The Impact of Contractual Arrangement and Risks on Maintenance Decisions: A Literature Review in Offshore Wind Farms

(1)

The Impact of Contractual Arrangement and Risks on

Maintenance Decisions: A Literature Review in

(2)

Page 2 of 64

Abstract

Since the past decade, the offshore wind farm has gradually grown into a thriving industry for sustainably producing green and renewable electricity. Research has revealed the factors that have significant roles in solving the problems related to the maintenance of offshore wind turbines. Hence, there are typical challenges in this industry, such as complex coordination amongst stakeholders (manufacturers, operators and service providers), hazardous environments, and long product life time (10-15 years). In this research, the maintenance of offshore wind turbines was investigated in terms of aspects that are relevant to decision-making, contractual arrangements, and risks.

By performing a systematic literature review, this research has recognised five aspects relevant to maintenance decisions according to the dispersion of articles, namely: the base of operations (3 articles); inventory management (7 articles); maintenance support organisations (13 articles); maintenance strategy (33 articles); and routing and scheduling (9 articles). In addition, there are six articles related to contractual arrangements and sixteen articles related to risks. Our findings indicate that contractual arrangements and the base of operation have the least attention in the literature, whereas maintenance strategy has the most attention from researchers.

Furthermore, by investigating cross-section literature, this research has found a relationship among these variables. The finding revealed a substantial connection between risks and maintenance decisions (13 articles), specifically for the operational risk that has interdependent relationships for most aspects of maintenance decision. Simultaneously, this research has found the adoption of contractual arrangements affecting both risks and maintenance decisions. To start with, contractual arrangements have various impacts against the risks and there are different responsibilities of every stakeholder to bear the risks. Also, the influence of contractual arrangements was indirect towards maintenance decisions, because the duty for decision-making was determined by the stakeholders in which this research has discovered.

Finally, this dissertation has both contributed and extended the studies of asset management, particularly the maintenance of expensive assets which were sparse in the literature. The contribution of this research can be used by practitioners and researchers to create an appropriate maintenance decision by considering contractual arrangements and risks and to further validate the framework with quantitative studies and empirical findings.

(3)

Page 3 of 64

Acknowledgements

First of all, I am really grateful for the blessing of Heavenly Father whom I serve and praise for. Then, I would like to take this chance and gives my highest gratitude to several people who have contributed to the completion of this research. First, I would like to thank my first supervisor and assessor Dr. Jasper Veldman from the University of Groningen who has fully supported me and provided the detailed feedback throughout the research. Then, similar appreciation is also given to my second supervisor, Prof. Christian Hicks from Newcastle University for a valuable insight on this project and the guidance on this project.

Also, I would like to give my thanks and gratitude to all people who help and assist me for the past 4 months, and with their help, I have been able to successfully complete this project. Special thanks for Andy Ma for his criticism and support on the grammar and structure of the thesis, and Jennifer who always encourages me at the toughest time on this project. Also, I would like to give gratitude for friends that embrace me in the time of boredom. Finally, I would like to thank my family, especially my parents who pray and give encouragement. Their support is very valuable so that I can strive and finish this project.

Groningen, 14 December 2016

(4)

Page 4 of 64

The Impact of Contractual Arrangement and Risks on

Maintenance Decisions: A Literature Review in Offshore

Wind Farms

MSc thesis, MSc Technology and Operations Management University of Groningen, Faculty of Economics and Business.

MSc thesis, MSc Operations and Supply Chain Management Newcastle University, Newcastle University Business School.

Author / researcher : Rio Ronald Hasudungan Student number : S3078175

Address : Celebestraat 25A-3

Groningen, 9718 JB

E-mail : r.r.hasudungan@student.rug.nl - r.r.hasudungan2@newcastle.ac.uk

Supervisor & assessor : Dr. Jasper Veldman Prof. Christian Hicks Institution : University of Groningen Newcastle University

Faculty of Economics and Business Newcastle University Business School

Address : Nettelbosje 2 5 Barrack Road

9747 AE Groningen Newcastle upon Tyne, NE1 4SE

(5)

Page 5 of 64

Globally, offshore wind energy has achieved remarkable success. A report from Global Wind Energy Council (2015) showed that in 2015, the total amount of electricity generated from wind energy was 12,107 MW. The trend in the past decade showed that offshore wind farms (OWFs) were preferable for power generation compared to onshore wind farms. Numerous OWFs have been installed in Europe, and the amount of electricity produced from this type of energy in this region reached 3,018.5 MW in 2015, a figure that indicates a solid growth of wind farm industry (EWEA, 2015). Currently, many OWFs have been installed in shallow water areas, such as the North Sea and the Baltic Sea. However, a recent trend has shown that some OWFs are moving from shallow water to deeper waters (Utne, 2010). The reason is that deeper water has more promising wind resources compared to shallow water; due to the latest technology advancement, a new generation of offshore winds farms is well-equipped with high capacity turbines that can generate more electricity (Crabtree et al., 2015).

Despite the fact that OWFs generate a considerable amount of electricity production, the exploitation of OWFs remains challenging and risky due to severe weather and harsh environments (e.g. high waves and strong wind) (Byon et al., 2010). This challenging arises throughout the whole OWFs lifecycle. The three main phases of OWF lifecycle are construction, operation and maintenance (O&M), and decommissioning (Tremeac & Meunier 2009). Among these phases, the operation and maintenance phase can be considered as the most challenging stage (Karyotakis, 2011). The reason for this difficulty is that the duration of an OWF’s operation can range from 15 to 20 years; in practice, it includes the complex process of maintaining components, dealing with challenging environments, and interacting with multiple stakeholders (Hassan, 2013; Utne, 2010; Joschko et al., 2015). Therefore, more research in OWF industry has focused on the maintenance phase and the aspects that are related to maintenance activities.

(11)

Page 11 of 64 for helping the operator to maintain OWFs (Utne, 2010; Kobbacy & Murthy, 2008), and the selection of a service provider can be directly done by the operator or based on an open tender.

