PosterPDF Available

SCADA alarms processing for wind turbine component failure detection

  • October 2016
  • Conference: The Science of Making Torque from Wind 2016
  • At: Munich, Germany
Authors:
Elena Gonzalez at Scottish Power Renewables
Elena Gonzalez
  • Scottish Power Renewables
Maik Reder at ANNEA.ai GmbH
Maik Reder
  • ANNEA.ai GmbH
Julio J. Melero at University of Zaragoza
Julio J. Melero
Download file PDF
Read file
Download citation

Abstract

Early detection of wind turbine failures facilitates the changeover from corrective maintenance towards a predictive approach, thus reducing the O&M costs. This study presents a cost-effective methodology for component failure detection purposes through the analysis of SCADA alarms. The approach categorises the alarms according to a reviewed taxonomy and then detect alarm patterns before actual failures by applying different techniques. Two case studies highlight therelationship between faulty behaviour in different components and between failures and adverse environmental conditions.
Content uploaded by Elena Gonzalez
Author content
All content in this area was uploaded by Elena Gonzalez on Oct 13, 2016
Content may be subject to copyright.
TIME-SEQUENCE ANALYSIS BEFORE AN ACTUAL FAILURE
Early detection of wind turbine failures facilitates the changeover from corrective maintenance towards apredictive approach, thus reducing the O&M costs.
This study presents acost-effective methodology for component failure detection purposes through the analysis of SCADA alarms. The approach categorises the alarms
according to areviewed taxonomy [1] and then detect alarm patterns before actual failures by applying different techniques [2].
Two case studies highlight the relationship between faulty behaviour in different components and between failures and adverse environmental conditions.
The methodology translates overwhelming data into valuable information. The
automatic use of SCADA data avoids the use of additional equipment and
complex signal analysis processes.
The case studies highlight the relationship between failures and faulty
behaviour in different components and/or adverse environmental conditions.
Certain alarms could then be directly related to upcoming component failures.
Further research work will aim at overcoming the drawbacks of the techniques
by applying more complex machine learning techniques.Moreover,
component-related alarms are being correlated to deviations from normal WT
operating conditions, in terms of performance.
This project has received funding from the European Union's Horizon 2020
research and innovation programme under the Marie Skłodowska-Curie grant
agreement No 642108.(AWESOME project -http://awesome-h2020.eu/).
HOW DOES THE SCADA ALARM SYSTEM WORK?
1. Wilkinson M, Hendriks B, Spinato F, Gomez E, Bulacio H, Tavner P, Feng Y and Long H.
EWEC 2010
2. Qiu Y, Feng Y, Tavner P J, Richardson P, Erdos G and Chen B (2012) Wind Energy 15
3. Reder M, Gonzalez E and Melero J J(2016) Journal of Physics: Conference Series
HISTORICAL DATA AVAILABLE (3 op. years)
WT make Technology Rated Power (kW) Nb of WTs Alarm system
A Geared Generator 1500 55 1
BDirect Drive 2000 37 2
CDirect Drive 2000 19 2
D Geared Generator 850 77 3
E Geared Generator 2000 133 4
F Geared Generator 1800 9 5
G Geared Generator 2000 76 5
Reviewed Taxonomy
system, sub-system, assembly
1-month prior to the
failure (affected WT)
Historical
SCADA alarms
Historical
Failure Logs
Component-related
failure data
Component-related
SCADA alarms
CATEGORISATION
PROBABILITY-BASED ANALYSIS BEFORE AN ACTUAL FAILURE
 ↔   →   ∩  =∅
A = B
  = 1
A
  > 0
BA
  = 0
B
A
B
,=,,,
A
B
,,,
,
A
B
,,,
,
A B A B A B
 ↔   →   ∩  =∅
INFORMATION
Grid conditions
Environmental conditions
WT operational state
Maintenance activities
Component malfunction
Manual stop or restriction
GEARBOX FAILURE
Failure following many alarms involving the hydraulic system (a2, a3, a4, a5)
and the cooling system (a6, a7, a8). A causal relationship between these
assemblies and the gearbox could be then assumed.
Alarms of a certain component resulted in afailure of another component.
YAW SYSTEM FAILURE
The high wind speed alarm (a26)was activated several times, regardless the
time-sequence patterns with the yaw-related alarms.The significant number of
times a certain alarm is triggered could be considered as an indicator of afailure
likely to occur in the future.
Probabilistic approach reveals the link between the failure and arecurrent
alarm, that could be used as an indicator.
