Search
Close this search box.
Search
Close this search box.

Modelling Wind Turbine Failures based on Weather Conditions

Wind turbine failure models can help to understand the components’ degradation processes and enable the operators to anticipate upcoming failures. Usually, these models are based on the age of the systems or components. However, our research shows that the on-site weather conditions also affect the turbine failure behaviour significantly. At ANNEA we follow a novel approach for modelling wind turbine failures based on the environmental conditions to which they are exposed to. Taking into account the environmental conditions that are more likely to provoke component failures. This could enhance predictive maintenance models significantly. 

It has been proven that not only the turbine age, but also certain combinations of weather conditions can affect their life-time negatively. During the useful life-time, the assumption of constant failure rates does not always hold true, especially when considering shorter time-intervals such as the failure occurrences on a monthly basis. There are significant variations in failure rates throughout the year that need to be taken into account by the wind farm operators. With this they can react properly to upcoming failure by initiating preventive or opportunistic maintenance actions. 

The industrial research has proved that the failure behaviour of wind turbines and their components is strongly influenced by the meteorological conditions the systems are exposed to. Nonetheless, no models have been developed yet to directly describe the wind turbine’s failure behaviour based on combinations of external covariates. 

In this article we describe a novel approach to model the failures of wind turbines during the useful life including the effect of environmental conditions. The model was applied to a 2 MW case study wind farm and the failures of the whole turbine system, as well as four main components. Failures are defined as events that can be associated to a component breakdown, which causes a wind turbine’s stop and needs intervention such as replacement or repair. 

A regression model based on a generalised linear model (GLM) is applied to the data. The model is set up with a Poisson response distribution and a logarithmic link function. Subsequently, a ridge regression is employed to estimate the model parameters. 

The model is applied to a case study including 30 turbines located in a wind farm in Spain. The wind turbines are geared, three bladed and pitch-regulated machines, with a rated capacity of 2MW each. The data set consists of failure logbooks, SCADA and meteorological tower (met-mast) data collected during three years of operation. The model covariates include the monthly average wind speed (WS) and turbulence intensity (TI), and the monthly maximum wind speed (MaxWs), measured at a height of 45 meters at the wind farm met-mast. Also included are the monthly mean relative humidity (RH), precipitation (Rain) and ambient temperature (Temp) taken from closely located weather stations. Additionally, as an indicator for the time operating at full capacity, the monthly mean power production (PWR) taken from the turbines SCADA systems divided by the rated capacity was chosen. 

At the start of the data collection the turbines were five years old. An average wind farm year is modelled, whereas the observation period is introduced into the model by means of an exposure variable called model offset. Thus, the model outcome can be considered as the rate of failure occurrence in an average operational year. 

In a first step the model is applied to the whole data base, without further distinguishing between the failed components. Subsequently, the failure data of four main components: the gearbox, generator, pitch and yaw system are extracted from the same set and the model is applied again. To analyse the importance of each input variable, the standardised model coefficients are compared. This is commonly done to interpret which of the covariates contributes the most to modelling the output and helps to see which weather conditions are important for modelling failures. 

A summary of the model performance metrics for the five different cases is shown in the following table:

 

Table reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

The R2 value ranges from 0 to 1 and shows the goodness of fit of the model to the data, where 1 indicates the best fit. It can be seen that in general the model performs well for all five failure classes. The models for the generator and pitch system, however, showed lower R2 and higher error values than the other ones. This indicates that for these two components additional covariates, which were not included in the model, could be of importance. This will be assessed in further studies. 
 

The figure below shows the original failures (black) and the modelled data (green) for wind turbine system failures without distinguishing between their components.

Source: https://iopscience.iop.org/article/10.1088/1742-6596/926/1/012012/pdf

Figure reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

The values are normalised to the maximum number of failures for confidentiality reasons. As an example for a separately modelled wind turbine’s component, Figure 2 displays the original and the modelled data for yaw system failures:

Source: https://iopscience.iop.org/article/10.1088/1742-6596/926/1/012012/pdf

Figure reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

It can be seen that in both cases the highest failure occurrences are recorded within the second to fourth month of the year. This is consistent with the previous research which states that failures mainly occur during the winter months and/or the transition periods between seasons. 

Figures 3 to 7 show the standardised coefficient magnitudes. Positive coefficients (blue) indicate that with increasing covariates the output increases as well, while negative coefficients (orange) respectively describe the opposite behaviour. As the input data was centered and scaled, these coefficients can be interpreted as the influence of every model covariate on the dependent variable. For an easier interpretation the plots show the magnitudes as fraction of the most important regressor, which is scaled to 1. 