Basically, the purpose of maintenance is to preserve assets at an optimal condition for a period (Tsang 2002). This condition is characterised by having a high level of availability with minimum operation and maintenance costs (Martin et al. 2016). According to Dinwoodie et al. (2015), there are three main elements in availability namely: reliability, accessibility, and maintainability (Karyotakis 2011). Therefore, to have a high level of availability, decision makers need to consider these three elements in designing and performing maintenance activities. Additionally, maintenance activities are costly, because it is widely believed that the costs incurred within the maintenance phase can be significant, which is 20-35% of the overall life cycle cost (Andrawus 2008; El-Thalji & Liyanage 2012; Shafiee 2015b). In order to reduce maintenance cost, several prominent aspects are cited in the literature. Among them are: the transportation method (Dalgic et al., 2015; Scheu et al., 2012), the maintenance strategy selection (Nielsen & Sørensen, 2011; Sarker & Faiz, 2016; Hofmann, 2011), and the inventory management (Jin, Tian, et al., 2012; Tracht et al., 2013).

Consequently, most studies in maintenance decisions of OWFs focused on technical aspects; however, in practice, the decisions on OWF maintenance depend not only on the technical aspects but also on the strategic levels of maintenance decisions. As recommended by Olsen et al. (2005) and Singh Panesar and Markeset (2008), maintaining an asset also requires proper contractual arrangements among the corresponding stakeholders. The importance of contractual arrangements ranges from managing complex procurement from multiple service providers, such as vessel providers, technicians, and spare part suppliers, to aligning different responsibilities of various parties (Olsen et al. 2005). Bijma (2015) identified contractual arrangements as legal agreements in the form of maintenance contracts between different parties, such as site owner, operator and service providers. The contractual arrangements govern the relationship between the parties as their responsibilities are critical for the successful execution maintenance activities (Olsen et al. 2005).

(12)

Page 12 of 64

economic loss or gain, physical damage or injury, or delay as a consequence of the uncertainty associated with any particular action or project’. Specifically, in this dissertation, risks are limited

to health and safety risks (EU-OSHA), price and quantity risks (TKI Wind op zee, 2015), and contractual risks (Van der Wal, 2014), all of which are vital in the maintenance phase of OWFs. The study of risk is necessary a necessity in this dissertation because research by Shafiee (2015a) suggested that neglecting risks have several consequences such as considerable loss of electricity production and significant costs to restore the wind turbine into a normal condition.

In conclusion, OWF maintenance decisions require proper contractual arrangements, and each stakeholder needs to consider risks. Gatzert & Kosub (2016) suggested that a proper and effective contractual arrangements can reduce the impact of risks associated with maintenance activities. Therefore, the purpose of this dissertation is to combine these variables and to create a holistic framework of maintenance decisions, contractual arrangements, and risks. However, in the literature, OWF maintenance has received little attention from researchers (Shafiee, 2015b) as the main focus were on the technical aspects of maintenance decisions and as well as relying on simulation or modeling. For instance, Ossai et al. (2016), Sinha and Steel (2015), and de Andrade Vieira and Sanz-Bobi (2013) proposed a model to identify the failure risk of critical components in wind turbine; on the other hand, Jin, Ding, et al. (2012) developed a contracting model to improve the reliability of a wind turbine, and Poore and Walford (2008) considered various types of contractual arrangements as a method of reducing maintenance costs.

(13)

Page 13 of 64

RQ1. “What are the relevant maintenance decisions, contractual arrangement types (including distinguished characteristics), and risks in the context of an offshore wind farm that is available in the extant literature?

RQ2. “Which and how do contractual arrangement and risks relate to maintenance decision in offshore wind farms?”

Referring the research questions, the scope of this research is too broad to be discussed in a single study because OWF maintenance may also include power station converter or subsea cables. Therefore, the scope of this research is limited to the maintenance decision of wind turbines. Furthermore, as mentioned above, the methodology used in this research was a qualitative method. As defined by Fink (2005), ‘a qualitative method is a systematic, explicit, and

reproducible method for identifying, evaluating, and synthesizing the existing body of completed and recorded work produced by researchers, scholars, and practitioners' (p. 3). This approach is

appropriate as there has been no theory that combines contractual arrangement, risks and maintenance decision in a single study.

(14)

Chapter 2 Methodology

Page 14 of 64

2. Methodology

The following section discusses the research methodology through systematic literature review, which will be used in this research.

2.1. Research Design

The purpose of this study is to fill the gap by identifying relevant literature and discussing the relationships between the variables. Therefore, this research was categorised as qualitative research with exploratory type. As suggested by Karlsson (2009), the aim of the exploratory study is to define and gain in-depth understanding of how the underlying factors are related and to explore the motivations or the reasoning behind a topic of interest.

(15)

Page 15 of 64 Figure 1. Research Design (adjusted from Fink (2005))

Firstly, this research has identified two research questions, which were posed in the previous

chapter. Secondly, the articles were searched using the Web of Science and EBSCOhost as the primary bibliographic databases. Thirdly, the selection of keywords or descriptors was based on each factor, namely maintenance decision, contractual arrangement, and risks. Furthermore, further research was performed to identify relevant aspects of maintenance decision (base of operation, inventory management, routing, and scheduling), contractual arrangement (e.g.

performance-based, joint-supervision and multiple contracts) and risks (e.g. operational risk,

(16)

Page 16 of 64

Fourthly, to obtain appropriate materials, several criteria were applied: (i) The articles should

be published between 2000-2016; (ii) the articles were peer-reviewed; (iii) the articles were associated with one or more of the following aspects: maintenance, contract or contractual arrangement, and risks; (iv) the articles should be related to the maintenance phase of wind turbines; (v) the articles were freely available by means of either RUG or Newcastle University proxy facilities; and (vi) The articles were selected from top quality journals (minimum impact factor was 1). Additionally, several criteria set as the boundaries to exclude irrelevant publications, include: (i) articles which discuss maintenance that relates to the environment or animals; (ii) articles without abstract and editorial notes; (iii) non-English articles; and (iv) trade publication articles.

Fifthly, the findings were reviewed by twofold: first, assess the appropriate journals based

on the criteria above and second, performs quick reading to find information associated with the descriptors and exclude irrelevant articles. Thus, to ensure the quality of this research, the journals are selected based on the rankings (top and very good quality) presented in RUG internet sites (see internet links in references list). This research also considers the impact factor of the journals, by checking the ISI web of knowledge, which can be accessed from the web of science electronic journal database. The impact factors in the last five years have been checked, and the scores range from 1 to 7.896.

Sixthly, Webster and Watson (2002) suggested that a forward and backward research should

be performed to supplement research and to identify whether additional materials exist.

