Safety costs in Arctic shipping: a proposal classification and estimation Open Access

The extreme climate and its unique environment clearly indicate that the wider Arctic region is still at high risk of ship operation. Accident and incident data covering 2005–2017 from six Arctic States (PAME, 2021) show 276 reported events in the region in 2017. The average number of incidents in a year increased by 40%, from 172 (2008–2012) to 244 (2013–2017). Among all accident types, equipment failure (618), marine pollution (612) and grounding (268) are the most common. This trend justifies investing in preventing Arctic shipping casualties.

Additionally, the remoteness and insufficient search and rescue (SAR) and salvage capability in the Arctic may aggravate accidents and lead to higher losses in terms of ship damage, cargo losses and casualties. The injury or death occurring in a harsh environment can be worsened by the difficulties of evacuation and repatriation of crew and passengers (Fedi et al., 2018). For example, on 27 August 2010, the passenger vessel MV Clipper Adventurer was grounded in Coronation Gulf, Nunavut, with 128 people on board. Fortunately, no injuries occurred, and there was little pollution. The estimated costs associated with the accident include $12 m in repair and salvage costs related to the ship’s hull, $2.6 m for loss of business, $445,361 in environmental costs and $350,000 in other costs (CBC, 2011, 2017). Some underwriters considered the sinking of MV MSC Concordia as an example that can represent the “worst-case scenario” in the Arctic, and one of them estimated that the costs associated with the accident could be around $1 bn if the case were in the Arctic (Fedi et al., 2018).

The growing marine traffic in Arctic waters increases marine pollution risks. A significant oil spill may have severe consequences in the Arctic, given its fragility – oil decomposes slowly in cold environments (Afenyo et al., 2016), infrastructure limitations, and a lack of local preparedness (Johannsdottir and Cook, 2019), challenging conditions (e.g. total darkness in winter, sub-zero temperatures and shifting ice) and inadequate means of collecting oil from ice-filled waters (National Research Council, 2014). Major oil spills in the Arctic can incur extremely high costs, including clean-up costs, community compensation, long-term pollution monitoring, wreck removal, lawsuits and administration of fines as well as indirect effects on fishing activities and tourism income (Johannsdottir and Cook, 2019; Allianz, 2021).

Consequently, Arctic coastal states and shipping companies involved in Arctic operations need to adopt effective safety control measures. Specific measures, such as the mandatory measures in the Polar Code implemented by the International Maritime Organization (IMO), can mitigate the risks related to Arctic waters navigation (Dalaklis et al., 2018a; Fedi et al., 2018, 2020).

However, the question remains: how much are maritime stakeholders willing to pay to prevent and respond to marine accidents, and how to determine safety investments?

Despite growing interest in Arctic shipping over the past decade, a thorough literature review suggests that the knowledge associated with safety costs is limited. To the authors’ best knowledge, no literature specifically discusses the safety costs in the Arctic shipping industry. A few studies (Furuichi and Otsuka, 2013; Lasserre, 2014; Pierre and Olivier, 2015; Faury and Cariou, 2016; Solakivi et al., 2019; Rigot-Müller et al., 2022) on the economic aspect focus on the competitiveness and profitability of the Northern Sea Route (NSR) or NorthWest Passage (NWP) and discuss capital cost, fuel costs, icebreaking fees, pilot costs and insurance costs. While an increasing number of studies on Arctic shipping safety (Drewniak et al., 2021; Benz et al., 2021; Gao and Erokhin, 2019; Giguère et al., 2017) advocate investing and strengthening the infrastructures, facilities, icebreakers and other technologies. A comprehensive safety cost analysis has not been undertaken. On the other hand, the costs of oil spills in Arctic waters (WSP Canada, 2014; Afenyo et al., 2019; Browne et al., 2022) have been discussed in the existing literature, but other accident costs have rarely been studied.

Consequently, this paper provides a comprehensive review of safety costs and paves the way for future safety investments for Arctic ship operators and Arctic coastal states.

The small number of accident samples associated with the region’s low volume of marine traffic challenges in-depth analysis. Despite such data-related shortcomings, such an analysis is necessary to orient the decision-makers in optimizing safety investments. From the perspective of economic evaluation, the costs of preventive and mitigating actions need to be compared with benefits in terms of avoiding accident costs (Ellis et al., 2007).

