Methods to improve wind turbine generator bearing temperature imbalance for onshore wind turbines

Abstract

For better annual energy production, wind turbine generator components are expected to perform efficiently and safely. Development of recent high-efficiency generators and motors leading their designs with less cooling capacity. Bearings are one of the most stressed components in the generator. Recent studies have indicated that bearing failure is the prime cause of generator failure, in wind turbine application. Grease lubrication deterioration was found to be the leading cause of motors and generators bearing failure. Grease service life for generators are closely associated with the operating temperature. One issue with the less cooling design is the higher bearing temperature. This led to marginal lubrication, premature bearing failure, and reduce generator reliability. To verify this and address the issue of inadequate and imbalanced bearing cooling; this paper presents recent experimentation performed on-air to air-cooled squirrel cage induction generator. This test addresses a potential issue with the IC6A1A6 cooled generator design and recommends the updates with corresponding standards. To get optimal bearing life and generator reliability, either allowed bearing operating temperature range should be reduced significantly, by developing a new cooling strategy or standards committee should come up with different intervals of lubrication for both ends of the bearings. Issues similar to with IEEE standard 841 high-efficiency motors can be avoided, where a study performed on the reliability of these motors used in refinery confirmed that motors are at greater risk. A proactive plan based on the result and recommendations in this paper will help to secure the safe wind turbine-generator operation.

Keywords

Wind turbine generator heating bearing greasing

Introduction

A wind turbine generator reliability study is performed and explained in this paper. The study was performed due to the findings by Shipurkar et al. (2015), Alewine et al. (2012), and Liu et al. (2018) that bearing failure to be the main cause of generator failure. Another main reason for performing this research is the recent finding of the new IEEE Standard 841 high-efficiency motor, which are at greater risk of a drive end bearing failure than the older replaced motor. Generator bearing service life depends on the bearing operating temperature rise and should be determined accordingly. Either allowable temperature rise should be reduced substantially (Singh et al., 2018) or different lubrication cycles should be chosen for DE (Drive-end) and NDE (Non-Drive-end) bearings. Generator operating at higher temperature leads to increase DE and NDE operating temperature. Higher bearing grease temperature reduces bearing and grease service life. Bearing and grease will be impacted more depending on which side bearing facing higher operating temperature. This paper is recommending three following changes to be included in respective standards, which are further clarified in the conclusion section. Based on the ambient (nacelle) temperature:

Different lubrication strategy should be defined for DE and NDE bearings

DE and NDE bearings should be rated for different lifetime

Allowed bearing temperature rise limits should be determined by NDE bearing temperature

DFIG (Doubly fed induction generator) and SCIG (Squirrel cage induction generator) are the two most commonly used generators for the onshore geared wind turbine. This paper discusses the issue with IC6A1A6 air to air-cooled SCIG. Winding insulation class and bearing insulation plays an important role in determining the thermal/electrical capability of the generator. Bearings and windings operating temperature determine the lifetime of the generator. Challenges associated with IC6A1A6 air to air-cooled generator has been discussed in detail. A solution is proposed to substantially reduce generator winding and bearing temperature (Singh et al., 2018). In this paper Issue of bearing temperature difference at drive end (DE) and non-drive end (NDE) in IC6A1A6 air to air-cooled generator has been addressed. An investigation has been done on SCIG with IC6A1A6 (as per IEC 60034-6) cooling configuration in a wind turbine. Thirty sensors were installed at a different location within the generator to monitor bearings and windings temperature for a while. Results demonstrated the temperature difference between NDE and DE. Results achieved are in accordance as predicted in theory. The paper has been organized in the following way: Section II is the literature review on the overview of wind turbine component failure frequency. This section discusses the generator operation challenges from an overall turbine point of view. Section III explains bearing failure studies, and section IV explains operating temperature and grease service life issues. This is one of the core issues addressed in this paper for wind turbine generators. Section V discusses the issue with the IC6A1A6 generator theoretically, which is justified through achieved results in the following Section VI. Section VII discusses the proposed grease lubrication multiplier and its bearing temperature dependence strategy. Section VIII is the conclusion and the recommendation to address this issue in the standards for the safer operation of the wind turbine generator.