P(a23) = P(a26) = 2.8%
P(a23|a26) = 100%
P(a26→a23) = 2.8%
P(a26) = 36.6%
Critical components
(see [3]) 1 –Blades
2 –Gearbox
3 –Generator
4 –Pitch system
5 –Yaw system
6 –Transformer
7 –Frequency converter
The Science of Making Torque from Wind
(TORQUE 2016)
DOWNLOAD THE FULL VERSION OF THE PAPER
SCADA alarms processing for wind turbine
component failure detection
Elena Gonzalez, Maik Reder and Julio J Melero
Fundación CIRCE
ABSTRACT
DATA DESCRIPTION
METHODOLOGY
RESULTS AND DISCUSSION
CONCLUSIONS AND OUTLOOK REFERENCES
ACKNOWLEDGEMENTS
CONTACT INFORMATION
Elena Gonzalez
Fundación CIRCE
C/Mariano Esquillor Gomez, 15 50018 Zaragoza (Spain)
egonzalez@fcirce.es
Network diagram of the SCADA
alarm patterns before the failure
Venn diagram
Network diagram of the SCADA alarm patterns before the failure
Exploring the Limits of Early Predictive Maintenance in Wind Turbines Applying an Anomaly Detection Technique
Article
Full-text available
The aim of the presented investigation is to explore the time gap between an anomaly appearance in continuously measured parameters of the device and a failure, related to the end of the remaining resource of the device-critical component. In this investigation, we propose a recurrent neural network to model the time series of the parameters of the healthy device to detect anomalies by comparing the predicted values with the ones actually measured. An experimental investigation was performed on SCADA estimates received from different wind turbines with failures. A recurrent neural network was used to predict the temperature of the gearbox. The comparison of the predicted temperature values and the actual measured ones showed that anomalies in the gearbox temperature could be detected up to 37 days before the failure of the device-critical component. The performed investigation compared different models that can be used for temperature time-series modeling and the influence of selected input features on the performance of temperature anomaly detection.
View
Wind Turbine Reliability - Maintenance Strategies
Chapter
  • Jan 2021
Wind turbine technologies are evolving rapidly in terms of complexity and size and there is an urgent need for cost-effective Operations and Maintenance (O&M) strategies to increase the profitability of wind power assets. Component reliability as well as maintenance strategies have the greatest effect on O&M planning and costs, especially in offshore applications. This Chapter provides an overview of the state of the art in the field of wind turbine and wind farm reliability and maintenance, discussing the main points, lessons and opportunities learnt from both industry and academic experience.
View
Reliability, availability, maintainability data review for the identification of trends in offshore wind energy applications
Article
Full-text available
This work presents a comprehensive review and discussion of the identification of critical components of the currently installed and next generation of offshore wind turbines. A systematic review on the reliability, availability, and maintainability data of both offshore and onshore wind turbines is initially performed, collecting the results from 24 initiatives, at system and subsystem level. Due to the scarcity of data from the offshore wind industry, the analysis is complemented with the extensive experience from onshore structures. Trends based on the deployment parameters for the influence of design characteristics and environmental conditions on the onshore wind turbines’ reliability and availability are first investigated. The estimation of the operational availability for a set of offshore wind farm scenarios allowed a comparison with the recently published performance statistics and the discussion of the integrity of the data available to date. The failure statistics of the systems deployed offshore are then discussed and compared to the onshore ones, with regard to their normalised results. The availability calculations supported the hypothesis of the negative impact of the offshore environmental conditions on the reliability figures. Nonetheless, similarities in the reliability figures of the blade adjustment system and the maintainability of the power generation and the control systems are outlined. Finally, to improve the performance prediction of future offshore projects, recommendations on the effort worth putting into research and data collection are provided.
View
Stochastic assessment of aerodynamics within offshore wind farms based on machine-learning
Article
Wind turbine flow field prediction is difficult as it requires computationally expensive computational fluid dynamics (CFD) models. The contribution of this paper is to propose and develop a method for stochastic analysis of an offshore wind farm using CFD and a non-intrusive stochastic expansion. The approach is developed through testing a range of machine-learning methods, evaluating dataset requirements and comparing the accuracy against site measurement data. The approach used is detailed and the results are compared with real measurements obtained from the existing wind farm to quantify the accuracy of the predictions. An existing offshore wind farm is modelled using a steady-state CFD solver at several deterministic input ranges and an approximation model is trained on the CFD results. The approximation models compared are Artificial Neural Networks, Gaussian Process, Radial Basis Function, Random Forest and Support Vector Regression. RBF achieves a mean absolute error relative to the CFD model of only 0.54% and the error of the SVR predictions relative to the real data, with scatter, was 12%, compared to 16% from Jensen. This approach has the potential to be used in more complex situations where an existing analytical method is either insufficient or unable to make a good prediction.
View
Offshore wind turbine reliability and operational simulation under uncertainties
Article
Full-text available
The fast‐growing offshore wind energy sector brings opportunities to provide a sustainable energy resource but also challenges in offshore wind turbine (OWT) operation and maintenance management. Existing operational simulation models assume deterministic input reliability and failure cost data, whereas OWT reliability and failure costs vary depending on several factors, and it is often not possible to specify them with certainty. This paper focuses on modelling reliability and failure cost uncertainties and their impacts on OWT operational and economic performance. First, we present a probabilistic method for modelling reliability data uncertainty with a quantitative parameter estimation from available reliability data resources. Then, failure cost uncertainty is modelled using fuzzy logic that relates a component's failure cost to its capital cost and downtime. A time‐sequential Monte Carlo simulation is presented to simulate operational sequences of OWT components. This operation profile is later fed into a fuzzy cost assessment and coupled with a wind power curve model to evaluate OWT availability, energy production, operational expenditures and levelised cost of energy. A case study with different sets of reliability data is presented, and the results show that impacts of uncertainty on OWT performance are magnified in databases with low components' reliability. In addition, both reliability and cost uncertainties can contribute to more than 10% of the cost of energy variation. This research can provide practitioners with methods to handle data uncertainties in reliability and operational simulation of OWTs and help them to quantify the variability and dependence of wind power performance on data uncertainties.