The main objective of this article is to identify safety costs and classify them on the basis of findings from relevant literature to provide suitable estimation methods for Arctic shipping. Following the literature review, section three proposes the classification of the safety costs and the estimation methods. Section four provides estimation checklists in practice and a test result with an Arctic ship operator. After the discussion, the conclusion summarizes the findings.

Safety costs exist as expenses from both the states and the enterprises. In a company, they usually relate to safe operation, including personal protective equipment (PPE) and safety management. In contrast, safety costs in a state are often fundamental, long-term and strategic expenditures (i.e. infrastructure and implementation of safety policies).

As a broad concept, safety cost does not have a widely accepted definition by scholars. The most common interpretation is that safety costs refer to health and safety costs (H&S costs) and are the sum of prevention costs and accident costs (Andreoni, 1986; HMSO, 1993; Brody et al., 1990). The former is used to ensure the safe operation (Ex ante), and the latter is used to restore the previous state post-accident (Ex post) (Brody et al., 1990).

Scholars have proposed various safety cost classifications in the literature, as listed in Figure 1. Pioneers in research on health and safety costs (Heinrich, 1931; Hinze, 1991) divided the costs of industrial accidents into direct costs and indirect costs. Later, Simonds and Grimaldi (1956) and Leopold and Leonard (1987) proposed dividing the safety costs into insured and uninsured costs. On this basis, Teo and Feng (2011) further expanded the uninsured costs of accidents into 14 possible components based on the existing literature.

In addition, scholars classify the safety costs according to other attributes. Laufer believes that from a management perspective, the distinction should be made not between uninsured and insured but rather between controllable and uncontrollable costs (Laufer, 1987). Controllable costs are expenses influenced by management involvement, whereas uncontrollable costs are cost allocations beyond the control of management. Another framework for health and safety costs (Riel and Imbeau, 1996) proposes to classify health and safety costs into three categories: insurance-related costs, work-related costs and perturbation-related costs. Other scholars classified safety costs into four distinct categories: economic and non-economic, internal and external, fixed and variable, direct (or visible) and indirect (or invisible) (Dorman, 2000; Battaglia et al., 2014). A new classification method proposes to divide the costs into safety costs and non-safety costs (Ibarrondo-Dávila et al., 2015; López-Alonso et al., 2016; Micheli et al., 2021). Safety costs include prevention costs and monitoring costs, while non-safety costs include tangible costs, intangible costs and extra costs.

The term “maritime safety” is commonly used in the maritime world, but there is no widely recognized definition (Praetorius et al., 2010). The United Nations (UN) describes the nature of maritime safety as follows: “Maritime safety is principally concerned with ensuring the safety of life at sea, the safety of navigation, and the protection and preservation of the marine environment” (UN General Assembly, 2008). Thus, the term “maritime safety” can be defined as “the combination of preventive and mitigative efforts and activities against potential hazards, to protect life and property at sea, and to preserve the marine environment, and those include regulations, navigational support, operational service, management of ships and seafarers, and advanced technologies.”

The safety costs in the shipping industry can be expressed in different ways. For example, according to the guidelines for formal safety assessment (FSA) of IMO, the costs of risk control measures (RCMs) should be expressed in terms of life cycle costs and may include initial, operating, training, inspection, certification and decommission (IMO, 2018). Therefore, scholars classified shipping safety costs into two components: guaranteed shipping safety costs and costs of shipping safety failure (Li and Zhang, 2014). The former represents the preventive costs by shipping companies and maritime administrations, while the latter refers to the loss resulting from accidents.

Authorities such as the United States Coast Guard (USCG) (U.S. Coast Guard, 1997) divided the costs of marine accidents into four categories: injury and death costs, property damage costs, oil spill costs and other costs. Examples of other costs are interruptions in operations, port fees and stock prices. It is noticeable that the effects of marine accidents on people, property, the environment, society and other parties have been fully considered under this classification.

Similarly, the IMO’s FSA guidelines present an example of a ship accident loss matrix composed of accident categories and categories of losses (IMO, 2018). Expanding into six categories, Grigalunas et al. (1986) considered the social costs of ship accidents: emergency response, clean-up and restoration, marine resource costs, recreation losses, losses to the tourism industry and other costs. A system diagram for the maritime traffic system (Walker, 2000) includes financial, economic, environmental and human consequences of shipping accidents.

Estimating marine accident costs (MAC) is complicated and unpredictable. Researchers usually base cost estimation on historical data from marine insurance companies, governments or other organizations. For instance, Talley (1995, 1999) and Talley et al. (2006, 2008) researched the determinants of property damage costs from the US Coast Guard data and investigated cost differentials among container, tanker and bulk vessels (Talley, 2002).

