Analysis of Offshore Wind Turbine Operation & Maintenance Using a Novel Time Domain Meteo-ocean Modeling Approach

This paper presents a novel approach to repair modeling using a time domain Auto-Regressive model to represent meteoocean site conditions. The short term hourly correlations, medium term access windows of periods up to days and the annual distribution of site data are captured. In addition, seasonality is included. Correlation observed between wind and wave site can be incorporated if simultaneous data exists. Using this approach a time series for both significant wave height and mean wind speed is described. This allows MTTR to be implemented within the reliability simulation as a variable process, dependent on significant wave height. This approach automatically captures site characteristics including seasonality and allows for complex analysis using time dependent constraints such as working patterns to be implemented. A simple cost model for lost revenue determined by the concurrent simulated wind speed is also presented. A preliminary investigation of the influence of component reliability and access thresholds at various existing sites on availability is presented demonstrating the ability of the modeling approach to offer new insights into offshore wind turbine operation and maintenance. INTRODUCTION Since the mid 1990s, there has been an exponential growth in the world wide installed capacity of wind turbines from around 6GW, concentrated in Northern Europe and the USA in 1996 to almost 200 GW spread across the world [1]. Onshore wind power is now considered the most mature renewable technology and operators have obtained significant experience in operation and maintenance (O&M) of wind farms. The most common approach for large onshore wind farms is a combination of scheduled maintenance, typically one to two visits per year and reactive maintenance, restoring components after failure. This approach has been deemed to be cost effective for operators and has allowed onshore availabilities of over 97% to be achieved [2]. In the last decade offshore wind energy has experienced exponential growth to a worldwide installed capacity of over 3GW focused in Northern European waters [3]. This expansion has coincided with the arrival of larger, multi MW machines suited to sites with higher mean wind speeds and has been driven by the decrease in available onshore sites and planning issues. This is particularly true in the UK where applications for large onshore wind farms have met with increasing planning difficulty and public resistance due to their visual impact. The shift towards offshore development has resulted in greater capacity currently being under development in the UK than onshore [4]. In addition, offshore projects currently at the scoping or development stage in Europe total exceed 100GW in capacity, [5]. It will only require a small proportion of these projects to be developed to create a significant market. The large capital expenditure for an offshore wind farm has resulted in a significantly different market structure from onshore wind. The market currently only exists for large scale developers and is dominated by a few OEMs and this trend is expected to continue. Offshore, there has been a lack of diverse operator experience and a conspicuous lack of failure databases such as those available for onshore [6-8]. In addition, many of the larger offshore wind farms are still operated under warranty. The result is that significant uncertainty exists surrounding offshore failure characteristics and early offshore wind farms have tended to adopt conventional operational strategies. This has resulted in poor availabilities of around 80% and a wide variation between operating years and different sites [9, 10]. Similar uncertainty exists around the costs of O&M with estimates ranging from 20 – 33% of overall project cost [11, 12]. Even at lower estimates this represents a huge financial