Wind turbine components failure frequency

Totally enclosed air to air-cooled (TEAAC) SCIG is a widely accepted generator for onshore geared wind turbine application. The nacelle is a compact environment to avoid contaminations and foreign objects entry. It is preferred to seal the nacelle properly, as the performance of most of the component are dependent on the temperature profile they are operating. An Investigation on the reliability of more than 6000 modern onshore wind turbines and their subassemblies in Denmark and Germany over 11 years has been done (Shipurkar et al., 2015; Spinato et al., 2009). Particularly investigation has been done for the changes in the reliability of generators, gearboxes, and converters in a subset of 650 turbines in Schleswig Holstein, Germany. Power electronic converter, the generator, and the gearbox account for approximately 0.25, 0.15, and 0.15 failures per turbine per year respectively (Shipurkar et al., 2015).

The generator is the main subassembly that is susceptible to failure. Following are the two failure modes in generator (Tavner, 2008):

Electrical Failure: core insulation; stator winding or insulation; rotor winding or insulation; brush gear; slip ring; commutator

Mechanical Failure: Bearing; rotor mechanical integrity; stator mechanical integrity

A quantitative review of the failure of more than 12,000 wind turbine generators has been done (Alewine and Chen, 2012). Fewer than half of the failures were electrical in nature. Electrical windings failure is one of the major reasons for generator failure. An overview of conditioning monitoring of rotating electrical machines has been done and potential root causes found to be ambient conditions, the excessive temperature in windings and bearings (Tavner, 2008). A study on a wind specific 12,000 generators for faults distribution over the sub-generator components found that failure of bearing and stator are major contributing factor (Shipurkar et al., 2015). This investigation is presented in Figure 1 (Shipurkar et al., 2015). The fault tree for the generator is presented in Figure 2 (Shipurkar et al., 2015).

Figure 1.

Distribution of faults over the components of generator.

Figure 2.

Fault tree for generators.

Few faults have been marked with red rectangles, where the temperature is assumed to be the root cause. Generator cooling is a well-known challenge for decades.

For better performance, one option is either to make the system capable enough to handle extreme temperatures or reduce the experienced temperature substantially. Various methods have been developed for advanced cooling as mentioned in Kilbourne and Holley (1956) and Singh et al. (2018). With the advancement in High-temperature superconductor technology, the generators can operate at extreme temperatures for a wind turbine of nearly 10MW (Jeong et al., 2017).

Bearing failure issues

Bearing failure is one of the most common reasons for the downtime associated with wind turbine generator failure. There are two aspects of the investigation that is, prevention and detection. Vibration measurement is the most commonly used bearing conditioning monitoring technique (Tandon and Choudhury, 1999). A good review for acoustic and vibration conditioning monitoring is provided in Tandon et al. (2007) and Stack et al. (2004, 2006). Mainly vibration analysis is performed to address bearing failure mechanical in nature only. As far as prevention is concerned, generator bearing can also be heated due to multiple causes. Bearing and common mode current are few electrical root causes. A bearing failure detection model by just measuring shaft voltage and current is presented by Asefi and Nazarzadeh (2018). Bearing insulation plays a critical role in high frequency circulating bearing current. A good comparison on predicting the reduction on circulating current with different insulation thickness is presented in Muetze and Binder (2006a). This paper presents the importance of proper greasing and operating conditions for drive-end and non-drive end bearing. Greasing interval plays a critical role. Grease dielectric strength breaks when the rotor potential grows with respect to the ground due to common-mode voltage. Discharge current leads to issues like fluting and pitting. This affects the life of the bearing (Busse et al., 1997; Muetze and Binder, 2006b; Magdun and Binder, 2013; Muetze et al., 2011; Von Jouanne et al., 1998).

Operating temperature and grease service life

Greasing service life depends on bearing size and grease operating temperature. Optimal grease service life for any given bearing size can be achieved in a narrow temperature region. Grease temperature inside bearing housing can be 11°C higher than operating bearing temperature (Gerstenkorn and Somes, 2016; NSK Americas, 2015; SKF, 1998). Figure 3 represents operating grease service life in a 40–55°C operating range. A 15°C higher temperature than 55°C will reduce the grease service life in half.

Figure 3.

Grease service life (SKF, 2003).