The assessment of losses due to oil spills in an accident can also be based on historical data. For example, O’Rathaille and Wiedemann (1980) derived the average social cost per collision and grounding based on the pollution incident record, pollution rates and the known historical costs incurred. Nevertheless, it may be difficult to quantify the indirect and hidden costs, such as reduced worker morale and productivity, eroding customer base, increased insurance costs and fines (Ornitz, 2001). Furthermore, Loureiro et al. (2009) suggested that the losses from oil spills in ship accidents should include environmental and passive losses and estimated the total environmental losses caused by the MV “Prestige” oil spill to be around 574 € m.

The literature search results in academic databases show that research on shipping prevention costs (SPC) and MAC is still insufficient because the specific costs (monetary values) are not often detailed and discussed. The limited access to insurer and state data following accidents hinders detailed quantitative analysis.

It is essential to define the scope of shipping safety costs before estimating them. Based on the literature discussed and adapted from Brody et al. (1990) and U.S. Coast Guard (1997), it can be concluded that shipping safety costs (SSC) include two broad categories: SPC and MAC, as expressed in Equation (1):

• (1)

  SPC represent the expenditures related to ship safety, health and safety standards, risk mitigation and elimination, reducing casualties and preventing marine accidents.

• (2)

  MAC are the sum of four components: injury and death costs, property damage costs, environmental damage costs and other costs.

To identify the diversity of safety costs, an analysis of the Arctic safety literature has been performed using keyword searches in Scopus, ScienceDirect and Google Scholar for the period from 2000 to 2023. Keywords used are the following: (1) Arctic shipping; (2) Arctic shipping safety; (3) Arctic shipping technology; (4) Arctic coastal states; (5) Arctic shipping companies; (6) Polar operation; (7) Arctic shipping investment; (8) Arctic shipping risk; (9) Icebreaker; (10) Polar Code and (11) Arctic shipping regulation.

Diverse by nature, safety in Arctic shipping is a complex topic with joint efforts and investments from multiple stakeholders in a wide range of fields (e.g. Arctic coastal states and shipping companies).

Stakeholders in Artic shipping safety mainly invests in personnel (e.g. classification societies), technologies (e.g. research institutes) and management (e.g. flag states). As a result, many costs are interrelated. For example, Arctic coastal states invest in icebreakers, while the shipowners pay icebreaking fees. The same applies to the ice pilotage and many other facilities.

The following five categories and sub-categories have been extracted from the literature review and summarized in Table 1.

This category includes costs incurred in safety-related infrastructure and facilities (and their maintenance costs). In the shipping industry, the conditions of infrastructure and facilities such as ports and rescue centers play an important role in shipping safety. For the Arctic coastal states, strategic investments range from improved ports, hydrographic mapping, SAR resources, navigational aids and icebreaking (Dalaklis et al., 2023a; Dimitrios et al., 2022; Giguère et al., 2017; Lajeunesse et al., 2011). For shipowners, the operation and certification of Ice Class ships are the main investments. The following bullet points list the sub-categories collected in the review.

• (1)

  Icebreakers from building or leasing to operation are important costs for coastal states (Dalaklis et al., 2018b; Drewniak et al., 2021). The estimated procurement costs of the three polar security cutters for the US Coast Guard are about $2.7 bn (Congressional Research Service, 2022), and the cost of leasing a similar icebreaker would be $142,605/day (Transportation Research Board, 2017).

• (2)

  Ports and hydrographic surveys constitute major coastal states efforts to safeguard navigation and provide support functions for vessels (Wang et al., 2019; Lajeunesse et al., 2011; Panahi et al., 2021). Under the Ocean Protection Plan (OPP), the Canadian government invested CAD 94.3 m in basic marine infrastructure (ports) in Northern communities since 2017 (Transport Canada, 2020) and committed a funding of CAD 84 m to improve the modern hydrographic service of mapping the seafloor in the Arctic from 2022, which directly contributes to safe navigation (DFO, 2022). The Russian Federation will invest 17.33 bn rubles from 2022 to 2035 in hydrographic surveys in the NSR (Government of the Russian Federation, 2022).

