Matthew A. Lackner

An Analytical Framework for Offshore Wind Farm Layout Optimization

A method is developed for using the levelized cost of energy as the objective function for offshore wind farm layout optimization problems. The method converts the cost of energy into a function of turbine position only. To accomplish this, wind speed data are first characterized by direction sector. Continuous functions are then fitted to the Weibull parameters for each direction sector. The wind direction probability density function and the turbine power curve are also transformed into continuous functions. For each turbine in the farm, the continuous function that describes the Weibull scale parameter can be scaled to reflect wake losses from other turbines. The function may also be adjusted according to the variation in wind speed with fetch. The annual energy production of the farm is thus modeled as a function only of the turbine positions. When combined with wind farm cost estimates, the levelized cost of energy is still only a function of turbine position and can then be used as an objective function within a variety of optimization algorithms.

Uncertainty Analysis in MCP-Based Wind Resource Assessment and Energy Production Estimation

This paper presents a mathematical framework to properly account for uncertainty in wind resource assessment and wind energy production estimation. A meteorological tower based wind measurement campaign is considered exclusively, in which measure-correlate-predict is used to estimate the long-term wind resource. The evaluation of a wind resource and the subsequent estimation of the annual energy production (AEP) is a highly uncertain process. Uncertainty arises at all points in the process, from measuring the wind speed to the uncertainty in a power curve. A proper assessment of uncertainty is critical for judging the feasibility and risk of a potential wind energy development. The approach in this paper provides a framework for an accurate and objective accounting of uncertainty and, therefore, better decision making when assessing a potential wind energy site. It does not investigate the values of individual uncertainty sources. Three major aspects of site assessment uncertainty are presented here. First, a method is presented for combining uncertainty that arises in assessing the wind resource. Second, methods for handling uncertainty sources in wind turbine power output and energy losses are presented. Third, a new method for estimating the overall AEP uncertainty when using a Weibull distribution is presented. While it is commonly assumed that the uncertainty in the wind resource should be scaled by a factor between 2 and 3 to yield the uncertainty in the AEP, this work demonstrates that this assumption is an oversimplification and also presents a closed form solution for the sensitivity factors of the Weibull parameters.

Controlling Platform Motions and Reducing Blade Loads for Floating Wind Turbines

Offshore sites hold great promise for the growth of wind energy. To tap the vast resource in deep water sites, new support structures, such as those that float, are needed. For floating structures to succeed, they must withstand the offshore wind and wave environment. Two new methods for controlling a floating turbine and reducing the platform and blade loads are presented. The first is a method for controlling collective blade pitch and reducing platform pitch motion, a significant problem for floating structures. The rated generator speed is made a function of the platform pitch velocity. When the platform is pitching upwind, the set point generator speed is set to a larger value, and vice versa. For constant generator torque, this approach essentially makes the rated power a variable that depends on the platform pitch velocity. Fundamentally, this control approach trades power variability for platform pitch variability. The results will show substantial reductions in platform pitch motion but minor increases in power variability. Second, an individual blade pitch controller (IPC) designed to reduce blade fatigue loads is implemented for a floating turbine. The IPC approach is commonly utilized for reducing the 1P fatigue loads on the blades. The goal of implementing this IPC approach is to investigate how traditional load reduction control, which is successful for onshore turbines, integrates and performs with floating turbines. The results will demonstrate the unique challenge of reducing blade loads on a floating turbine.