View
Wind turbine reliability data review and impacts on levelised cost of energy
Article
Full-text available
Reliability is critical to the design, operation, maintenance, and performance assessment and improvement of wind turbines (WTs). This paper systematically reviews publicly available reliability data for both onshore and offshore WTs and investigates the impacts of reliability on the cost of energy. WT failure rates and downtimes, broken down by subassembly, are collated from 18 publicly available databases including over 18 000 WTs, corresponding to over 90 000 turbine‐years. The data are classified based on the types of data collected (failure rate and stop rate) and by onshore and offshore populations. A comprehensive analysis is performed to investigate WT subassembly reliability data variations, identify critical subassemblies, compare onshore and offshore WT reliability, and understand possible sources of uncertainty. Large variations in both failure rates and downtimes are observed, and the skew in failure rate distribution implies that large databases with low failure rates, despite their diverse populations, are less uncertain than more targeted surveys, which are easily skewed by WT type failures. A model is presented to evaluate the levelised cost of energy as a function of WT failure rates and downtimes. A numerical study proves a strong and nonlinear relationship between WT reliability and operation and maintenance expenditure as well as annual energy production. Together with the cost analysis model, the findings can help WT operators identify the optimal degree of reliability improvement to minimise the levelised cost of energy.
View
Offshore wind turbine fault alarm prediction
Article
Full-text available
Offshore wind operations and maintenance (O&M) costs could reach up to one third of the overall project costs. In order to accelerate the deployment of offshore wind farms, costs need to come down. A key contributor to the O&M costs is the component failures and the downtime caused by them. Thus, an understanding is needed on the root cause of these failures. Previous research has indicated the relationship between wind turbine failures and environmental conditions. These studies are using work‐order data from onshore and offshore assets. A limitation of using work orders is that the time of the failure is not known and consequently, the exact environmental conditions cannot be identified. However, if turbine alarms are used to make this correlation, more accurate results can be derived. This paper quantifies this relationship and proposes a novel tool for predicting wind turbine fault alarms for a range of subassemblies, using wind speed statistics. A large variation of the failures between the different subassemblies against the wind speed are shown. The tool uses 5 years of operational data from an offshore wind farm to create a data‐driven predictive model. It is tested under low and high wind conditions, showing very promising results of more than 86% accuracy on seven different scenarios. This study is of interest to wind farm operators seeking to utilize the operational data of their assets to predict future faults, which will allow them to better plan their maintenance activities and have a more efficient spare part management system.
View
Wind turbine reliability: A comprehensive review towards effective condition monitoring development
Article
The current work presents a review of wind turbine reliability studies. Thirteen reliability studies were identified in the scientific literature, highlighting the lack of public reliability data. We present the differences across the studies, with great effort being made to unify the various studies to obtain comparable results. To this end, we have endeavoured to develop a wind turbine taxonomy that is common across the different studies, formed by thirteen assemblies, and the failure rates and downtimes from each study have been normalised. The results establish differences between the least reliable assemblies, categorised as critical, and the most reliable ones. Small differences emerge between onshore and offshore locations, and between studies on European wind farms and others in the U.S. and China. The influence on the total failures and downtime of the most recent studies is evaluated against older studies. These results will contribute to elucidate the right direction for condition monitoring design and development, and therefore to improve reliability and availability of wind turbines.
View
Key Performance Indicators for Wind Farm Operation and Maintenance
Article
Full-text available
  • Oct 2017
Key performance indicators (KPI) are tools for measuring the progress of a business towards its goals. Although wind energy is now a mature technology, there is a lack of well-defined best practices to assess the performance of a wind farm (WF) during the operation and maintenance (O&M) phase; processes and tools of asset management, such as KPIs, are not yet well-established. This paper presents a review of the major existing indicators used in the O&M of wind farms (WFs), as such information is not available in the literature so far. The different stakeholders involved in the O&M phase are identified and analysed together with their interests, grouped into five categories. A suggestion is made for the properties that KPIs should exhibit. For each category, major indicators that are currently in use are reviewed, discussed and verified against the properties defined. Finally, we propose a list of suitable KPIs that will allow stakeholders to have a better knowledge of an operating asset and make informed decisions. It is concluded that more detailed studies of specific KPIs and the issues of their implementation are probably needed.
View
ResearchGate has not been able to resolve any references for this publication.