• (3)

  Ice class ships are also recognized in the literature as indispensable for operating in the region with or without icebreaker assistance (Gao and Erokhin, 2019). For shipowners, buying an ice class ship signifies a capital increase ranging between 5 and 146%, depending on the nominal ice class (Sibul and Jin, 2021).

• (4)

  SAR facilities enable intervention in the Arctic region to reduce the negative consequences of marine accidents, which is a significant investment for coastal states (Dalaklis et al., 2018b). Among the members of the Arctic Council, Russia had invested 910 m rubles ($30.1 m) in establishing ten SAR centers along the NSR (Buixadé Farré et al., 2014). Recently, Russia announced plans to invest 98.2 bn rubles ($1.67 bn) in constructing 30 SAR vessels for the NSR from 2025 to 2030 (Mikhailova and Tabata, 2024).

• (5)

  To overcome electromagnetic disturbances and gaps, communication and navigation facilities need enhanced attention (Kvamstad et al., 2009; Benz et al., 2021). For example, Russia plans to cover the NSR with the Differential Global Positioning System (DGPS) (Ostreng et al., 2013), and numerous signal-enhancing stations have been established to increase the reliability of satellite service (Milaković et al., 2018).

• (6)

  Investment in oil spill response facilities and hardware stockpiles is necessary in the Arctic area (Lajeunesse et al., 2011; Dimitrios et al., 2022). The Canadian Coast Guard has over 20 clean-up equipment depots in the Arctic (Council of Canadian Academies, 2016), but only one was designed for airlift (Tanker Safety Expert Panel, 2014).

Costs under this category are expenses that occurred on safety measures placed onboard the ship and external services paid by the shipping companies to the coastal administration or other parties to ensure the safety of the voyages.

• (1)

  Icebreaker assistance and pilotage services must be available and reliable for ship operators (Rajagopal and Zhang, 2021). The cost of pilotage in NSR is high ($10,000 for the 14,357 gross tons multi-purpose vessel “Yong Sheng” NSR voyage in 2013, at the rate of $2,000/day) (Liu et al., 2016). Various estimates for the icebreaking and pilotage costs per ton of cargo are given in previous studies, such as $15 (Li et al., 2015), $14 (Yan et al., 2015) and $20 (Wang and Shou, 2013).

• (2)

  Real-time communication facilitating the immediate transmission of important messages (Buixadé Farré et al., 2014) and aids to navigation and related services are necessary (Tseng and Cullinane, 2018). For instance, some shipping companies operating in the NSR use a range of communication and broadband services provided by MORSVYAZSPUTNIK, a Russian state company (Borisova et al., 2020).

• (3)

  Availability of weather and ice information services are critical factors influencing the route choice (Wang et al., 2018). For the MV “Yong Sheng” case, the ship was provided weather analysis services for navigation by Weathernews Inc. (Zhang and Zhao, 2015). In 2009, ships from the Beluga Shipping Group used their own meteorologists to provide the ice navigator with up-to-date data (Østreng et al., 2013a).

• (4)

  Increased risks mean augmented insurance premiums to cover seafarers and ships (Tetley, 2002; Lajeunesse et al., 2011). The premium can reach $130,000 for an ice-strengthened Panamax-size bulk carrier for a single voyage (Lackenbauer and Lajeunesse, 2014). Premium is also estimated as per day, e.g. $800/day (Arpiainen and Kiili, 2006), $1,150/day (Østreng et al., 2013b) and $3,344/day (Verny and Grigentin, 2009).

• (5)

  Other prevention and mitigation measures refer to equipment required by coastal states regulations (e.g. NSR regulations) and voluntary safety measures placed onboard by shipowners. For the MV “Yong Sheng” transit, the ship installed additional communication and navigation equipment, including MF/HF NBDP (Narrow Band Direct Printing), satellite telephone, GPS compass and 122.5 MHz VHP (Zhang and Zhao, 2015).

This category covers prevention costs spent on work personnel from shipowners and other organizations. Investment in personnel is essential to effectively mitigate the negative impacts of human elements contributing to Arctic accidents.

• (1)

  Seafarers with Arctic experience and extra crew strengthen the ship’s capacity to operate safely in the Arctic region. Shipowners may increase salaries beyond the market to retain specialized crew (Lasserre, 2014) and arrange additional crew members to mitigate the fatigue risk common to most ships (Smith, 2007). In many cases, the salary of experienced seafarers working on the Arctic routes is assumed to be 110% of that on the Suez Canal Route (Shou and Feng, 2015), while Song and Zhang (2013) assumed the salary to be 128% of that on the southern routes.