Forward research is performed by investigating publications citing these papers, and backward

(17)

Page 17 of 64

Finally, this dissertation has investigated the extracted papers using critical analyses to

investigate the relevant aspects of maintenance decisions, contractual arrangements, and risks. Critical analyses have been performed to check relationship between these aspects and a framework was developed along with the propositions. The framework was the results of inductive approach that aimed to generate a meaning from the collected data (in this case literature) to investigate the patterns and connections, and to build a theory (Saunders 2012).

2.2. Distribution of Literature

This research referred to 88 articles with three categories. Accordingly, within the topic of maintenance decisions, there are five aspects: the base of operation (3 articles), inventory management (7 articles), support organisation (13 articles), maintenance strategy (33 articles), and routing and scheduling (9 articles). Additionally, 6 articles were related to contractual arrangements, and 16 articles have discussed risks in the maintenance phase of OWFs. The distribution of each literature is provided in Table 1 below and Figure 2 illustrates the dispersion of the literature based on the year of publication.

Articles Topics Number of articles 1. Maintenance Decision 1.1. Base of operation 3 1.2. Inventory management 7

1.3. Maintenance support organisation 13

1.4. Maintenance strategy 33

1.5. Routing and scheduling 9

2. Contractual Arrangement 6

3. Risk in Maintenance 16

Total 88

(18)

Page 18 of 64 Figure 2. Distribution of literature by year (per 16 November 2016)

(19)

Chapter 3 Result

Page 19 of 64

3. Result

3.1. The Scope of Research

The scope of this study is narrowed down into maintenance decisions from the perspective of a wind turbine service provider due to the following reasons: first, a wind turbine comprises critical components such as wind blade, gearbox, generator, and bearings having relatively high prices which are 38% of the total OWFs cost (Karyotakis, 2011). Second, because the role of wind turbines is vital in generating electricity, it is very important to keep each component from failure. After all, if the critical components malfunction, the wind turbine needs to be switched off to prevent significant electricity loss (Besnard, 2009). Third, the role of wind turbine service provider is vital for maintaining wind turbine in a good condition; however, findings indicate that wind turbine service provider ordinarily retains knowledge about the failing behaviour of each wind turbine (Walford, 2006; Bijma, 2015). Therefore, there is a need to investigates the characteristic of contractual arrangements which are typical in offshore wind industry.

(20)

Chapter 3 Result

Page 20 of 64 Figure 3. Maintenance Logistic Framework (Shafiee 2015b)

3.2. Aspects of Maintenance Decision

3.2.1. Selection for the base of operation (BoP)

The BOP may have the same meaning and function as a headquarter, which is a central place for all maintenance activities. Several considerations in selecting appropriate headquarter include the distance to the nearby OWFs and the format of the BoP. The distance is considered an important factor because in the sea, the weather can easily change. So, when severe weather occurs, the base is deemed necessary to provide shelter and accommodate technicians at idle time (Besnard, 2009).

(21)

Chapter 3 Result

Page 21 of 64 while a platform in the middle of the sea is an accommodation place that is erected near OWF, and a custom-made vessel (or mothership) is a ship that can accommodate technicians and carry necessary equipment or spare parts.

Overview. In the literature shows that a few studies consider the BOP as the focus of the study.

De Regt (2012) suggested that an appropriate location can be determined and solved using algorithm formula. In a single location, the problem can be formulated as ‘Weber’ problem, which can be solved with ‘Weiszfeld’ algorithm. On the other hand, in multiple locations, the problem can be solved with ‘the adaptive location-allocation’ algorithm. Recently, a paper by Akbari et al. (2017) proposed a port selection model that has a similar function with the BOP. In the article, an empirical research was performed by industry experts to extract information about the most critical criteria that need to be considered in selecting a port. Each criterion is weighted accordingly, and by applying analytical hierarchy process (AHP), the research can determine the most suitable port to support maintenance activities. Besnard et al. (2013) proposed a model to determine the most optimal technique for selecting a location by considering transportation method, maintenance scheduling activities and economic aspects of the site.

3.2.2. Inventory management (IM)

Inventory management is essential for maintaining wind turbines, and it is becoming more critical for any wind farm located in remote areas, such as deep sea water where maintenance access is limited (Utne 2010). Inventory management in the OWF is associated with activities such as warehousing and spare parts stock policy. Among these activities, spare parts management has a more important role in inventory management, because most wind farm components are very expensive (Cheng & Prabhu 2012). Therefore, maintaining a reasonable amount of spare parts is vital, because ordering some missing spare parts directly from OEM requires significant time and may cause long operational downtime (El-Thalji & Liyanage 2012).

Overview. The existing literature showed that significant research had been done relating to

(22)

Chapter 3 Result

Page 22 of 64 replacement of gearbox for more than ten years, and reduce the cost to purchase new components. Dewan (2013) investigated the critical components of offshore wind turbines and proposed a stochastic model to determine the best policy for inventory management (single-item versus multi-item) linked with supporting organisations such as technician and transportation modes. Tracht et al. (2013) developed a generic model for spare part planning with several considerations such as weather condition and vessel readiness to be deployed. In addition, Tracht et al. (2013) has concluded that utilization of appropriate transportation modes could reduce maintenance cost. Meanwhile, Jin, Tian, et al. (2012) developed a model that is optimised with genetic algorithm method to determine the most optimal maintenance strategy and spare parts management for third-party spare parts provider. Gallo et al. (2012) also designed a model to maximize profit by examining the different warehouse locations and its relation with maintenance scheduling. On the other hand, Lindqvist and Lundin (2010) proposed a model associated with spare parts and stock policy in order to determine the most optimal and cost efficient strategy for inventory management by analysing various critical components combined with supporting organisation. Nnadili's (2009) research focuses on planning an inventory policy based on the maintenance strategy selection and weather condition in floating OWF.

3.2.3. Maintenance support organisation (MSO)

(23)

Chapter 3 Result

Page 23 of 64 gearbox, is immense. Additionally, an operator needs to have sufficient technicians and transportation ships in the base because the likelihood of failure in a wind turbine is probabilistic. Thus, it is deemed necessary to have a support organisations standby in the base for a certain period (e.g. during summer seasons) (Shafiee, 2015b; Utne, 2010).

