**1. Introduction**

The continued growth in the size and complexity of offshore wind turbines means that more profitable O&M actions will be needed to optimize the upper ranges of robustness for RAMS, in order to fulfill the size increase [1].

Previous research has indicated that O&M constitutes up to 20–30% of the overall cost of OWTs during their lifetimes. However, lowering the O&M cost per unit power will rely on larger OWTs, due to the greater cost per failure of smaller OWTs, their high demand for palliative actions (e.g., corrective maintenance), and their loss of production during

**Citation:** V. Taboada, J.; Diaz-Casas, V.; Yu, X. Reliability and Maintenance Management Analysis on OffShore Wind Turbines (OWTs). *Energies* **2021**, *14*, 7662. https://doi.org/10.3390/ en14227662

Academic Editors: Davide Astolfi, Eugen Rusu, Kostas Belibassakis and George Lavidas

Received: 29 July 2021 Accepted: 11 November 2021 Published: 16 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

downtimes [1]. Therefore, increasing turbine size implies decreasing O&M costs. Larger OWTs provide a lower number of individual machines that need to be conserved and could therefore provide lower O&M costs [2]. The design and modeling of O&M costs is essential to the screening of cost-effective maintenance strategies and decision-making, as well as the development of specific methodologies for O&M. In addition, design and modeling increase trust for wind energy investors financing OWTs. Therefore, this analysis is a significant step for the growth of wind power [3]. The O&M costs quantified and measured in this paper are the cost for personnel, spare parts, and vessels required for the accomplishment of maintenance requirements of the wind farms. Normally, maintenance is understood as a general concept that includes all interventions (inspections, repairs, replacement of components/elements, etc.). The analysis of current and previous O&M strategies for OWTs takes into account industrial achievements made in the oil and gas industry and the manufacturing industry in order to identify the most important functional drivers for O&M planning, and management for OWTs. Thus, previous trials and achievements in other industries act as an input driver for O&M in the offshore wind industry.

To gain insight into current advances in O&M knowledgebase standardization, offshore wind farm models are based on today's state-of-the-art OWTs, approximately 25 years after the first generation of conventional OWTs was designed, manufactured, and installed.

On the other hand, the use of larger wind turbines generates much greater uncertainty. Operation and maintenance costs represent a large part of the total life cycle cost (LCC), with operation and maintenance costs being approximately 22 to 40% of the overall total cost of an offshore wind farm [4,5]. Those costs are related to the risk cost incurred by the profit lost due to downtimes of OWTs.

O&M activities account for around <sup>1</sup> <sup>4</sup> of the life-time costs of a regular offshore wind farm. Over the next twenty years, offshore wind O&M will turn into a significant industrial sector in its own right [3]. For instance, in the UK government's forecasts for the deployment of offshore wind, O&M activities for more than 5500 OWT's could be worth almost £2bn/year by 2025. The graphs are shown below in Figure 1.

**Figure 1.** UK O&M spending by strategy class during years 2013–2025, data from [6].

Figure 2 represents a simple understanding of O&M research results for common offshore wind projects at different distances from the nearby O&M harbor. From the analysis, the junction points are at around 12 nautical miles (NM) (to have helicopter support) and at 40 NM (to trigger offshore wind-based strategies). However, it is vital to remember that there also many site-specific external aspects (environmental conditions, aviation regulations, safety considerations, and suitability) of existing ports that affect decisions about the exact positions of these junction points [3].

**Figure 2.** O&M cost as a function of distance from the O&M port, data from [5].

On the other hand, the prominence and challenges of O&M for OWTs are recognized in both academia and industry. The availability of OWTs is much less favorable and their costs can be more than 1.5 times higher than onshore wind. Furthermore, onshore wind turbines are capable of achieving 95–99% availability and producing electricity at a reasonable price in the market. There is clear cost reduction potential for O&M, which contributes around 30% of the total cost of offshore wind.

The emphasis of this document is to research and develop methods to improve and optimize the efficiency of operation and maintenance in offshore wind farms. Efficiency is related to the optimization of maintenance organization in offshore wind farms. The decrease of O&M costs is directly addressed in this document and the research results are supportive.

