Towards automated and integrated O&M data collection - standardising workflow processes for the offshore wind industry

Alexios Koltsidopoulos Papatzimos 1 ,2, Tariq Dawood2, Philipp Thies3
1Industrial Doctoral Centre for Offshore Renewable Energy (IDCORE), Edinburgh, UK, 2EDF Energy R&D UK Centre, London, UK, 3University of Exeter, Penryn, UK


A significant amount of operation and maintenance (O&M) data are being generated daily from offshore wind farms. This includes a variety of monitoring systems, maintenance reports and environmental sources. The challenge with having a wide diversity of data in inhomogeneous types and formats, is the considerable human effort involved in the initial extraction, transformation and loading (ETL) stages for these data to be processed and analysed. Although several commercial solutions are available, aiming to improve data management to support O&M decision making, the initial ETL phase is still a work-intensive process. One of the main reasons is that the organization and structure of the data flow does not allow easy access to the data. Due to the rapid growth of offshore wind turbine installations, there is a need to automate and integrate some of these processes in order to reduce the human effort and the associated costs. The aim is to facilitate a responsive, data driven decision making for O&M. This paper shows the results of re-structuring and automating the data flow from daily maintenance procedures to achieve a more efficient data analysis. These early results also indicate that less man-hours and a smaller number of people need to work on data collection. The framework and the steps followed will be of interest to offshore wind farm developers and operators to automate their data collection workflow.


A flexible and easily adjustable ETL framework has been created and is being tested, building on the existing expertise of data collection at an operational offshore wind farm. More precisely, the methodology that has been achieved with the collaboration of the operational and R&D teams, comprises the following three stages: i) the combination of current procedures via joint worksheets in extractable formats, ii) the automation of data collection, categorization, integration, filtering and pre-processing, by high-level programming and text mining and iii) the automated loading into a database for storing and analysis. Moreover, to further test and evaluate the ETL and integration processes, an integrated database including the generated monitoring data has been created.


The results of this automated and integrated data collection show a significant reduction in the man-hours and the number of people needed for data collection and processing. The previous multistage data pre-processing procedure has been simplified and automated, allowing easier and faster data validation, integration of multiple data sources, rapid visualization and automated failure root cause analysis. This has also allowed the creation of three layers of information from a single platform, for the R&D, O&M and asset management teams. Additionally, the integrated database created, allows targeted data analysis to support data driven O&M strategies and operational and asset management decisions.


Although a lot of tools exist, trying to improve data management and failure identification and prediction, there is still a large amount of human effort required in the initial stages of the collection, organization and filtering of input data. The proposed method for automating and integrating the data collection workflow has been established and is being tested at an operational offshore wind farm with the results showing significant reductions in human involvement. This approach will help to efficiently data-manage an increasing amount of O&M data in the offshore wind sectors.


This paper and presentation will inform the offshore wind farm developers and operators on how to better plan and structure their data collection and management framework and how they can automate and integrate processes in order to reduce costs, reduce man-hours and facilitate quicker and data-driven decision making. The standardization of the ETL processes will be one of the key industry drivers to achieve further O&M cost reduction through possible data sharing under the same quality standards and conditions.