Overview. In the literature, there has been research that discusses maintenance support

(24)

Chapter 3 Result

Page 24 of 64

No Articles

Method Transportation Method Parameter

Modeling or Simulation Optimization CTV Vessel Heavy lift or Jack Up vessel Helicopter Charter Rate Accessibility Failure Rate Electricity price 1 Sperstad et al. (2016)      2 Martini et al. (2016)      3 Gundegjerde et al. (2015)        4 Dalgic et al. (2015)        5 Dalgic, Lazakis, Dinwoodie, et al. (2015)        6 Dalgic, Lazakis, Dinwoodie, McMillan, Revie, et al. (2015)  *Mothership Vessel    7 Cao (2015)        8

Dinwoodie & McMillan

(2014)     9 Dalgic et al. (2014)      10 Halvorsen-Weare et al. (2013)         11 Besnard et al. (2013)      12

Van Bussel & Bierbooms

(2003)      

Table 2. Maintenance support organisation literature database

*Note: - Accessibility is associated with weather, wind speed and wave height.

(25)

Chapter 3 Result

Page 25 of 64 3.2.4. Maintenance strategy (MS)

Principally, maintenance strategy can be divided into three types: preventive maintenance (PM), corrective maintenance (CM), and opportunistic maintenance (Karyotakis, 2011). The difference is that corrective maintenance is performed after a failure appears, while preventive maintenance is carried out before a malfunction (Nilsson & Bertling, 2007). Additionally, in opportunistic maintenance, repair tasks are simultaneously conducted for more than one components to conserve both time and cost (Taghipour & Banjevic, 2012). In corrective maintenance, impermanent repair can be provided first so that the system can be function again before conducting proper and permanent repair later (Karyotakis, 2011).

On the one hand, time-based maintenance is performed regularly and is carried out with predetermined interval or schedule (Nilsson & Bertling, 2007); on the other hand, age-based maintenance relies on life distribution of the components (Alsyouf, 2011). Meanwhile, a condition-based maintenance is performed based on component’s health conditions. There are several advantages for this type of maintenance, such its capability to evade an unnecessary investigation from time-based maintenance and precipitous failure from corrective maintenance (Tian et al. 2011a).

In OWFs, a corrective maintenance is performed one to four times annually, with an average repair duration of 2-4 days (Utne, 2010). However, to achieve maximum results, reconditioning or overhauling is performed every four years and after twelve years, the parts need to be substituted with new ones (Van Bussel et al., 2001). For PM, planning and schedule is necessary because maintaining wind turbine is constrained by the weather condition of a season. For instance, when there is a potential failure spotted during the summer, maintenance decision needs to be made immediately either to substitute the component or to postpone the maintenance until the next summer (Utne, 2010). Ultimately, good maintenance strategy should combine both PM and CM.

Overview. In the literature, numerous studies have been aimed at either optimising or simulating

(26)

Chapter 3 Result

Page 26 of 64 opportunistic maintenance (6 articles). In Table 3, the largest share of studies focuses on condition-based maintenance compared to other strategies. The finding is obvious because this type of maintenance has several advantages such as (i) reducing maintenance costs (Besnard, 2013), (ii) preventing unnecessary wind turbine shutdown (Zhang et al., 2015), (iii) ensuring a high level of accuracy (Besnard & Bertling, 2010), and (iv) reducing failure rates and increasing wind turbine availability (Byon et al., 2011).

Maintenance Strategies Articles 1. Preventive Maintenance:

1.1. Time-Based Andrawus (2008); Karyotakis (2011)

1.2. Age-Based Shafiee & Finkelstein (2015); Ding & Tian (2012); Jin et al. (2013); Sarker & Faiz (2016)

1.3. Condition-Based May et al. (2015); Shafiee (2014); Tchakoua et al. (2014); Zhang et al. (2015); Flory (2013); Besnard (2013); Crabtree et al. (2014); Byon (2013); Yang et al. (2012); García Márquez et al. (2012); Tian et al. (2011b); Sheng & Veers (2011); Byon et al. (2010); Besnard & Bertling (2010); Hameed et al. (2009); McMillan & Ault (2008); Zhang et al. (2015); Van Horenbeek et al. (2013); Byon et al. (2011)

2. Corrective Maintenance Andrawus et al. (2007); Van Bussel et al. (2001); Obdam et al. (2007)

3. Opportunistic Maintenance Zhang et al. (2016); Shafiee (2014); Ding & Tian (2011); Besnard et al. (2009); Al-Najjar & Alsyouf (2003)

Table 3. Maintenance strategy literature database

3.2.5. Routing vessels and scheduling maintenance activities (R&S)

(27)

Chapter 3 Result

Page 27 of 64 sufficient number of technicians who possess the suitable skills to perform the tasks (Irawan et al., 2016).

Overview. Recently, a paper by Froger et al. (2016a) investigates maintenance scheduling in the

context of energy industry market and reviews the regulation and policies adopted in a particular environment. Irawan et al. (2016) proposed a new model to solve routing and scheduling problems by considering several bases of operation and the opportunity to service multiple wind farms simultaneously. Froger et al. (2016b) developed an integer linear programming and a constraint programming model to solve scheduling problems by taking technicians into account due to their possession of specific skill. Additionally, the research contemplating the duration of tasks as well as the resource requirements are the effect caused by numerous task execution modes. Xiang et al. (2016) recommended a scheduling model that considers peak regulation penalty fee to design an appropriate maintenance schedule. Both Stålhane et al. (2015) and Odeskaug and Raknes (2015) studies have developed a routing and scheduling model for vessel fleets combined with technician expertise. In the latter research, the model is extended with additional consideration for maintaining multiple wind farms, and the synergy effect of a joint vessel fleet between two or more wind farms in the same area shows the robustness of the model. Perez-Canto and Rubio-Romero (2013) proposed an optimal preventive maintenance schedule that considers availability and reliability. Meanwhile, Amayri et al. (2011) created an optimal model for condition-based maintenance that considers various types of wind turbines and lead times.

3.3. Contractual Arrangement

(28)

Chapter 3 Result

Page 28 of 64 depending on the contractual arrangement adopted, the involvement of these mentioned parties may differ from one OWF to another.

Category of Stakeholders Description of Responsibilities

Site owner The rightful entity that owns the OWFs

Original Equipment Manufacturer (OEM)

Conducting maintenance activities and guaranteeing availability (in warranty period)

The Wind Farms Operator Responsible for the operations and the control of the site

Independent maintenance service provider (IMSP)

Operates with independent brands to maintenance OWFs, which often provide lesser cost compared to OEM.

