Xin Yuan
PhD
Classification, modelling, and parameterisation of tall wind profiles
​
Host Organisation
Technical University of Delft (TUD)
​
Company
​TNO
Project Description
My project aims to develop new wind profiles suitable up to 500m tall, based on observations, reanalyses, and numerical modeling. While some research has been conducted on defining parametric profiles that extend above the surface layer, the validation of these profiles is limited, and they require additional parameters that are not easily accessible. The availability of more detailed reanalysis datasets has opened the possibility to better study the wind speed profile beyond the surface layer. Machine learning techniques can reduce the large amount of data into a limited number of key profile types. These clusters provide an opportunity to set boundary conditions for numerical models (mesoscale or microscale) to gain insight into important parameters which drive the shape of wind speed profiles beyond the surface layer. I will investigate how a combination of profile clustering and numerical modeling can be used to provide robust wind speed profiles accurate to 500m using parameters that can easily be inferred from observations or available reanalyses. These profiles can then be used to evaluate the loading conditions of 25+ MW turbines for design and performance calculations.
Supervisors
Main Supervisor: Prof. Simon Watson (TUD)
External Academic Supervisor: Prof. Joachim Reuder (UiB)
External Industry Supervisor: Dr. Peter Eecen (TNO), Dr. Wouter Engels (TNO)
Background
I am from China and hold a B.Sc. in Atmospheric Science from Chengdu University of Information Technology (China), as well as an M.Sc. in Integrated Climate System Sciences from Universität Hamburg (Germany). My experience as a wind resource engineer at SANY Renewable Energy deepened my focus on wind energy and strengthened my motivation to pursue research in this field. As a PhD candidate at TU Delft, I am now working to bridge atmospheric science with wind energy and to develop a reliable and robust method for determining tall wind speed profiles.
Motivation
A critical observation during my time in the wind energy industry shaped my decision to pursue a PhD. While working as a wind resource engineer, I saw firsthand how existing models struggle to capture the true complexity of the atmosphere, especially at higher altitudes. This was not just a technical limitation. It was a barrier standing between innovative ideas and real-world impact. I found myself wanting not just to use models but to improve them, not just to apply atmospheric science to wind energy but to push it forward. This program offers the opportunity to develop more accurate and physically meaningful approaches that can truly advance the field.