Offshore Wind Turbine Load Reduction Employing Optimal Passive Tuned Mass Damping Systems

Offshore wind turbines can capture the high-quality offshore wind resource but suffer from increased loading from waves and ice. Reducing these loads through structural control techniques has the potential to be an economically viable solution. Both fixed-bottom and floating substructures are considered in this paper, which will investigate a fixed-bottom monopile as well as a barge, spar buoy, and tension-leg platform for the floating platforms. A set of optimum passive tuned mass dampers are developed by creating a limited degree-of-freedom model for each of the four offshore wind platforms. These models are then integrated into an optimization function using a genetic algorithm to find a globally optimum design for the tuned mass damper. The tuned mass damper parameters determined by the optimization are applied to a series of wind turbine design code simulations using FAST. A sensitivity analysis of the tuned mass damper parameters and a study on the effect of misaligned wind and waves on load reductions are also conducted. Results from these simulations are presented, and tower fatigue damage reductions of up to 20% are achieved for the various tuned mass damper configurations.

Uncertainty Analysis in Wind Resource Assessment and Wind Energy Production Estimation

This paper presents a mathematical approach to properly account for uncertainty in wind resource assessment and wind energy production estimation. The evaluation of a wind resource and the subsequent estimation of the annual energy production (AEP) is a highly uncertain process. Uncertainty arises at all points in the process, from measuring the wind speed to the uncertainty in a power curve. A proper assessment of uncertainty is critical for judging the feasibility and risk of a potential wind energy development. Many current methods for assessing uncertainty either oversimplify the process or make faulty assumptions, leading to erroneous estimates of uncertainty. The approach in this paper yields a more accurate and objective accounting of uncertainty, and therefore better decision making when assessing a potential wind energy site. Three major aspects of site assessment uncertainty are presented here. First, a method is presented for combining uncertainty that arises in assessing the wind resource. Second, uncertainty in wind turbine power output and energy production is characterized. Third, a method for estimating the overall AEP uncertainty when using a Weibull distribution is presented. While it is commonly assumed that the uncertainty in the wind resource should be scaled by a factor between two and three to yield the uncertainty in the AEP, this work demonstrates that this assumption is an oversimplification, and also presents a closed form solution for the sensitivity factors of the Weibull parameters.

Probabilistic Models for Wind Turbine and Wind Farm Performance

A Markov model for the performance of wind turbines is developed that accounts for component reliability and the effect of wind speed and turbine capacity on component reliability. The model is calibrated to the observed performance of offshore turbines in the north of Europe, and uses wind records obtained from the coast of the state of Maine in the northeast United States in simulation. Simulation results indicate availability of 0.91, with mean residence time in the operating state that is nearly exponential and has a mean of 42 days. Using a power curve typical for a 2.5 MW turbine, the capacity factor is found to be beta distributed and highly non-Gaussian. Noticeable seasonal variation in turbine and farm performance metrics are observed and result from seasonal fluctuations in the characteristics of the wind record. The input parameters to the Markov model, as defined in this paper, are limited to those for which field data are available for calibration. Nevertheless, the framework of the model is readily adaptable to include, for example: site specific conditions; turbine details; wake induced loading effects; component redundancies; and dependencies. An on-off model is introduced as an approximation to the stochastic process describing the operating state of a wind turbine, and from this onoff process an Ornstein–Uhlenbeck (O–U) process is developed as a model for the availability of a wind farm. The O–U model agrees well with Monte Carlo (MC) simulation of the Markov model and is accepted as a valid approximation. Using the O–U model in design and management of large wind farms will be advantageous because it can provide statistics of wind farm performance without resort to intensive large scale MC simulation. [DOI: 10.1115/1.4004273]

Active Structural Control of Offshore Wind Turbines

The application of control techniques to ofishore wind turbines has the potential to signiflcantly improve the structural response of these systems. Active control is investigated, and compared to a previous investigation into passive control by the authors. A limited degree of freedom model is identifled with synthetic data and used to design a family of controllers using H1 multivariable loop shaping. The controllers in this family are then implemented in full degree of freedom time domain simulations. The performance of the controllers is quantifled using the reduction in fatigue loads of the tower base bending moment. The performance is calculated as a function of active power consumption and the stroke of the active mass damper. The results are compared to the optimal passive system, and the additional achievable load reduction using active control is quantifled.