Logistic service provider Provides logistic service packages to support the activity

Fleet provider Offers transportation vehicles such as vessel, crane or helicopter

Spare part suppliers Provides components for replacement

Customer A customer can be referred to utility companies

Table 4. Responsibilities of Maintenance Stakeholders (Bijma 2015)

Establishing a contract is an important part prior to performing maintenance activities because in this stage, the scope of repairing task and the responsibilities of the sub-contractor need to be well defined (Stremersch et al. 2001). According to Murthy et al. (2015), a maintenance service contract is a legal document agreed by both parties (the owner or operator and the sub-contractor), and it involves technical, economic and legal matters. In OWFs, maintenance activities are often performed by OEM and it is based on the agreement between an operator and service providers. Thus, this agreement is designed in the form of a contractual arrangement in which the relationship and responsibilities of these parties are officially stated.

(29)

Chapter 3 Result

Page 29 of 64 guarantees the certain level of availability (Andrawus et al. 2007). Therefore, other parties’ responsibility (see Table 4) is limited, because OEM and its subsidiary companies are more responsible for performing most of the maintenance activities.

However, after the warranty period ends, the owner needs to decide whether to renew the maintenance agreement, to develop expert for in-house maintenance, or to hand it over to the service providers (Petersen et al., 2016). In terms of contractual arrangement, this research adopts several types of maintenance contract commonly used in offshore oil and gas (O&G) industry due to the similarities such as capitally intensive, complex coordination and management of stakeholders, and weather condition limitation (Singh Panesar & Markeset, 2008). Thus, based on this similarity, this research identifies three main contract types: (i) performance-based contract, (ii) work-package contract, (iii) facilitator contract (Tsang, 2002). Additionally, more knowledge regarding maintenance contract is derived from a report by Hassan (2013) who suggested hands-on, hands-off, and hybrid approach as typical maintenance contracts in OWFs. Consequently, to deliver comprehensive explanation about contractual arrangements, the combination of those two researches is described below.

First, in a performance-based contract, the site owner intends to hand over a wide-range of maintenance activities to a single service provider (Tsang, 2002). This type of contract mandates the owner to supervise the maintenance activities performed by a service provider while focusing on the availability of the wind turbines. Availability has an important role because it is a key factor or benchmark when the contract between an owner and a service provider is being designed. Two types of guarantee commonly offered by service providers are time-based and energy-based availability (Conroy et al., 2011). The difference can be seen from the proposed package;

time-based means that the service provider guarantees that the wind turbines will be able to operate

within the defined period (Conroy et al., 2011), whereas energy-based means that the service provider guarantees the yield produced by wind turbines, and this yield is determined by wind strength, machine and farm design (Dai et al., 2015).

(30)

Chapter 3 Result

Page 30 of 64 include developing internal expertise and capabilities to minimise dependency on OEM. Even though the cost of externalising maintenance activities is much higher, it is favourable to perform maintenance with internal resources because it is economically feasible in the long term (Murthy et al., 2002). This type of arrangement is frequently found in large-scale wind farms with a large utility company as the major shareholder (Lema et al., 2014). The core business of these companies is mainly related to electricity productions which show great interest for OWF market (Markard & Petersen, 2009). However, to maintain such large-scale of OWF, both the owner and the operator need to be accompanied with OEM several years after the expiration of the warranty period. Then, OEM technicians gradually remove their responsibilities to the new in-house maintenance teams.

Third, in multiple contracts, a site owner establishes multiple contracts with various parties and leases most of the resources (Hassan, 2013). A concept of Tsang (2002) about work-package

contracts and facilitator contracts is appropriate to explain the type of job done by a service

provider because the fundamental difference of these contracts depends on both maintenance activities and equipment utilisation. Work-package contract is determined based on the number of maintenance tasks or activities, while in facilitator contract, the owner only rents physical assets owned by a service provider and operates the assets independently (Tsang, 2002). However, in the context of wind turbine maintenance, the concept of facilitator contract is not relevant for usage as the assets (wind turbines) are not leased to other parties.

Overview. The literature shows that contractual arrangement in maintenance phase has received

(31)

Chapter 3 Result

Page 31 of 64 Lema et al. (2014) identifies several service provider companies and the types of contracts offered. In their research, most utility companies are interested in adopting join-supervision contract for a large-scale OWF. Jin, Ding, et al. (2012) proposed a model in the context of joint-supervision contracts and found that optimum cost can be obtained by considering critical spare parts stock (e.g. gearbox) and wind turbine availability. Danilovic et al. (2010) investigated factors that shape customer perception and satisfaction related to the effectiveness of contractual arrangements. Poore and Walford (2008) suggested that in the early years after the operation, it is preferable to outsource maintenance activities as it may decrease maintenance costs.

3.4. Risks during Maintenance Offshore Wind Farm

As mentioned in the research conducted by Gatzert and Kosub (2016), several risks associated with maintenance phases include the price and quantity risk, the operational risk (e.g. downtime and failures), and the contractual risk. Additional aspects of risk are taken from Ashrafi et al. (2015) who suggested that health and safety risks needed to complement the above-mentioned risks.

3.4.1. Price and quantity risk

Both price and quantity risks influence each other. The correlation can be seen as the amount of electricity produced by wind farms corresponds with the predetermined electricity price given by the government. The nature of the price and quantity risks also differs, depending on the regulation in the corresponding country. In this research, the price and quantity risk is taken from a report by TKI wind Op Zee (2015), and it focuses on the calculation of the electricity price and the amount of subsidy given for a certain period respectively. The purpose of adding these risks is related to the feasibility of performing maintenance, which needs to be adjusted with the electricity price in a certain period.

(32)

Chapter 3 Result

Page 32 of 64 based on the electricity price in the market and whether the newly corrected market price is set annually or monthly. The report suggests that it is more favourable to have the market price closer to an hourly index because it is more accurate to the real situation in comparison with the monthly or annual average. Therefore, in the context of maintenance of OWFs, an operator is exposed to the price risk when there is an attempt to perform maintenance.