Figure 4 represents the estimation service life of grease in a lubed-for-life application. This provides important information on how to grease service life depends on the speed and diameter of the bearing. This also provided input to our study, on how frequently bearing should be lubricated. The re-lubrication strategy should follow the operating ranges of bearing. A recommendation is made, where bearing lubrication interval changes with operating temperature exponentially (Gerstenkorn and Somes, 2016; SKF, 2011; Tech Talk Publication, 2009). One of the failure analyses of bearing in a wind turbine has been presented by Sankar et al. (2012). Singh et al. (2018, 2019a, 2019b, 2019c, 2019d, 2020) and Singh and Sundaram (2020) performed similar studies on the impact of heat, improper filtration, and converter on winding and bearing of electrical generator. Manufacturing deformation leading winding and bearing issues have been presented in detail by Singh and Sundaram (2020). Study presenting an Intermediate stage non-drive end bearing failure and damage of outer race of the bearing presented (Sankar et al., 2012).

Figure 4.

Bearing temperature difference in IC6A1A6 generator.

Theory

A totally enclosed air to air-cooled solution is a well-known generator cooling solution for the wind turbine application. SCIG is a widely used generator with this configuration. Figure 4 shown below represents air to air-cooled generator solution and based on IEC60034-6 is represented as IC6A1A6. IC6A1A6 has two cooling systems that is, internal and external. An internal cooling system is a closed-loop where a shaft-mounted fan helps to recirculate air inside the generator and transfers the heat from the generator stator core into an air to air heat exchanger. An external cooling system is achieved with the help of an electrical fan. Where the flow is created through the nacelle and the heat exchanger. Heat is removed in the form of hot air outside nacelle. One of the challenges is to keep uniformity of the temperature inside the generator during operation.

This diminishes the possibility that some areas are more thermally stressed than others. One of the challenges of the IC6A1A6 solution is the temperature difference inside the generator. Based on airflow, the drive end is mentioned in Figure 4 with a blue arrow and assumed to be lower temperature. The Brown arrow represents the medium temperature. Red arrow and associated bearing areas are assumed to be higher temperature. This flow can be described as a flow inside a tube with constant heat flux in the external wall. The following equation represents heat exchange between cooling air and generator internal surface as

Q = UA Δ T

(1)

Where,

Q Heat transferred between surface and air [W]

U Overall heat transfer coefficient [W/ $m^{2} . ° C$ ]

A Area across where heat is transferred [ $m^{2}$ ]

$Δ$ T Temperature difference between generator internal surface and cooling air [ $° C]$

By using log mean temperature difference, equation can be rewritten as

Q = UA Δ T_{m}

(2)

Where $Δ T_{m}$ can be described as follows

Δ T_{m} = \frac{(T_{h} - T_{a_{1}}) - (T_{h} - T_{a_{2}})}{\ln (T_{h} - T_{a_{1}}) - (T_{h} - T_{a_{2}})}

(3)

$θ$ Generator windings hot temperature

$T_{a_{1}}$ Cooling air inlet temperature

$T_{a_{2}}$ Cooling air outlet temperature

$T_{h}$ Hot winding temperature

From this equation, we can see lower the inlet air temperature represented via blue arrows as represented in Figure 4 leads to higher heat transfer. Generator losses are constant, so heat load is constant. The drive-end part of the generator should face a lower temperature as Ta1 is lower. Drive end bearing is located in that area and it should also be lower in temperature based on the theoretical model. Winding located in the drive-end area should face lower temperature comparative to the middle and non-drive end area. Ta2 will always be higher in value due to losses inside the generator. Bearing and windings at the non-drive end should face comparatively higher temperatures than the drive end. For better performance of the generator, It is usually recommended not to have a big temperature difference between the drive end and non-drive end. As per National Electrical Manufacturers Association (NEMA) classification Generator winding for class F insulation system maximum allowed hot spot temperature is 155°C. For the onshore turbine operation with a full converter, temperature rise of approximately 15°C in generator windings due to converter harmonic losses should also be taken into account, and few C should be reserve for this purpose. Operating the same generator at a higher load point with the same cooling system and same design lead to higher losses in the generator. This will eventually lead to a higher temperature rise in the windings and higher bearing temperature.

