Characterisation of offshore winds for energy applications

Abstract

Offshore wind energy is regarded as a key component of future energy systems. However, rolling out wind energy on a large scale requires scientific, technological and economic breakthroughs. Wind power forecasting, resource assessment, wind farm optimization, and turbine load assessments all rely on engineering models that simulate the interaction between the wind and the turbine(s). These models often represent the wind in an idealised fashion, which introduces uncertainties that translate into financial risk for investors. This thesis addresses these uncertainties by re-evaluating common assumptions about the (offshore) wind field, studying the physics that governs winds in coastal areas, evaluating the representation of offshore winds in numerical weather prediction models, and proposing alternative methods to represent the offshore wind climate in engineering models. To characterize uncertainties in the simulated wind climate, the concept of anomalous wind events is introduced. An important and illustrative example is the low-level jet. During a low-level jet event, the wind speed does not increase monotonically with height, as is often assumed. Instead, it reaches a maximum in the lower atmosphere (typically around wind turbine hub height), and then decreases with height. As such, low-level jets can substantially impact power production and wind loads on the turbine. This research finds that low-level jets occur often over the North Sea. Moreover, numerical weather prediction models struggle to adequately represent this phenomenon. A climatology based on observations is also biased, because the observations are limited in time and space. This work shows that a reliable climatology combines field observations with output of reanalysis products. Examination of the physics governing low-level jet characteristics reveals the complexity of the wind field in coastal areas, and over the North Sea in particular. The thermal contrast between land and sea plays a pivotal role in the manifestation of the wind field. This feature is not unique for the North Sea, but the relatively small scale and complex coastlines in this area pose an extra challenge in describing the local climatology. It is shown that the newest generation of downscaled reanalysis datasets, represented by the Dutch Offshore Wind Atlas and the New European Wind Atlas represent this complex wind field with reasonable accuracy. As the resolution and quality of weather models and reanalysis products increase, direct coupling between meteorological forcing data and wind energy engineering models becomes an appealing alternative to the use of idealized inflow fields. However, this paradigm shift will come with a substantial increase in computational demands and calls for innovative case selection, or dimensionality reduction. The results presented in this thesis provide a starting point to select relevant parameters, and possible directions are extensively discussed.