Comparison of the Dynamic Wake Meandering Model, Large-Eddy Simulation, and Field Data at the Egmond aan Zee Offshore Wind Plant: Preprint

The focus of this work is the comparison of the dynamic wake meandering model and large-eddy simulation with field data from the Egmond aan Zee offshore wind plant composed of 36 3-MW turbines. The field data includes meteorological mast measurements, SCADA information from all turbines, and strain-gauge data from two turbines. The dynamic wake meandering model and large-eddy simulation are means of computing unsteady wind plant aerodynamics, including the important unsteady meandering of wakes as they convect downstream and interact with other turbines and wakes. Both of these models are coupled to a turbine model such that power and mechanical loads of each turbine in the wind plant are computed. We are interested in how accurately different types of waking (e.g., direct versus partial waking), can be modeled, and how background turbulence level affects these loads. We show that both the dynamic wake meandering model and large-eddy simulation appear to underpredict power and overpredict fatigue loads because of wake effects, but it is unclear that they are really in error. This discrepancy may be caused by wind-direction uncertainty in the field data, which tends to make wake effects appear less pronounced.

Implementing the Dynamic Wake Meandering Model in the NWTC Design Codes

This paper focuses on implementing the dynamic wake meandering (DWM) model in the National Wind Technology Center (NWTC) design code framework. The DWM model calculates the wake deficit and the meandered wake center behind a wind turbine. The advantage of incorporating the DWM model into the NWTC design codes is to provide the ability to simulate unsteady wake effects and examine their impact on power generation, structural loads, and turbine control schemes, while maintaining an acceptably low computational cost. In the present paper, the method of implementing the DWM model into the NWTC design codes is described. In addition, the performance of the DWM model is verified by comparing the generated turbine power and blade loads with high-fidelity large-eddy simulation (LES) and field data for North Hoyle and Lillgrund wind farm. It was found that the results from the DWM model are in good agreement with the LES results and the field data. Further efforts are in pursuit to allow simulations of a full wind farm with an arbitrary wind turbine layout and incoming flow directions which are briefly described in this paper.

Simulation-Length Requirements in the Loads Analysis of Offshore Floating Wind Turbines

The design standard typically used for offshore wind system development, the International Electrotechnical Commission (IEC) 61400-3 fixed-bottom offshore design standard, explicitly states that “the design requirements specified in this standard are not necessarily sufficient to ensure the engineering integrity of floating offshore wind turbines” [1]. One major concern is the prescribed simulation length time of 10 minutes for a loads-analysis procedure, which is also typically used for land-based turbines. Because floating platforms have lower natural frequencies, which lead to fewer load cycles over a given period of time, and ocean waves have lower characteristic frequencies than wind turbulence, the 10-min simulation length recommended by the current standards for land-based and offshore turbines may be too short for combined wind and wave loading of floating offshore wind turbines (FOWTs). Therefore, the goal of this paper is to examine the appropriate length of a FOWT simulation — a fundamental question that needs to be answered to develop design requirements.To examine this issue, we performed a loads analysis of an example FOWT with varying simulation lengths, using FAST, the National Renewable Energy Laboratory’s (NREL’s) nonlinear aero-hydro-servo-elastic simulation tool. The offshore wind system used was the OC3-Hywind spar buoy, which was developed for use in the International Energy Agency (IEA) Offshore Code Comparison Collaborative (OC3) project, and supports NREL’s offshore 5-MW baseline turbine. Realistic metocean data from the National Oceanic and Atmospheric Administration (NOAA) and repeated periodic wind files were used to excite the structure. The results of the analysis clearly show that loads do not increase for longer simulations. In regard to fatigue, a sensitivity analysis shows that the procedure used for counting half cycles is more important than the simulation length itself. Based on these results, neither the simulation length nor the periodic wind files affect response statistics and loads for FOWTs (at least for the spar studied here); a result in contrast to the offshore oil and gas (O&G) industry, where running simulations of at least 3 hours in length is common practice.Copyright