A review of wind turbine bearing condition monitoring: State of the art and challenges
Introduction
The market forecasts that by 2018 the wind energy cumulative gigawatts (GW) will be 43% higher than of 2015׳s GW, as illustrated in Fig. 1 [1]. Nowadays, wind energy is a growing and reliable renewable energy source in the world. However, the industry still experiences premature main components (MC) failures, which increase the operation & maintenance (O&M) costs [2]. For instance, early cracks in bearings have been detected with less than 3 years of operation or 5–10% of the expected life [3]. Beyond that, it is estimated that the rate of gearbox failure is one incident per 145 turbines per year [4]. As the demand for wind energy continues to grow substantially, reducing O&M costs and improving reliability have become top priorities in wind turbine (WT) maintenance strategies. O&M cost is a priority because it is responsible for 20-30% of the life cycle cost on onshore installations and up to 30% on offshore installations [5]. In addition, reliability is a fundamental attribute that need to be guaranteed [6]. In the past decade the key research collaborations have focused on improving WT reliability, prediction and diagnosing incipient failures of WT sub-assemblies [7]. Manufactures are trying to improve the reliability of their systems, e.g. gearbox, through same configuration or by developing new configurations [8]. Furthermore, bearings cause high impact on cost and reliability. For instance, it has been shown that bearings cause more than 50% of faults on gearboxes, and therefore 50% of all costs associated to these faults [4], [9]. Consequently, wind turbine bearing condition monitoring (WTBCM) has also become a priority to limit premature failures.
WTs are exposed to extremes, variable and around the clock 24/7 weather conditions, which generate rapid changes on temperature, air pressure, wind shear, wind speed and total load. Due to these variations, WTs undergo constantly changing global and local dynamics and loads. Therefore, the components have to work on intense and variable mechanical stress, which may lead to the occurrence of failures. Besides mechanical stress, the bearing temperature and the different temperatures cause temperature stress along the machine, e.g. shaft׳s and gears’ temperature, together with lubrication problems can also accelerate bearings faults [10]. Also, for bearings at different locations the loads and stresses are different. For instance, misalignment in the drive train leads to abnormal loads and accelerates the wear of the bearings placed at that specific location [11]. To avoid such premature failures, it is necessary to develop better WT designs and also apply reliable and cost-effective condition-monitoring (CM) techniques [12]. Due to all these factors, condition monitoring systems (CMS) are expected to grow from 25 to 36.1% in 2015 according to sellers and buyers perspectives respectively [13].
This paper is structured in seven sections. Section 2 describes CM and why it is important to monitor bearings. Section 3 presents the state-of-the-art review on data acquisition techniques for condition monitoring applied to wind turbine׳s bearings. Section 4 emphasizes a literature review on fault detection and diagnoses methods used by different authors. Section 5 presents a literature review on remaining useful life techniques for bearings. Section 6 focuses on technical, financial and operational challenges to implement the CM. Finally, Section 7 presents the concluding remarks.
Section snippets
Condition monitoring definition and importance of bearings
First and foremost, it is important to have a clear understanding of what is CM and how to define it. Although each author presents CM a little bit differently, in general, one can say that CM is a monitoring process or a sensitive tool that focuses on early detection of faults, failures and wear of machinery with the intention to minimize downtimes and O&M costs, and consequently, maximize production [14], [15], [16], [17]. Beyond that, since the CMS detects failure at an incipient phase it
Acoustic Measurement
Acoustic monitoring is similar to vibration monitoring; however, there are some main differences. Vibration sensors are rigid mounted to the component involved, while acoustic sensors are attached to the component by flexible glue with low attenuation. Vibration sensors register the local motion, whereas, the acoustic sensors "listen" to the component through sound level meters [21], [22]. These devices have a microphone that converts pressure levels and variations both internally and
Acoustic Measurement
Oh et al. (2009) showed a study (including experiment simulating lubricant starvation) on the precursor parameters, such as acoustic noise, vibration and lubricant temperature to detect degradation on fan bearings. Acoustic emission was the best precursor parameter to represent the health of dry bearings [43].
Elforjani et al. (2010) applied AE monitoring on slow speed shafts and bearings. Experimental cases with different conditions were executed separately to shaft and bearings. Bearings’ test
Remaining useful life methods for bearings
Dempsey et al. (2011) used oil debris analysis to indicate bearing damage progression and estimate the RUL on roller bearings. Experiments were conduct at a test rig and results show that the oil debris can be used to indicate progression and RUL, but further tests are required due to variance in spall propagation rates [108].
Butler et al. (2012) presented a methodology for the estimation of the remaining useful life (RUL) of the main bearing for a commercial wind turbine. A residual model is
Technical, financial and operational challenges
There are several challenges to do WTBCM, however, we will focus on five main challenges that enclose technical, financial, operational and management issues from CMS purchase up to the WF monitoring stage. Some of these topics were discussed among the participants and during forums on the fourth annual wind farm data management and analysis on November 2014 [115].
- I.
The first of all challenges that a CMS faces is in the purchasing phase because the pay back of the system cannot be “exactly”
Concluding remarks
In conclusion, it could be highlighted that wind turbine bearings should be monitored due to its high impact on downtime and component replacement. Most of gearbox and generators faults are due to bearings failures, and the bearing replacement cost is lower than the gearbox/ generator׳s replacement and downtime cost. Therefore, it is also very profitable to identify and replace faulty bearings.
WTBCM can be done through several commercial systems that use diverse techniques. However, it is
References (118)
- et al.
Reliability assessment of generic geared wind turbines by GTST-MLD model and Monte Carlo simulation
Renew Energy
(2015) - et al.
Performance assessment of wind turbine gearboxes using in-service data: current approaches and future trends
Renew Sustain Energy Rev
(2015) - et al.
Analyzing bearing faults in wind turbines: a data-mining approach
Renew Energy
(2012) - et al.
Condition based maintenance optimization for wind power generation systems under continuous monitoring
Renewable Energy
(2011) - et al.
Integrated modulation intensity distribution as a practical tool for condition monitoring
Appl. Acoust
(2014) - et al.
The application of spectral kurtosis on Acoustic Emission and vibrations from a defective bearing
Mech Syst Signal Process
(2011) - et al.
Cyclostationarity of Acoustic Emissions (AE) for monitoring bearing defects
Mech Syst Signal Process
(2011) - et al.
Natural roller bearing fault detection by angular measurement of true instantaneous angular speed
Mech Syst Signal Process
(2010) - et al.
EEMD-based wind turbine bearing failure detection using the generator stator current homopolar component
Mech Syst Signal Process
(2013) - et al.
Electrical monitoring of crack propagation during quasi-static loading and ballistic impact of alumina plates
J Eur Ceram Soc
(2013)