Source: https://iopscience.iop.org/article/10.1088/1742-6596/926/1/012012/pdf

Figure reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

As shown in Figure 3 the input parameters TI, Temp and WS have the highest importance for modelling the data including failures of any wind turbine’s component. High monthly mean turbulence intensity, high mean wind speed and low temperatures play a significant role. This is consistent with the previous studies that stated that higher mean wind speeds and low temperatures can be correlated to higher number of wind turbine’s failures. Low PWR values seem to influence the failure behaviour as well, but not very dominantly. As the active power output is usually positively correlated to wind speed, having positive coefficient magnitudes for wind speed and negative ones for the PWR variable this might seem contradictory. However, under faulty conditions the wind turbine is often performing below the expected capacity and under-performance can be seen as indicator for component failures. 

Figure reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

The different components react differently to certain combinations of environmental conditions. Thus, these should be analysed separately in order to obtain more meaningful results. 

For the pitch system model (Figure 4) low temperatures and high monthly maximum wind speeds are significant. In addition to that, high relative humidity and turbulence intensity play a role when modelling the pitch system failures. These are the conditions, where the pitch system is mostly active in order to regulate the rotor speed and thus is subject to higher stresses and possible damages. 

Figure 5 shows the standardised coefficient magnitudes for yaw system failures. The differences between the coefficient magnitudes are not as large as they are for the other wind turbine’s components. Many meteorological factors seem to play a role in this model. A clear under-performance of the turbine can be seen, as rising wind speeds and falling PWR values lead to increased numbers of yaw system failures. This leads to the assumption that despite the higher mean wind speeds, the wind direction changed frequently and the yaw system had to be constantly searching for the best wind direction. Thus, a possible yaw-misalignment resulted in higher wear and under-performance. However, this should be investigated in more detail by including the wind direction in further studies. 

Figure reference: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012). View the article online for updates and enhancements.

Modelling the number of gearbox failures is influenced mostly by high wind speeds, as displayed in Figure 6. With higher mean wind speeds, the load on the gearbox is increasing and the component is more likely to fail. 

The generator failure model (Figure 7) is mostly driven by increasing turbulence intensity and power output. Elevated TI introduces higher loads on the generator that has to adapt to these varying input speeds. Positive coefficients for the variables PWR and WS indicates that no under-performance was recorded before the failures. Additionally, it states that with higher power production, the generators are more likely to fail. Generator failures usually occur abruptly due to sudden changes in turbulence intensity and highly varying wind conditions. Furthermore, the amount of precipitation plays a significant role, as water intrusion highly affects electronic equipment. 

Hence, the herein presented models serve to identify which environmental parameters influence the failure behaviour of certain wind turbine components. This information can help to anticipate failures and significantly enhance predictive maintenance models.

This is a summary of the original article: “Modelling Wind Turbine Failures based on Weather Conditions” by Maik Reder and Julio J. Melero (2017 J. Phys.: Conf. Ser. 926 012012).

View the article online for updates and enhancements.
4.5 2 votes
Article Rating
Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x
A note to our visitors

This website has updated its privacy policy in compliance with changes to European Union data protection law, for all members globally. We’ve also updated our Privacy Policy to give you more information about your rights and responsibilities with respect to your privacy and personal information. Please read this to review the updates about which cookies we use and what information we collect on our site. By continuing to use this site, you are agreeing to our updated privacy policy.

Click here to view our updated Privacy Policy

en_USEnglish

Book A DEMO


How can our predictive maintenance and asset optimisation platform do to make your day-to-day activities easier?

With ANNEA’s domain knowledge in the renewable industry, engineering expertise, and predictive maintenance modelling techniques, we reduce operational costs while driving performance improvements, increasing energy production.

– No extra hardware required
– Using only 10-min SCADA Data
– Brand agnostic
– All-in-one Platform

Submit and our team will get in touch with you to schedule a platform demonstration.

Book A DEMO

How can our predictive maintenance and asset optimisation platform do to make your day-to-day activities easier?

With ANNEA’s domain knowledge in the renewable industry, engineering expertise, and predictive maintenance modelling techniques, we reduce operational costs while driving performance improvements, increasing energy production.

– No extra hardware required
– Using only 10-min SCADA Data
– Brand agnostic
– All-in-one Platform

Submit and our team will get in touch with you to schedule a platform demonstration.

Want to join annea?

 

Interested in joining our team? We’re always on the lookout for talented individuals who are passionate about making a difference in the renewable energy sector.

If you believe you have what it takes to contribute to our innovative projects and drive positive change, we’d love to hear from you.

Please fill out the form to submit your unsolicited application. Let’s work together to create a sustainable future!

Want to ANNEA?

Interested in joining our team? We’re always on the lookout for talented individuals who are passionate about making a difference in the renewable energy sector.

 

Interested in joining our team? We’re always on the lookout for talented individuals who are passionate about making a difference in the renewable energy sector.

If you believe you have what it takes to contribute to our innovative projects and drive positive change, we’d love to hear from you.

Please fill out the form below to submit your unsolicited application. Let’s work together to create a sustainable future!