Early detection of faults and stall effects associated to wind farms

https://doi.org/10.1016/j.seta.2021.101441Get rights and content

Highlights

  • Early detection of faults in wind farms with automatic analysis.

  • Faults mathematically modeled and experimentally verified in 13 wind turbines.

  • Stall effect influence in wind farms and its mathematical description.

Abstract

Wind turbine generator (WTG) is the most common equipment installed around the world, in the renewable energy field, due to lowest cost compared with solar, hydro and geothermal technologies. The main components are the nacelle, blades, generators, gearbox, pitch system, converters, transformers, tower and yaw system for the normal operation [1]. Our research has an exhaustive analysis of the influence of damages in the blades and the nacelle control in thirteen wind turbine generators, in a real wind farm with information from 2019 and 2020, associated to 2,695,680 operation point with wind velocity, active power, wind direction and yaw angle, with an interesting evaluation of the influence of icing in blades. Besides, this experimental study has demonstrated the reduction of the production until 2.5% in early condition, and 23.8% with the total influence of the blade dynamic impact and deficient yaw control, called stall effect with a real measurement of the lost production. Although the last years, many authors have investigated new ways to improve the detection of failures in this technology, some restrictions have been identified during the evaluation in real conditions.

Introduction

In the last decade, the wind turbine generator technology has been spreading around the world, therefore, it has the cheapest energy cost, the new wind farms are growing in a rate of 6.3% [8] although, the energy companies has still difficulties in the maintenance of the wind technology [2]. Other important aspect is the repowering in large power plants, it is required in countries as Brazil, therefore, it has constraints as time and energy restrictions [3]. In large scale, the WTGs are designed with reliability criterion in wind farms, equipment similar near to 3 MW per WTG and 50 to 150 WTGs, in each wind farms. In this context, the maintenance team should evaluate three challenges: Logistics and the supply change associated to the transportation through countries [4], energy conversion systems associated to the technology and modern control systems [5] and finally, the condition monitoring and fault detection, associated to large power plants, in order to provide the energy required in the power purchase agreement with the local government [6].

Since 2020, several authors have investigated about the reliability of the wind farms and failures modes, however, experimental studies should incorporate to validate the new perspectives, in order to solve real problems with solutions associated to maximize the production.

In 2021, associated to the winter season, several wind farms in USA has been affected with ice condition, and faults associated to yaw misalignment caused by ice with and without heating system [18]. Therefore, the early detection of faults is a priority, in order to normalized the production of the WTGs and reduce the lost production caused by ice [19]. The research article question is the following: Is possible to detect in a early condition faults by yaw misalignment and ice influence in the wind turbine generator with data from production? The research article is organized as follows: In Section “Description and model developmentDescription and model development” description of the blade and yaw alignment is done with a brief systematic review and mathematical modeling of the system is presented; therefore, in Section “Case study” a case study in a real wind farm with thirteen WTGs, associated to active power vs wind turbine generators and wind direction vs yaw angle analysis. Later, the Section “Conclusions” is a discussion about the main results and contribution to the failure analysis theory; finally, the last section has conclusions.

Section snippets

Systematic review

This research article has been analyzed with the PICO methodology, the algorithm implemented for the reference evaluation has detected 79 research articles, in December 12th, 2020, as follows: ((Wind farm) OR (Wind turbine)) AND ((Yaw) OR(pitch)) AND(failure) NOT((storage) OR(structural) OR(atmospheric) OR(horizontal)). The main contribution is the following.

  • The component investigated in the last year is the bearing, however “the root causes of premature failure of bearings are still much less

Distribution of the wind turbine generator

The case the study is for thirteen WTGs, with a location in the north of Perú, in the Fig. 9. About the WTG a good indicator, as a first approach, it is the R2 between neighborhood with WTGs. In this case study, The worst R2 found between V11 and V12 (in the Fig. 10), it could be an failure in the anemometer and yaw systems, it is due to lowest R2 and correlation with wind velocity average each ten minutes in Fig. 12; usually the average is among 0.55 and 0.89. Furthermore, an interesting case

Conclusions

A novel evaluation of the failure in wind farms with lower efficiency than manufacturer design. The mathematical model for the analysis of the power and wind curve is considered with a complete evaluation in a wind farm of thirteen wind turbine generators. The main root cause failure is the blade problems, mainly associated to contamination and Yaw angle defects. The analysis allows to compare online the information in all the wind farm. It has been experimentally validated with a wind farm.

List of variables

  • V is the wind speed average.

  • X is the mechanical state of the turbine.

  • P(X=1) = p probability of availability of WTG.

  • P(wt) is the power produced in the WTG according the wind velocity, through the time and performance of the WTG.

  • F(v) is the wind velocity function.

  • v is the wind velocity in a specific time.

  • T is the time, with T > 0.

  • b turbine constant.

  • vci is a limit wind velocity, for an specific analysis.

  • vr is the rated wind speed.

  • βi=1 wind velocity range though a stage and position, with wTG

CRediT authorship contribution statement

Ricardo Manuel Arias Velásquez: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing - original draft, Visualization, Writing - review & editing, Supervision, Project administration. Freddy Antonio Ochoa Tataje: Validation, Formal analysis, Investigation, Data curation, Visualization, Writing - review & editing. María del Carmen Emilia Ancaya-Martínez: Validation, Formal analysis, Investigation, Data curation, Visualization, Writing - review &

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (21)

There are more references available in the full text version of this article.

Cited by (0)

View full text