Quantity – Quantity risk considered in this research is related to the subsidy scheme given by the government to support a certain number of hours to be paid; in this context, the subsidy grant is calculated based on the full load hours (TKI Wind Op Zee, 2015). Full load hours refer to the numbers of equivalent hours that a machine (wind turbines) generates a certain amount of electricity at its full capacity in a year (EC, 2014). Moreover, quantity risk is related to whether or not the subsidy on full load hours is capped because capped full load hours are unfavourable for the site owner or because the profit is limited. Quantity risk in maintenance phase is important to ensure project continuity because when the subsidy is insufficient, site owners and operators need to re-evaluate the maintenance strategy or propose a new price.

3.4.2. Operational risk

The design of an offshore wind turbine needs to satisfy the availability requirements guaranteed by the OEM. Indeed, satisfying the availability requirement is not an easy task as it is subject to the operational risk that occurs in the maintenance phase of OWF. Thus, operational risk in this context is related to component failures and port authorities. First, several critical parts inside a wind turbine are essential to produce electricity, and among them are three parts that often fail. These parts are rotors, bearings, and gearbox, all of which have become the primary focus in wind turbines maintenance (Bhaba, Sarker, & Faiz, 2016). However, as the knowledge about these components is limited, the majority of OWF components are supplied by OEM (Jin, Ding, et al., 2012).

(33)

Chapter 3 Result

Page 33 of 64 issued, which can take up to several weeks. Consequently, planning and scheduling maintenance activities are necessary to reduce OWFs downtime, and spare parts need to be stocked nearby so that maintenance can be performed immediately after the issuance of the permit.

3.4.3. Contractual risk

According to Irimia-Diéguez et al. (2014, p. 413), a contractual risk is defined as ‘the

renegotiation of the contract, such as midstream change of project scope and issues caused by imprecision and vagueness in the contract.’ Crocker and Reynolds (1993) argued that the

underlying idea of an optimal contract is the completeness of contractual clauses and the sufficient balance between costs and benefits. Consequently, when the contract is implemented, neither party feels aggrieved. However, establishing OWFs maintenance contract is complicated because there is a possibility of having poor contract between the operator and the service providers as identified by Van der Wal (2014). He argued that poor contract may lead to delayed process, extra costs, and, in the worst-case scenario, additional work because of technical inexperience. Therefore, an operator needs to consider a solid track record in establishing a contract with service providers. In a contractual agreement, important considerations are factors such as a good reputation with strong track record performance and high financial rating (EWEA, 2013). Among these, the latter factor is important because OWFs maintenance cost is so high that service providers should have sufficient budget to cover these expenses. Consequently, proper financial stability is necessary for service providers to assure the operator that maintenance activities can be performed accordingly without any detention.

3.4.4. Health and safety risk

(34)

Chapter 3 Result

Page 34 of 64

Overview. In the extant literature, research about risks in maintenance phase are more

pronounced in various journal databases. However, as the knowledge of risks has been increasing in the last few years, there is a growing consideration of risks as part of maintenance activities. The summary of risks found in the literature is summarised in Table 5 below. Moreover, it was noticed that most research focused more on operational risk than other types of risks. Sixteen articles written by different authors discuss this risk as the main point of their research. Furthermore, findings show that most research aims to reduce the amount of failure in the maintenance phase and increase wind turbines availability. A comprehensive study of risks can be found in a paper written by Gatzert and Kosub (2016) who had deliberately discussed the classification of risks in OWFs, including risks mitigation, risks transfer and risk avoidance.

No Articles

Category of risk Price &

quantity

Operational Contractual Health and safety

1 Ossai et al. (2016) 

2 Gatzert & Kosub (2016)   

3 Zitrou et al. (2016)  

4 Ashrafi et al. (2015)  

5 Sinha & Steel (2015) 

6 Shafiee (2015a)  

7 Staid & Guikema (2015) 

8 Shafiee & Dinmohammadi (2014) 

9 Bangalore (2014) 

10 de Andrade Vieira & Sanz-Bobi (2013)



11 Douard et al. (2012) 

12 Kahrobaee & Asgarpoor (2011) 

13 Nielsen & Sørensen (2011) 

14 Vijayakumar & Rao (2007) 

15 Van (2012)  

Table 5. Category of risks during maintenance

(35)

Chapter 3 Result

Page 35 of 64

3.5. Cross-section literature

This section will identify cross-section literature that discusses the correlation between the variables (maintenance decisions, contractual arrangements, and risks).

3.5.1. Risks and maintenance decisions

Overview. In the literature, there are several studies that correlate the risks and maintenance

(36)

Chapter 3 Result

Page 36 of 64 maintenance activities. Lindqvist and Lundin (2010) studied critical components that frequently experienced failure, and they developed the most optimum model to manage spare parts stocks. The cross-section literature between risks and maintenance decisions is provided below (see Table 6).

No Author Risks type Correlation with Maintenance

Decisions 1 Ossai et al. (2016) Operational risk MS 2 Gatzert and Kosub

(2016)

Operational risk, price and quantity, contractual

IM, MSO, MS

3 Zitrou et al. (2016) Operational risk MS, IM, MSO 4 Ashrafi et al. (2015) Operational risk, health and

safety risk

MS, MSO, IM

5 Sinha and Steel (2015) Operational risk MS 6 Shafiee (2015a) Operational, health and safety MS 7 Shafiee &

Dinmohammadi (2014)

Operational risk MS

8 Bangalore (2014) Operational risk MS

9 de Andrade Vieira & Sanz-Bobi (2013)

Operational risk MS

10 Nielsen and Sørensen (2011)

Operational risk MS, MSO, R&S

12 Van (2012) Health and safety MSO

13 Lindqvist and Lundin (2010)

Operational risk IM

Table 6. Cross-section literature between risks and maintenance decisions

3.5.2. Contractual arrangements and maintenance decisions

Overview. In the literature, there are few articles that discuss contractual arrangements and

(37)

Chapter 3 Result

Page 37 of 64 3.5.3. Risks and contractual arrangements

Overview. In the literature, risks and contractual arrangements are rarely discussed. Especially

(38)

Chapter 4 Framework and Propositions

Page 38 of 64

4. Framework and Propositions

In the following chapter, the result of systematic literature review will be synthesized to develop a comprehensive framework of maintenance decisions, contractual arrangements, and risks.

4.1.

Framework of maintenance decisions, contractual arrangements, and risks

The result derived from the systematic literature review showed that there is a relation between maintenance decisions, contractual arrangements and risks in the context of the OWFs. The following paragraph will explain the relation between each variable.

