Failure Prognostic Schemes and Database Design of a Software Tool for Efficient Management of Wind Turbine Maintenance

Abstract

This is an author produced version of a paper published in Wind Engineering (ISSN 0309-524X, eISSN 2048-402X) This version may not include final proof corrections and does not include published layout or pagination. Citation DetailsCitation CopyrightItems in 'OpenAIR@RGU', Robert Gordon University Open Access Institutional Repository, are protected by copyright and intellectual property law. If you believe that any material held in 'OpenAIR@RGU' infringes copyright, please contact openair-help@rgu.ac.uk with details. The item will be removed from the repository while the claim is investigated. This is the author's reviewed version, the full published article may be found at http://dx.doi.org/10.1260/0309-524X.39. AbstractWind Turbines require numerous and varied types of maintenance activities throughout their lifespan, the frequency of which increases with years in operation. At present the proportion of maintenance cost to the total cost for wind turbines is significant particularly for offshore wind turbines (OWT) where this ratio is ~35%. If this ratio is to be reduced in-spite of adverse operating conditions, pre-mature component failures and absence of reliability database for wind turbine components, there is a need to design unconventional maintenance scheme preferably by including novel failure prediction methodologies. Several researchers have advocated the use of Artificial Neural Networks (ANN), Bayesian Network Theory (BNT) and other statistical methods to predict failure so as to plan efficient maintenance of wind turbines, however novelty and randomness of failures, nature and number of parameters involved in statistical calculations and absence of required amount of fundamental work required for such advanced analysis have continued to maintain the high cost of maintenance. This work builds upon the benefits of condition monitoring to design methods to predict generic failures in wind turbine components and exhibits how such prediction methods can assist in cutting the maintenance cost of wind turbines. This study proposes using a dedicated tool to assist with failure prediction and planning and execution of wind turbine maintenance. The design and development of such an all-inclusive tool will assist in performing administrative works, inventory control, financial calculations and service management apart from failure prediction in wind turbine components. Its database will contain reference to standard management practices, regulatory provisions, staff details and their skillsets, service call register, troubleshooting manuals, installation guide, service history, details of customers and clients etc. that would cater to multiple avenues of wind turbine maintenance. In order to build such a software package, a robust design of its database is crucial. This work lists prerequisites for choosing a physical database and identifies the benefits of relational database software in controlling large amounts of data of various formats that are stored in such physical databases. Such a database wou...

A Comprehensive Review of Artificial Intelligence and Wind Energy

“…The most common source of information is supervisory control and data acquisition (SCADA) data [19][20][21] or vibration signals [22][23][24]. Combinations with other techniques are common to improve performance, such as with Decision Trees [25,26] [113,114] and R 2 adjustment [115,116] are also applied, together with simulations [117][118][119], prototypes [120,121] and online classification systems [122,123]. It is not uncommon to compare different techniques to decide the best choice [124,125].…”

Bayesian Network Modelling for the Wind Energy Industry: An Overview

“…The fatigue life after each inspection was estimated using a Bayesian updating method. In another study, Sinha and Steel (2015) proposed a BN model to incorporate qualitative information into the estimation of failure probability. They considered four major factors, namely, system faults, operational factors, human factors and external factors (environment) in the analysis.…”

Research on Operation Safety of Offshore Wind Farms

“…Representative qualitative analysis methods are Fault Mode Analysis Method [14], Tree Chart Analysis [15], SWOT (Strengths Weaknesses Opportunities Threats) [16]. Representative quantitative methods are Bayesian method [17][18][19], Monte Carlo Analysis [20], and reliability-based design optimization tools [21,22]. However, the operation of the offshore wind farm is systematic.…”