The research presented throughout this document analyzes the existing approaches and methods used for access, design, operation, maintenance planning, and life cycle engineering in offshore wind farms.

#### *1.1. Challenges and Solutions for OWT Maintenance Activities*

#### 1.1.1. Weather Conditions

The meteorological window is represented in the model by a time series accounting for significant wave height and wind speed when determining the hourly time. The weather forecast notes when a given set of offshore or marine activities (operations, construction, etc.) can be carried out within their maximum limits for wave height, wind speeds, etc. Specifically, marine operations are planned based on a reference period; the operation reference period is (TR) = planned operation period (TPOP) + estimated maximum contingency time (TC) [6]. Incorporating wave height and wind speed into a weather window is crucial to ensuring the accessibility of offshore wind farms. For operations to be considered not limited by meteorological factors, it is necessary that the planned operating time (TPOP) be less than 72 h and the reference period (TR) be less than 96 h.

The meteorological time series are created using a Markov chain model based on historical meteorological record input data from the specific site of an offshore wind farm. The Markov chain model reproduces and recreates random time based on models and estimated stochastic probabilities [4].

Failure occurrence can fit an exponential probability distribution dependent on failure rates. Given the failure rates *λ* (e.g., *λyear* <sup>04</sup> = 1, *λyear* <sup>510</sup> = 0.75, etc.) for a component/element in an OWT, the distribution probability function for the time duration Δ*t* until a failure happens on that explicit component/element, is set as:

$$
v \left(\Delta t\right) = \lambda e^{\lambda \Delta t} \tag{1}$$

where (Δ*t*) is the time interval until the next fault. Two cases are defined:


The maintenance model, therefore, is able to repeat simulations. Each one takes weather scenarios as diverse and random, and uses arbitrary times for failures to account for doubt in the times for failure rates and weather effects [7].

#### 1.1.2. Weather Delays and Repair Timing

The total downtime per failure is the sum of the downtime originating because of:


Safety weather window and work shift constraints create expected maintenance delays, which are statistically determined for the given time duration (rm & rM) based on the environmental time series sum for the offshore location, with the vessels considered limited by wave height and wind speed [7].

Downtime repair comprising of waiting for weather (without the effect of queuing) is referred to as *d<sup>s</sup> <sup>m</sup>* (minor repairs) and *d<sup>s</sup> <sup>M</sup>* (major repairs). The average failure rate (*λS*) and repair time (*d<sup>s</sup> CM*) per failure and per season is calculated as:

$$
\lambda^S = \lambda\_m^S + \lambda\_M^S \tag{2}
$$

$$d\_{CM}^S = \frac{\lambda\_m^S d\_m^S + \lambda\_M^S d\_M^S}{\lambda^S} = \frac{1}{\mu^S} \tag{3}$$

where *μ<sup>S</sup>* is the resulting repair rate [8].

#### 1.1.3. Accessibility

As stated above, both wave height and wind speed are essential to guaranteeing the safety and accessibility of an offshore wind farm. Accessibility itself is particularly essential for offshore wind power systems, to guarantee reduction of the great financial risks due to doubts to the accessibility and reliability of OWT [9].

Maintenance technicians' transportation to the OWTs shall be carried out by workboats, which are limited by wave height [8].

#### 1.1.4. Operation and Maintenance Plan

Maintenance planning is the prioritization of maintenance tasks ahead of available resources (for example, personnel, maintenance equipment, and spare parts). Maintenance planning involves all maintenance tasks, and the optimization process can achieve great savings. Mainly, cost savings are correlated with current assets (fuel, mobilization costs, production losses, and logistics costs) [9].

Managing operation and maintenance activities to reduce OPEX (operating expenses) costs is one of the most decisive challenges of offshore wind farms, due to the distribution of maintenance varying with time depending on the performance of OWTs and their subassemblies, as well as the weather window. Thus, to determine operation and maintenance activities, project managers need to have a clear understanding of sub-assembly history, background, performance, and weather [9,10]

Maintenance program activity triggers are usually failures of a component/element or a time interval based on operational service principles.