4.1.1. The influence of risk in maintenance decision

The influence of risks in maintenance decisions will be discussed by dividing risks into two categories: (a) contractual risk as well as price and quantity risks (b) operational as well as health and safety risks. The difference between these two categories is the degree of influence as most risks in the former has an influence in maintenance decisions, whereas the latter has an interdependent relationship.

(a) Price and quantity

A decision to select either location or a format of the BOP is not relevant to the price and quantity risk. This irrelevance is because the price of electricity and the numbers of full load hours are determined by the government. However, a different condition appears in a decision to adopt a certain type of maintenance strategy (e.g. PM or CM or both). The influence of price risk is relatively high because in some countries, the electricity price can fluctuate (TKI wind Op Zee, 2015). Therefore, it is favourable to forecast the electricity price (monthly or annually) so that a suitable maintenance strategy can be arranged. Furthermore, the influence of price and quantity risk is not relevant to the decision of inventory management, support organisations, and routing and scheduling. It is pertinent that these three aspects need to work efficiently in order to preserve the conditions of offshore wind turbines and generate steady power.

(39)

Page 39 of 64 The influence of contractual risk to the format or the location of a base of operation is relatively low, because it shows that an operator prefers to develop an area as the central point for maintenance activities. Therefore, the influence of the contractual risk is limited in this context. However, regarding inventory management and support organisations, the influence of a contractual risk is relatively high since these two aspects are associated with one or more service providers. The influence of contractual risk is related to the decision to select suitable service providers and designing a proper agreement that contains responsibilities and legal conditions between an operator and a service provider. Moreover, the influence of contractual risks is relatively high to the decision for routing and scheduling because it affects the terms and conditions for utilizing suitable vessels as well as administering the responsibilities of technicians.

(c) Operational Risk

(40)

Page 40 of 64 number of component failures is relatively low; and (ii) the implementation of PM can reduce the expensive cost of replacing major components and increase their lifespan. Maintenance strategies need to be correlated with decisions for routing and scheduling because planning a PM or CM is different. While component failures occur, a decision to select an optimal route should be combined with the schedule of the technician, the routing of the vessel, and the list of spare parts to be deployed (Irawan et al., 2016).

(d) Health and safety risk

(41)

Page 41 of 64 Figure 4. The influence of risks to maintenance decisions

Based on these findings, the following propositions are developed:

P1a. Price and quantity risk has a low influence on every aspect of maintenance decisions, except

for such decisions as selecting the appropriate maintenance strategies.

P1b. Contractual risk has a low influence on the decisions to select the BOP and maintenance

strategy, whereas it has a high influence on decisions related to inventory management and support organisation.

P1c. Operational risk has an interdependent relationship for every aspect of maintenance

decisions, except for decisions related to support organisation (high influence).

P1d. Health and safety risk has an interdependent relation to the BOP, support organisation, and

(42)

Page 42 of 64 4.1.2. The influence of contractual arrangements

The influence of contractual arrangements are two-folds: The first is that contractual arrangements have various impacts on the risks associated with the maintenance of offshore wind turbines, and there are different responsibilities to bear risks for every stakeholder. The second is that the contractual arrangements do not directly affect decision making in an OWF; rather, this research notices that the influence of contractual arrangements is associated with stakeholders.

The utilisation of a performance-based contract provides more security and reduce the impact of risks as electricity production is guaranteed by service providers in terms of either energy-based or time-based availability. However, different conditions occur for the joint-supervision contract, in which the influence of this type of contract has medium effects on reducing the impact of risks because both parties (the service provider and the operator) are responsible for maintaining offshore wind turbines and bearing the risks together. Furthermore, in multiple contracts, the impact of risks is increasing because the operator solely handles the responsibilities for maintaining wind turbines. As for this type of arrangement, the operator is predisposed to bear the contractual risks because multiple service providers are involved. To sum up, these below are the responsibilities of every stakeholder – including the site owner, operator, OEM, and service providers to bear different types of risks depending on the contractual arrangements (see Table 7).

Risk Category Element

Contractual arrangement Performance-based Joint Supervision Multiple Contract Price and Quantity Risk

Price Owner Owner Owner

Quantity Owner Owner Owner

Operational Risk

Component Failures OEM OEM/Operator Operator/SP General Maintenance

Procedure

OEM OEM/Operator Operator/SP

Port Authorities OEM OEM/Operator Operator/SP Contractual risk

(43)

Page 43 of 64

Health and Safety Risk

Health and Safety All Parties All Parties All Parties

References (Utne 2010; Hassan 2013; Gatzert & Kosub 2016; Irimia-Diéguez et al. 2014; Lema et al. 2014)

Table 7. Stakeholder perspective to bear risk in different form of contract

Additionally, the influence of contractual arrangements on maintenance decisions is not direct because it is associated with multiple parties such as site owners, operators, and service providers. In short, it is similar to the influence of contractual arrangements and risks. Consequently, it is preferable to investigate the correlation between each stakeholder for every aspect of maintenance decisions in a different contract setting (see Table 8) and then discover the influence of contractual arrangements on maintenance decisions.

Table 8. The responsibilities of stakeholders in various contract setting

Note: “SP” stands for service providers such as a vessel provider, a spare parts provider, and consultancy companies.

In performance-based contracts, an OEM has significant influences for most aspects of maintenance decisions, except the decision to decide either a location or a format for the BOP because it is usually done by an operator. In addition, joint-supervision contracts, the influence of an OEM is mainly associated with inventory management (to provide spare parts and extra equipment), support organisations (to deploy technician and OEM vessels), and maintenance strategies (to guide an operator to maintain wind turbines). This influence is due to the capabilities and expertise owned by the OEM to guide the operator in early years, whereas an operator has more influence on the decision to select the BOP together with the OEM influencing

Maintenance Decision

Responsible Party

Performance-based Joint-Supervision Multiple Contracts Base of Operation Site Owner & Operator Site Owner & Operator Site Owner & Operator

Inventory Management OEM Operator & OEM Operator & OEM

Support Organisation OEM Operator & OEM Operator & SP

Maintenance Strategies OEM Operator & OEM Operator & SP

(44)

Page 44 of 64 support organisation and relatively having a small impact on other aspects of maintenance decisions.