Test procedures and results

For testing bearing and winding temperature and mainly to monitor temperature difference between drive and non-drive end of the squirrel cage induction generator, 30 temperature sensors have been installed on the generator. Figure 5 represents different locations across generator stator winding and bearing, where temperature monitoring sensors were installed at a different location. Turbine has been monitored for three weeks and the following test results were obtained:

Figure 5.

Temperature sensors location on squirrel cage induction generator.

Drive end and non-drive end bearings temperatures

Two specific temperature sensors were placed at the non-drive end and drive end bearing, each to monitor the temperature variation. The turbine was operating, and temperatures were monitored for regular three days. Figure 6 represents the bearing temperature difference between NDE and DE of squirrel cage induction generator with IC6A1A cooling configuration as represented in the schematic of Figure 4.

Figure 6.

NDE and DE bearing temperature difference in IC6A1A6 generator.

As predicted in theory and represented in Figure 4 non-drive end bearing temperature appears to be higher during all three days of testing. During high wind and full production, NDE bearing winding reached up-to 80°C and DE temperature is at 44°C. The maximum temperature difference between NDE and DE bearing is 36°C. Test data confirms that in wind turbine operation bearing temperatures will be different at NDE and DE with IC6A1A6 solution. During the three days of testing maximum temperature encountered was 80°C which can be even higher with a higher generator load point and in extreme load cases. The temperature difference between the two ends eventually might lead to thermal imbalance in the system, and certain areas of electrical windings would be overheated. One main conclusion of this study is NDE bearing will be facing higher thermal stress during lifetime comparative to DE bearing in IC6A1A6 cooling solution, and hence it needs extra cooling or lubrication.

Phase windings temperatures

Phase winding in Figure 5 is located at positions 28, 29, and 30 which are almost at the same position. As presented in Figure 7 temperatures of the phase winding U, V and W closely follow each other. During the three days of testing maximum temperature encountered was 114°C which can be even higher with extreme generator load point, in the case of a fully rated converter temperature experienced by windings will be even higher due to harmonic losses. As per NEMA classification, the maximum allowed hot spot temperature for generator winding class F insulation system is 155°C and if temperature rise reaches or exceeds its limit then either a new cooling system needs to be developed or machine needs to be operated at a lower load point.

Figure 7.

Phase windings temperatures.

Grease lubrication interval multiplier and its bearing temperature dependence

Re-lubrication is an important process that keeps bearing functional under safe operation limits. Bearing lubrication interval multiplier and its dependence on temperature for a particular bearing type has been demonstrated (Gerstenkorn and Somes, 2016; SKF, 2011; Tech Talk Publication, 2009). What happens if the non-drive end and drive- end bearing operates under different temperature conditions? Surely bearing lubrication interval multiplier should not be the same, otherwise, it might affect the lifetime of the bearing. This issue has been discussed in this paper for the generator in wind turbine application. IEC60034 is the IEC standard for rotating electrical machines and part 6 that is, IEC60034- 6 is dedicated to various methods of cooling (IC code). IC6A1A6 cooling that is, air to air-cooled generator have different bearing temperature as presented through results in Section V. Different operating conditions for the bearing leads to different lubrication interval multiplier for both DE and NDE bearing. Figure 8 representing a situation where DE and NDE bearing face a temperature difference (T) of 36°C. DE bearing is 36°C cooler than NDE bearing. From Figure 8 it is clear that the ratio of the re-lubrication interval at 82.16°C for DE and NDE is approximately 1:5. Ratio grows up further exponentially if operating temperature increases. This situation should definitely be taken into account for the safe generator operation.

Figure 8.

Bearing operating temperature versus re-lubrication interval multiplier.

