This paper presents an engineering asset management model for wind turbines that integrates all dimensions impacting the life-cycle of a wind farm. Both technical modules (reliability, supply-chain…) and economic modules (present values of cash-flows, valorization of generation losses) are created and linked together into a global one. The weather module, based on artificial neural networks, used to quantify the accessibility of the turbines for installation or maintenance tasks is described in depth. The paper concludes with test cases demonstrating the importance of integrating such a realistic weather model with the assets models, as well as showing how the global model can be used to support operational decision making. This study is of interest to wind farm operators and service providers seeking to optimize their installation and operational strategies and reduce the overall offshore wind farm costs.
Offshore wind power is a recent electricity generation technology when compared to other renewable energies and if the installed power, around 19GW in 2017, is still largely beneath onshore wind production (about 540GW), it has been growing exponentially since the years 1990 (Sawyer, Liming & Fried 2018). Offshore wind speeds, higher than inland ones, make the offshore energy sector very promising to support the energy transition toward a decarbonized electricity production, although recent strike prices around Europe have put very challenging targets to the offshore wind industry. In order to make sure that those prices are met, the industry needs to reduce its installation and O&M costs, which account for 18 to 23% of the total cost of an offshore wind farms according to (Tavner, 2012). Thus engineering and asset management are useful to optimize the life-cycle management of offshore wind assets (Anastasia, Andrew and Feargal, 2018).
Engineering asset management models have been developed in the industry for decades and have been applied successfully to the energy production for different types of generation (nuclear, fossil, hydropower.). If such models exist for offshore wind turbines and proved to be efficient tools to optimize offshore assets management (Scheu, Matha, Hofmann, Muskulus, 2012 or Douard, Domecq & Lair, 2012), they usually rely on specific assumptions that may differ from one model to the other, making validation and verification difficult as shown in (Dinwoodie et al., 2015). In this paper we will describe a new model that integrates all the dimensions of wind turbine assets, both technical and economic. The main novelty of the presented tool is its flexibility to model any offshore wind farm operations, being agnostic on whether these are installation or maintenance ones. We will focus on the weather model that uses artificial neural networks, to generate weather time series statistically consistent with historical data. Several examples will then highlight the importance of using such models instead of relying on the historical weather data. We will also show how the described integrated model can be used to support life-cycle decision making.
