Proceedings of the IEEE Prognostics and Health Management, Austin, TX, June 2015

PHM Based Predictive Maintenance Optimization for Offshore Wind Farms

Xin Lei1, Peter Sandborn1, Roozbeh Bakhshi1, Amir Kashani-Pour and Navid Goudarzi1

1CALCE, Department of Mechanical Engineering, College Park, MD, 20742


In this paper, a simulation-based real options analysis (ROA) approach is applied to evaluate the predictive maintenance options created by PHM for multiple turbines in offshore wind farms managed under outcome-based contracts known as power purchase agreements (PPAs). When a remaining useful life (RUL) is predicted for a subsystem in a single turbine, a predictive maintenance option is triggered. If predictive maintenance is implemented before the subsystem or turbine fails, the option is exercised; if the predictive maintenance is not implemented and the subsystem or turbine runs to failure, the option expires and the option value is zero. The time-history cost avoidance and cumulative revenue paths are simulated considering the uncertainties in wind and the RUL predictions. By evaluating a series of European real options based on all possible predictive maintenance opportunities, the maintenance opportunity with the maximum value can be obtained. In a wind farm, there may be multiple turbines concurrently indicating RULs. To model multiple turbines managed via an outcome based contract (PPA), the cumulative revenue and cost avoidance for each turbine depends on the operational state of the other turbines in the farm, the amount of energy that has been delivered and will be delivered by the whole farm. A case study is presented that determines the optimum predictive maintenance opportunity for a farm under a PPA, the optimum predictive maintenance opportunity for the same farm managed via an asdelivered contract, and the optimum predictive maintenance opportunities for individual turbines managed independently.

Complete article is available from the publisher and to the CALCE consortium members.

[Home Page] [Articles Page]
Copyright © 2011 by CALCE and the University of Maryland, All Rights Reserved