Conclusion

Paper explained the investigations performed in regard to the impact of imbalanced wind turbine generator cooling on reliability. Paper addressed the areas where the electric generator is susceptible to failure. Windings and bearing turned out to be two major components within the generator which are responsible in the majority of cases leading to failure. Temperature distribution across generator and its relation to bearing and winding failure discussed. Squirrel cage Induction generator configuration with heat exchanger and IC6A1A6 cooling configuration was chosen for the investigation. The theory explained the functioning and provided a prediction that the existing configuration might lead to the difference in bearing operating temperature. This was justified with testing results at a wind turbine testing facility for three days. NDE bearing encounters higher temperatures throughout the three days operation. The highest temperature difference between DE and NDE bearing was found to be 36°C. Although, all phase windings follow almost the same temperature profile. It can be concluded that NDE bearing needs additional cooling or additional lubrication cycles. This will help NDE bearing to overcome thermal stress during lifetime and during extreme load cases. It is found through results that the NDE bearing lubrication cycle should be approximately five times more than DE bearing lubrication cycle (at 82°C).

Following are the recommendations:

Different lubrication strategy should be defined for DE and NDE bearings (Re-lubrication interval could be the same only when both bearings are operating at the same temperature)

DE and NDE bearings should be rated for a different lifetime, as both are operating under different temperature conditions

Allowed bearing temperature rise limits should be selected based on NDE bearing temperature, as NDE bearing is facing the highest temperature

Implementation of the above-mentioned recommendations could avoid failure and facilitate better and safe performing air to air-cooled squirrel cage IC6A1A6 induction generator, for wind turbine applications.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Gopal Singh

References

Alewine

Chen

(2012) A review of electrical winding failures in wind turbine generators. IEEE Electrical Insulation Magazine 28(4): 8–13.

Asefi

Nazarzadeh

(2018) A fast transient model for bearing fault analysis in induction machine drives. IEEE Sensors Journal 19(5): 1897–1904.

Busse

Erdman

Kerkman

, et al. (1997) The effects of PWM voltage source inverters on the mechanical performance of rolling bearings. IEEE Transactions on Industry Applications 33(2): 567–576.

Gerstenkorn

Somes

(2016) Impacts of reduced motor cooling on reliability. IEEE Transactions on Industry Applications 53(1): 739–744.

Jeong

J-S

D-K

Hong

J-P

, et al. (2017) Design of a 10-MW-class HTS homopolar generator for wind turbines. IEEE Transactions on Applied Superconductivity 27(4): 1–4.

Kilbourne

Holley

(1956) Liquid cooling of turbine-generator armature windings. Electrical Engineering 75(5): 436–441.

Liu

Ren

, et al. (2018) Comparative analysis of bearing current in wind turbine generators. Energies 11(5): 1305.

Magdun

Binder

(2013) High-frequency induction machine modeling for common mode current and bearing voltage calculation. IEEE Transactions on Industry Applications 50(3): 1780–1790.

Muetze

Binder

(2006a) Calculation of influence of insulated bearings and insulated inner bearing seats on circulating bearing currents in machines of inverter-based drive systems. IEEE Transactions on Industry Applications 42(4): 965–972.

10.

Muetze

Tamminen

Ahola

(2011) Influence of motor operating parameters on discharge bearing current activity. IEEE Transactions on Industry Applications 47(4): 1767–1777.

11.

Muetze

Binder

(2006b) Don’t lose your bearings. IEEE Industry Applications Magazine 12(4): 22–31.

12.

NSK Americas (2015) Roller Bearing Life in High Temperatures. Ann Arbor, MI: NSK Motion and Control Tech Talk Publication.

13.

Sankar

Nataraj

Raja

(2012) Failure analysis of bearing in wind turbine generator gearbox. Journal of Information Systems and Communication 3(1): 302.

14.

Shipurkar

Polinder

, et al. (2015) A review of failure mechanisms in wind turbine generator systems. In: 2015 17th European Conference on power electronics and applications (EPE’15 ECCE-Europe), Geneva, Switzerland, 8–10 September 2015, pp.1–10. New York: IEEE.

15.

Singh

(2020) Wind Turbine Generator Overheating Solution. Electronic Theses and Dissertations, p.133. https://stars.library.ucf.edu/etd2020/133

16.

Singh

Sundaram

(2020) Manufacturing deformation impact on the performance of electrical generator for the wind turbine application. Wind Engineering. Epub ahead of print 12 November 2020. DOI: 10.1177/0309524X20968816.

17.

Singh

Lentijo

Sundaram

(2019a) The impact of the converter on the reliability of a wind turbine generatorProceedings of the ASME 2019 power conference. ASME 2019 power conference, Salt Lake City, UT, USA, 15–18 July 2019, p.V001T06A016. ASME. https://doi.org/10.1115/POWER2019-1966

18.

