Analyzing a Decade of Wind Turbine Accident News with Topic Modeling

Abstract

Despite the significance and growth of wind energy as a major source of renewable energy, research on the risks of wind turbines in the form of accidents and failures has attracted limited attention. Research that applies data analytics methodologically in this context is scarce. The research presented here, upon construction of a text corpus of 721 selected wind turbine accident and failure news reports, develops and applies a custom-developed data analytics framework that integrates tabular analysis, visualization, text mining, and machine learning. Topic modeling was applied for the first time to identify and classify recurring themes in wind turbine accident news, and association mining was applied to identify contextual terms associated with death and injury. The tabular and visual analyses relate accidents to location (offshore vs. onshore), wind turbine life cycle phases (transportation, construction, operation, and maintenance), and the incidence of death and injury. As one of the insights, more incidents were found to occur during operation and transportation. Through topic modeling, topics associated most with deaths and injuries were revealed. The results could benefit wind turbine manufacturers, service providers, energy companies, insurance companies, government bodies, non-profit organizations, researchers, and other stakeholders in the wind energy sector.

Topic Modelling in Computer Security Discourse: a Case Study of Whitepaper Publications and News Feeds

“…Data analysis has been implemented within the machine learning principles. G. Ertek and L. Kailas define machine learning as computer algorithms relying on training and capable of identifying patterns, determining insights, and predicting [Ertek, Kailas: 2021]. Unsupervised machine learning "aims to discover patterns or structures in a dataset without considering any target attribute" [Ertek, Kailas 2021: 8-9].…”

Innovative Energy-Efficient Prefabricated Movable Buildings for Smart/Co-Working: Performance Assessment upon Varying Building Configurations

“…• the electric energy purchased from the national electric grid is zero; • the values of PES, ∆CO 2 , and ∆OC are equal to 100%, meaning that the corresponding building configuration is energetically self-sufficient (thanks to the electric energy produced by both the wind turbine and the photovoltaic panels coupled with three electrical storages) and can be used for stand-alone applications. (15) and Equation (18), respectively) as a function of the case studies (cases 1, 2, and 3 belong to building configuration A, cases 4, 5, and 6 belong to building configuration B, and cases 7, 8, 9, and 10 belong to building configuration C). This figure underlines that:…”

“…Among the renewable-based energy systems, the global renewable power sector is predicted to grow by 50% between 2019 and 2024, with solar photovoltaic (PV) panels accounting for about 60% of the expected expansion thanks to their capability to harness electrical energy from solar energy [13]. The use of wind power is also rapidly growing in popularity in the last few years, with wind turbines allowing to generate electricity from wind with a power output ranging from a few hundred watts to hundreds of kilowatts and even megawatts [14,15]. In terms of power generation, it should also be highlighted that coupling electric energy batteries with photovoltaic panels or wind turbines could assist in improving the electric energy generated and self-consumed, thus (i) lowering the costs associated with the electricity for end-users, (ii) reducing the overload on electric infrastructure in the case of peaks, and (iii) enhancing the reliability and quality of the electric grid [16].…”

Digitalization Workflow for Automated Structuring and Standardization of Maintenance Information of Wind Turbines into Domain Standard as a Basis for Reliability KPI Calculation

“…Other methods combine WT sensor data and maintenance reports for the purpose of predictive maintenance [11]. Further approaches focus on textual information of accidents at WTs to generate knowledge [13,14]. However, none of the before mentioned approaches consider domain standards like RDS-PP ® [2] or RDS-PS [3] for describing the affected components in a uniform manner.…”