PO090

Unlocking the potential of big data in the design of offshore wind structures

Jorge Parra, Irina Cortizo
Atkins Limited, London, UK

Abstract

The offshore wind energy industry has seen significant growth allowing the industry to move towards serialization and lower LCOE.


With the industry moving at high speed, the amounts of information that becomes available daily is overwhelming. Harnessing the right information in a timely and efficient manner is key to maximize its impact and use. Unlocking the potential of Big Data in the relative young offshore wind industry will prove a game changer, key for sustainable and efficient business growth allowing for design standardization and parametric design to become a closer reality.


This paper will present a case study where data trending and analytics have been used for the concept design of offshore substation platforms (OSPs). The data trends derived allow to clearly visualize the industry path to date in terms of OSP design as well as the expected trends for future developments. A set of outputs primarily in the form of structural weights are then obtained. A cost model considering both CAPEX and OPEX is built around this data and their considered accuracy allowing multiple potential concepts to be compared quickly and efficiently by several stakeholders.
Being able to find the most appropriate design solution applicable for each wind farm whilst factoring the lessons learned on the way, will allow for design standardization and extrapolation in an efficient manner. This in turn will be key for driving the LCOE down ensuring the industry can thrive in a non-subsidized environment.

 

Method

For the case study presented, OSP related data is gathered from Atkins internal sources as well as from publicly available external sources. The data is then systematically processed into a searchable database. A factoring approach, similar to that used for weight control during design projects, is used to reflect the various degrees of accuracy associated with the data.

Parameters such as wind farm capacity, water depth, substructure type or distance to shore are considered in the model. These parameters are then analyzed to derive trends using regression techniques. The learning from previous trends observed in similar O&G structures has been also incorporated into the model. Lastly, a feedback loop is introduced whereby these trends are updated as more information becomes available.

Results

A database of OSP data has been created containing details of more than 60 different assets. The data has been analyzed and the following trends have been derived:

 

 

The correlation between these trends and the scatter of data is studied to inform the potential data users of the opportunities and limitations of the information provided. The trends will be presented in a graphical visual form.

 

Conclusions

This paper will demonstrate the value in using a trending approach to determining crucial information associated with OSPs. Weight estimates can be efficiently derived with a reasonable accuracy for early planning and concept design phases, which will in turn allow the industry to reduce the time and investment required in the early phases of any development.

The methodology presented can be applied beyond OSP design scenarios and therefore, it becomes apparent that the use of trends can allow the industry to map out different offshore wind farm configurations to better inform initial key strategic decisions.

This paper will highlight the industry benefit in harnessing big data and maintaining a continuous feedback loop to improve and guarantee the accuracy of trend predictions like the ones presented.

 

Objectives

Delegates will learn through the presentation of this paper what key parameters are expected to have a significant impact on the derivation of weight estimates for OSPs. They will also gain an understanding of the spread of the data associated with the offshore substations assets in European waters.

A visualization of the "direction of travel" of the industry is shown with regards to estimated structural weights for OSPs. This will broaden the understanding of developers, installers and manufacturers of the path that the industry is likely to take in the near and medium future.

Additionally, delegates will gain an appreciation for the significant impact that collaboration, knowledge and data sharing can have in unlocking the potential of big data within the offshore wind energy industry.