Review of failures and condition monitoring in wind turbine

XIX International Conference on Electrical Machines - ICEM 2010, Rome

Review of Failures and Condition Monitoring in

Wind Turbine Generators

Z. Daneshi-Far, G. A. Capolino, H. Henao

Abstract – Increasing wind power generation quantity in power systems needs obviously reliable operation. Therefore, accurate condition monitoring and fault diagnosis are almost mandatory. This paper aims to report recent works on condition monitoring and fault diagnosis for wind turbine generators. Wind turbines are subjected to different sort of electromechanical systems which extract kinetic power of the wind and convert it into electrical power. By increasing the amount of installed wind turbines, it becomes mandatory to force wind power generation to be a reliable source of energy. Of course, obtaining a reliable power from wind Φ

failures, thus before stating condition monitoring and fault diagnosis methods it is necessary to identify what kind of failures can be found in the real world. As a result, the gearbox is one of the most critical components of the wind turbine in which failures could stop generation for a long time. Recently, several condition monitoring and fault diagnosis techniques have been introduced in order to minimize downtime and maintenance cost while increasing energy availability and life time service of the wind farms. Vibration sensors have been used for long time in wind turbine condition monitoring systems to collect data of the generator health. Actually, using such sensors is expensive, difficult to install and sometimes impossible to be implemented in already installed wind turbines. Therefore, researchers have been focused on using electrical signature analysis from sensors in order to detect wind turbines faults and generator health in a cheap, easy to install and accessible way.

Index Terms - Wind turbine, Condition monitoring, Fault diagnosis, Predictive maintenance, Failure in wind turbines,

Vibration, Gearbox, Shaft, Bearings, Spectral methods,

Wavelet analysis.

W

I. INTRODUCTION

ind power generation is highly increasing in all the

world and especially in France. Europe aims for 20%

renewable power generation by 2020 in order to

follow the Kyoto protocol. France planed to increase

renewable power generation to 23% by 2020 from which

wind generation targets would be 19GW off-shore wind

turbines and 6GW on-shore which means increasing about

six times of installed wind energy capacity on 2009 [1]. At

the beginning of 2009, wind power generation in France has

been 3,404MW which cover 3.9% of electric power

consumption of the country. At the end of 2009, it has been

increased to 4,492MW which means 32% wind power

generation increase in the year 2009 while wind power

generation ratio in world is 31% for 2009 [2].

Wind turbines are subject to different sort of failures.

Therefore, before stating condition monitoring and fault

diagnosis methods in wind turbines, the different kinds of

failures as well as their downtime consequences have to be

reviewed [3]-[7]. Wind turbines are complex

This work has been supported by the research funds of the Regional Council of Picardie, Amiens, France. Z. Daneshi-Far, G. A. Capolino and H. Henao are with the Laboratory

of Innovative Technologies – Power Group, University of Picardie, 80000 Amiens, France (e-mails: Zahra.Daneshifar@,

Gerard.Capolino@, Humberto.Henao@ ).

978-1-4244-4175-4/10/$25.00 ©2010 IEEE

turbine generators which are subject to many types of failures, leads to the need of high performances condition monitoring systems.

Wind turbine condition monitoring systems allow collecting data from the main components of a wind turbine such as the generator, the gearbox, the main bearing, the shaft and the yaw system. The purpose is to minimize downtime and maintenance cost while increasing the energy availability and the life time service of wind turbine components. An ideal condition monitoring system is supposed to monitor all the components by using a minimum number of sensors.

The selection of suitable monitoring parameters in wind turbines is a matter of fact in which several works have been performed [8]. The main existing monitoring systems are

using vibrations as indexes to detect fault in the global system [9], [10]. In this way, several vibration sensors should be installed on the generator, the gearbox, and the

bearings. Therefore, the vibration-based monitoring system is costly and it is difficult to be implemented considering that it is designed with numerous vibration sensors [11]-[13]. Consequently, some recent publications have recommended

the usage of the electrical machine parameters in order to

recognize both early electrical and mechanical faults in the

wind turbine while it is generating and without stopping it

[14].

Recently, lot of works has been done in order to use

electrical parameters for monitoring the electrical machine

health [15]-[18]. In wind turbines, because of the noisy

environment due to the presence of power electronics,

signal to noise ratio of measured signals is low and the usage

of proposed techniques is often more problematic than in lab

environment [8], [19]. Several contributions have been

proposed to overcome this drawback by using proper signal

processing techniques [20]-[26]. Some of them focus only

on the electrical machine fault diagnosis [20]-[22] while

other focus on both bearings and gearboxes [23]. Other

contributions are dealing with detection of electrical and

mechanical faults from the generator electrical

measurements [24]-[26].

Several reviews on wind turbine condition monitoring have been already published [27]-[29]. …… 此处隐藏：5823字，全部文档内容请下载后查看。喜欢就下载吧 ……