscale data set, and the correction factor is applied to characteristic wind speeds of the third set of storms to provide corrected wind speeds. A wind speed associated with the return period is calculated from the corrected characteristic wind speeds.