Singh

Matuonto

Sundaram

(2019b) Impact of imbalanced wind turbine generator cooling on reliability In: 2019 10th international renewable energy congress (IREC), Sousse, Tunisia, 2019, pp.1–6. New York: IEEE.

19.

Singh

Matuonto

Amos

, et al. (2019c) System of systems strand tilt analysis perspective on MediuMHigh voltage stator bar and a non-destructive testing case study. In: 2019 IEEE international systems conference (SysCon), Orlando, FL, USA, 2019, pp.1–8. New York: IEEE.

20.

Singh

Saleh

Amos

, et al. (2018) IC6A1A6 vs. IC3A1 squirrel cage induction generator cooling configuration challenges and advantages for wind turbine application. In: Proceedings of the ASME 2018 power conference collocated with the ASME 2018 12th international conference on energy sustainability and the ASME 2018 nuclear forum. Volume 1: fuels, combustion, and material handling; combustion turbines combined cycles; boilers and heat recovery steam generators; virtual plant and cyber-physical systems; plant development and construction; renewable energy systems. Lake Buena Vista, FL, USA. 24–28 June 2018, p.V001T06A002. ASME. https://asmedigitalcollection.asme.org/POWER/proceedings-abstract/POWER2018/51395/V001T06A002/277415

21.

Singh

Sundaram

Matuonto

(2020) A solution to reduce overheating and increase wind turbine systems availability. Wind Engineering. Epub ahead of print 18March 2020. DOI: 10.1177/0309524X20910992.

22.

Singh

Sundaram

Saleh

(2019d) Addressing reduced ingress protection class & proper filter selection for open ventilated (IC3A1) wind turbine generator In: 2019 10th international renewable energy congress (IREC), Sousse, Tunisia, 2019, pp.1–5. New York: IEEE.

23.

SKF (1998) Improved Grease-Life Model for y-Bearings. February 1998. https://www.skf.com/binaries/pub12/Images/0901d196802a2b8f-13728-EN-Y-bearing-and-Y-b-units_tcm_12-129182.pdf

24.

SKF (2003) Improved Grease Life Model for y-Bearings. August 2003. SKF Evolution Mag., Tech. Rep. https://evolution.skf.com/us/improved-grease-life-model-for-y/

25.

SKF (2011) SKF Bearing Maintenance Handbook. Tech. Rep., pp.193–195. https://www.skf.com/binaries/pub12/Images/0901d1968013be94-SKF-bearing-maintenance-handbook---10001_1-EN(1)_tcm_12-463040.pdf

26.

Spinato

Tavner

van Bussel

GJW

, et al. (2009) Reliability of wind turbine subassemblies. IET Renewable Power Generation 3(4): 387–401.

27.

Stack

Habetler

Harley

(2006) Fault-signature modeling and detection of inner-race bearing faults. IEEE Transactions on Industry Applications 42(1): 61–68.

28.

Stack

Harley

Habetler

(2004) An amplitude modulation detector for fault diagnosis in rolling element bearings. IEEE Transactions on Industrial Electronics 51(5): 1097–1102.

29.

Tandon

Choudhury

(1999) A review of vibration and acoustic measurement methods for the detection of defects in rolling element bearings. Tribology International 32(8): 469–480.

30.

Tandon

Yadava

Ramakrishna

(2007) A comparison of some condition monitoring techniques for the detection of defect in induction motor ball bearings. Mechanical Systems and Signal Processing 21(1): 244–256.

31.

Tavner

(2008) Review of condition monitoring of rotating electrical machines. IET Electric Power Applications 2(4): 215–247.

32.

Tech Talk Publication (2009) Guide to Electric Motor Bearing Lubrication. Exxon Mobil Corporation, Tech. Rep.https://www.mobil.com/lubricants/-/media/Files/global/us/industrial/tech-topics/tt-electric-motor-bearing-lubrication-guide.pdf

33.

Von Jouanne

Zhang

Wallace

(1998) An evaluation of mitigation techniques for bearing currents, EMI and overvoltages in ASD applications. IEEE Transactions on Industry Applications 34(5): 1113–1122.