Eventually, for multiple contracts, the influence of OEM are more pronounced in the decisions associated with inventory management (to supply critical spare parts), and it allows limited influences on other aspects. Additionally, an operator has a relatively high influence for the aspects that are related to the selection for the BOP, support organisations, and maintenance strategies because the operator has the experience to maintain wind turbines. In this type of contract, the service providers have additional influences, all of which are associated with two aspects: (i) support organisations such as vessels with specific abilities (e.g. jack-up and crane ships) or highly skilled technicians; and (ii) the utilisation of software created by a consultancy company to help an operator to design a suitable maintenance strategy. In sum, the influence of contractual arrangements on maintenance decisions depends largely on the various stakeholders who are responsible for preserving offshore wind turbines. The influence of contractual arrangements to maintenance decisions is illustrated in Figure 5 (see next page).

The following propositions were created based on these findings:

P2a. There are various effects of contractual arrangements (performance-based,

joint-supervision, and multiple contracts) of the risks due to the capabilities of stakeholders to bear the impacts of the risks.

P2b. The influence of contractual arrangements on the maintenance decision is not direct because

(45)

Page 45 of 64 Figure 5. The influence of contractual arrangements on

maintenance decisions

Performance-Based Contracts

Stakeholders Maintenance Decisions

Base of Operation OEM Inventory Management Operator Support Organisation SP Maintenance Strategies High Low Degree of influence

Multiple Contracts

OEM

Stakeholders Maintenance Decisions

Base of Operation Operator Inventory Management Support Organisation Maintenance Strategies SP High Low Degree of influence:

Joint-Supervision Contracts

OEM Operator Base of Operation

Stakeholders Maintenance Decision

(46)

Page 46 of 64 To sum up all the findings, this dissertation has created a final framework of maintenance decisions, contractual arrangements and risks in the context of OWFs. The holistic framework is illustrated in figure 6 below, and the detailed explanation about the correlation among each variable has been presented in the previous section.

Contractual Arrangement

Responsibilities PBC or Warranty OEM Purchase or Lease Owners Joint Supervision Performance Based Joint Supervision Multiple Contracts Risks B Risks A Price and Quantity Risk Operational Risk Contractual Risk Health and Safety Risk Maintenance Decision Base of Operation Inventory Management Support Organisation Maintenance Strategies Routing and Scheduling SP High Interdependent Low Degree of influence:

(47)

Page 47 of 64

4.2. Risk Mitigation on the Maintenance of Offshore Wind Turbines

As the commissioning of OWFs is growing, the risk is becoming one of the biggest concerns for their owners. This challenge is due to the reason that the method of funding the project is via debt, and in some cases, owners use their wind farms as a collateral asset (EWEA, 2013). Thus, managing risks in OWFs is an important issue that deserves further consideration because the lifetime of wind turbines is relatively long (around 15-20 years). In this section, the risk mitigation will be further discussed, including several propositions.

Price and quantity risk

The price and quantity risk is difficult to mitigate because the governments regulate the operations of OWFs (see TKI Wind Op Zee [2015] for more information). In several countries in Europe, the government subsidises the electricity price, which in this case depends on the subsidy scheme, the impact of price, and the quantity risks. However, if the price is not subsidised, the owner needs to maximise the output and reduce the impact of any other risks.

Operational risk

Operational risk can be mitigated by using a proven wind turbine technology, which has been rigorously tested by OEM, or by using a wind turbine that is common in the market. The reason to select a wind turbine that is common in the market is that it is very easy to repair or to replace components and are cheaper to maintain. Another reason is that it can rely on an independent service provider rather than an OEM. Also, to reduce the impact of operational risk, a new monitoring technology can be used to complement the existing SCADA system. As suggested in research carried out by Netland (2014), the application of robots can be used to check whether or not the failure signal from SCADA system is real, and it is favourable for OWFs located far away from shore. After all, the cost of deploying human technicians is quite expensive, and if the failure signal turns out to be erroneous, the operator may suffer from huge losses.

Contractual risk

(48)

Page 48 of 64 outstanding performance from the service providers (Sorrell, 2007). Therefore, to mitigate the contractual risk, the operator and service providers should design the contract rationally, including the scheme of penalty, if the service provider cannot deliver the intended results. Health and safety risk

In the last decade, studies to resolve health and safety specifically for the offshore wind farms flourished. The results can be seen from several government reports as follows: EU-OSHA (2013) and Renewable UK (2014). To sum up the information from these reports, we have include a summary to mitigate the health and safety risk as follows: (i) a standardized training protocol that is specifically created for offshore wind farm technicians, (ii) a stand-alone communication system with an ability to cover the whole site of OWFs, (iii) an advance gangway system from a vessel to the wind turbine (e.g. Ampelmann gangway), (iv) the utilization of a drone as the part of surveying equipment to inspect the wind turbine blade (iv) a regular fitness assessment to check technicians for their health. Finally, the risk mitigation technique is wrapped up as the following propositions:

P3a. The price and quantity risk can be mitigated by purchasing additional insurance from the

third parties in order to ensure that both the loss of electricity produced and such other losses are paid so that the target availability is always attained.

P3b. The impact of the operational risk can be summarised into three key points: One, the

operator needs to employ a wind turbine that has been technologically proven by the OEM; two, it is preferable to purchase wind turbines that are common in the market; three, the operator should deploy robots to check whether or not the failure signal from SCADA monitoring system is correct.

P3c. The contractual risk can be mitigated by designing an appropriate contract that includes

incentives and penalty schemes to justify the results from a service provider.

P3d. The impact of health and safety risk can be mitigated by ensuring standardised training for

The Impact of Contractual Arrangement and Risks on Maintenance Decisions: A Literature Review in Offshore Wind Farms

The Impact of Contractual Arrangement and Risks on

Maintenance Decisions: A Literature Review in

Abstract

Acknowledgements

The Impact of Contractual Arrangement and Risks on

Maintenance Decisions: A Literature Review in Offshore

Wind Farms

Table of Contents

List of Figures and Tables

List of Figures

List of Tables

List of Abbreviations

Glossary

1. Introduction

2. Methodology

2.1. Research Design

3. Result

3.1. The Scope of Research

3.2. Aspects of Maintenance Decision

3.3. Contractual Arrangement

3.4. Risks during Maintenance Offshore Wind Farm

3.5. Cross-section literature

4.

Framework and Propositions

Framework of maintenance decisions, contractual arrangements, and risks

Performance-Based Contracts

Multiple Contracts

Joint-Supervision Contracts

Contractual Arrangement