Optimum performance and reliable operation of wind turbines and/or wind farms are crucial parameters for stakeholders such as OEMs, operators, owners and investors. The efficiency and life time of wind power heavily depends on operation and maintenance efforts. Every manufacturer or owner wants to have wind turbines with high reliability, which can be defined as the probability that a wind turbine will perform its designed function for a certain period of time under stated conditions. In this regard, IEC WG26 is now developing a technical specification defining information categories from which reliability metrics can be identified and reported.

In a previous study, a technique for measuring wind speed and direction by using a roll-rotating three-axis ultrasonic anemometer was proposed and verified by wind tunnel tests. In the tests, instead of a roll sensor, roll angle was trimmed to make no up flow in the transformed wind speeds. Verification was done in point of the residual error of the rotation effect treatment. In this study, roll angle was measured from the roll motor encoder and the transformed wind speed and direction on the test section axis were compared with the ones provided to the test section. As a result, up to yaw $20^{\circ}$ at a wind speed of 12 m/sec or over, the RMS error of wind speed was within the double of the ultrasonic anemometer error. But at yaw $30^{\circ}$, it was over the double of the ultrasonic anemometer error. Regardless of wind speed, at yaw $20^{\circ}$ and $30^{\circ}$, the direction error was within the double of the ultrasonic anemometer error. But at yaw $10^{\circ}$ or less, it was within the error of the ultrasonic anemometer itself. This is a very favorable characteristic to be used for wind turbine yaw control.

A damage detection method for the tripod support structure of offshore wind turbines is presented for structural health monitoring. A finite element model of a prototype tripod support structure is established and the modal properties are calculated. The degree and location of the damage are estimated based on the neural network technique using the changes of natural frequencies and mode shape due to the damage. The stress distribution occurring in the support structure is obtained by a dynamic analysis for the wind turbine system to select the output data of the neural network. The natural frequencies and mode shapes for 36 possible damage scenarios were used for the input data of the learned neural network for damage assessment. The estimated damages agreed reasonably well with the accurate ones. The presented method could be effectively applied for damage detection and structural health monitoring of various types of support structures of offshore wind turbines.

Wind lidar application is increasing and its calibration method is required to use wind lidar as an alternative to the meteorological mast. A nacelle lidar calibration method is now being discussed in IEC 61400-50-3 (Wind energy generation systems - Part 50-3: Use of nacelle-mounted lidars for wind measurements), and the method is mainly based on the wind lidar beam line of sight (LOS) wind speed calibration suggested by DTU as DTU E-0020 (Calibrating Nacelle Lidars). In this paper, a LOS wind speed calibration method is introduced and a calibration example performed on Jeju island is presented. The results showed a slope of 1.011 and R2 of 0.997, which means that the LOS wind speed is highly correlated with the reference wind speed and is comparable. But LOS wind speed calibration requires a very long time due to its principle and environmental conditions, and a calibration method that can overcome this problem of uncontrollable environments needs to be developed.

In this study, the annual power production of a wind farm according to obstacles and wind data was predicted for the Garyeok-do wind farm in the Saemangeum area. The Saemangeum Garyeok-do wind farm was built in December 2014 by the Korea Rural Community Corporation. Currently, two 1.5 MW wind turbines manufactured by Hyundai Heavy Industries are installed and operated. Automatic weather station data from 2015 to 2017 was used as wind data to predict the annual power production of the wind farm for three consecutive years. For prediction, a commercial computational fluid dynamics tool known to be suitable for wind energy prediction in complex terrain was used. Predictions were made for three cases with or without considering obstacles and wind direction errors. The study found that by considering both obstacles and wind direction errors, prediction errors could be substantially reduced. The prediction errors were within 2.5 % or less for all three years.

In this study, a three-dimensional Navier-Stokes simulation around Wido Island was performed to analyze the wake effect behind an island. A 10 m/s wind speed and pressure boundary conditions were assigned for the inflow and outflow boundary conditions, respectively. Wido Island was modeled using GIS data. A prevailing wind from the north-northwest direction was determined based on QuikSCAT satellite data. A computational domain of $40km{\times}20km{\times}5km$ covering Wido Island was applied for numerical analysis. Sixty points were specified to extract the wind speed data. A wind speed profile inside the atmospheric boundary layer was compared with a wind profile using a simple power law. It turns out that the wake effect decreases the mean wind speed by 5% more or less, which corresponds to a 14% decrease in wind energy. Thus, the installation of a meteorological mast or development of a wind farm behind Wido Island is not highly recommended.

This paper introduces an active power dependent standard characteristic curve, Q(P) to compensate for voltage variations due to the output of distributed generation. This paper presents an efficient control method of grid-connected inverters by comparing and analyzing voltage variation magnitude and line loss according to the compensation method. Voltage variations are caused not only by active power, but also by the change of reactive power flowing in the line. In particular, the system is in a relatively remote place in a coastal area compared with existing power plants, so it is relatively weak and may not be suitable for voltage control. So, since it is very important to keep the voltage below the normal voltage limit within the specified inverter capacity and to minimize line loss due to the reactive power. we describe the active power dependent standard characteristic curve, Q(P) method and verify the magnitude of voltage variation by simulation. Finally, the characteristics of each control method and line loss are compared and analyzed.

To maximize power generation and reduce the construction cost of a commercial utility-grade wind turbine, the size of the wind turbine should be large. The initial design of the 12 MW University of Ulsan(UOU) Floating Offshore Wind Turbine(FOWT) was carried out based on the 5 MW National Renewable Energy Laboratory(NREL) offshore wind turbine model. The existing 5 MW NREL offshore wind turbines have been expanded to 12 MW UOU FOWT using the geometric law of similarity and then redesigned for each factor. The resonance of the tower is the most important dynamic responses of a wind turbine, and it should be designed by avoiding resonance due to cyclic load during turbine operations. The natural frequency of the tower needs to avoid being within the frequency range corresponding to the rotational speed of the blades, 1P, and the blade passing frequency, 3P. To avoid resonance, vibration can be reduced by modifying the stiffness or mass. The direct expansion of the 5 MW wind turbine support structure caused a resonance problem with the tower of the 12 MW FOWT and the tower length and diameter was adjusted to avoid a match of the first natural frequency and 3P excitation of the tower.

To establish an offshore wind turbine test site, a wind resource assessment of the candidate site is required as a preliminary procedure. The wind resource assessment must be performed with at least one year of wind data. If the assessment is performed with short-term wind data, the results cannot validate the wind conditions of the candidate site. This study performs wind resource assessment of Kokunsangun-do to investigate the wind conditions of the candidate site. The wind data is measured by the Automatic Weather System (AWS) of the Korea Meteorological Administration, located at Maldo. The data is for five years, measured from 2013 to 2017. Measured wind data is statistically processed with a 10-minute average scheme to find out the dominant wind direction and wind power density, with yearly wind speed distribution (Weibull-based). This study contributes to build a database of wind energy resources around Maldo. Also, the results of this study could be used for the establishment of an offshore wind turbine test site.