• (2)

  Ice pilots are well-trained and experienced in navigation in ice environments, and more importantly, they are familiar with the local weather and environmental conditions. Pilotage services have to recruit and retain qualified ice pilots. The salary of an ice pilot may largely depend on the local salary level, experience, qualifications and others. For example, the median wage of a ship pilot is CAD 38/hour and the higher wage is CAD 66/hour in Canada (Government of Canada, 2024).

• (3)

  Safety inspectors and safety administration staff conversant with Arctic shipping conditions are necessary (Chircop et al., 2018). For example, one of the initiatives in the Canadian Ocean Protection Plan is increasing the number of marine safety inspectors in Northern communities to improve the safety of marine vessels and their crews (Chircop et al., 2018; Transport Canada, 2021). The costs (salary and benefits) of having one marine safety inspector from Transport Canada in the northern communities can be around CAD 150,000.

• (4)

  Emergency teams and well-qualified SAR personnel are essential to effective emergency response (Benz et al., 2021). In many cases, local workforces (e.g. police, volunteers, Coast Guard Auxiliary) could support SAR staff, especially in minor incidents or remote communities (Byers and Covey, 2019). Salaries for a marine communications and traffic services officer supporting SAR in the Canadian Coast Guard can range from CAD 49,712 to 70,700, and those for a maritime SAR coordinator can range from CAD 93,148 to 106,393 (Coast Guard Canadian, 2024).

• (5)

  The need for personnel training has been stressed in numerous studies (Larsson et al., 2018), especially for crew members (Fedi et al., 2020; Khan et al., 2020). Besides the Polar Code requirements, practical training onboard and training for emergency response in the Arctic region (e.g. cold weather survival and medical procedures through telemedicine) should be provided as well. For example, the tuition for the USCG-approved basic and advanced Polar Code operations can be $1,400 and $2,200, respectively (MMA, 2024).

• (6)

  Extreme temperatures and severe weather require work personnel to wear suitable PPE, which should be specially designed for use in the Arctic environment (Power et al., 2016).

This category comprises the costs of developing and adopting new technologies contributing to Arctic shipping safety by shipowners, government agencies or other parties. It covers the purchase of new technologies, research and development costs and adoption costs.

• (1)

  Developments in communication and navigation technologies (Reid et al., 2020; Wang et al., 2021) are growing to facilitate the safe conduct of navigation. An example of recent development is research led by Kongsberg Seatex AS on investigating how broadcasting navigation augmentation messages can be leveraged for safe and efficient shipping navigation in the Arctic (Bohlmann and Koller, 2020).

• (2)

  Ice monitoring and weather forecasting techniques require merging multiple organizations’ data and prediction models. The US National Oceanic and Atmospheric Administration (NOAA) invests in research to advance sea ice forecasting, such as data assimilation and model experiments, to increase the accuracy of ice-edge predictions (Scott, 2020). C-CORE started to develop a Polar Code application software with a fund of CAD 145,882 from National Resources Canada in 2021. The application aims to provide the latest data in near-real-time and integrate multiple functions, including sea ice charts, sea ice satellite imagery, offline mode, Polaris Maps and ice drift products (C-CORE, 2024).

• (3)

  Drones in the Arctic can supplement many tasks at lower capital, fuel and crew costs (Lajeunesse et al., 2011). Autonomous underwater vehicles can also become hydrographic survey tools (Johansson et al., 2023). For example, since 2017, the Canadian government has invested CAD 36 m to procure drones for Arctic maritime surveillance (Transport Canada, 2020).

• (4)

  Big data analytics and artificial intelligence (AI) ability to integrate and process data will find future applications in Arctic shipping (Bohlmann and Koller, 2020; Dalaklis et al., 2023b; Dimitrios, 2023). The Artificial Intelligence Based Virtual Control Room for the Arctic (AI-ARC) project is a €7 m project launched in 2021 and funded by the European Union with the purpose of creating a new system based on AI that facilitates the creation of a situational picture in the Arctic Ocean region (LUAS, 2021).

• (5)

  Other technologies include anti-icing solutions, Arctic voyage optimization, icebreaking enhancement and Arctic oil spill response technology. Safe maritime operations under extreme conditions: the Arctic case (SEDNA), a project funded by the EU Horizon 2020 program with a budget of €6.5 m, created and demonstrated other technologies applied to Arctic navigation, such as a human-centered “Safe Arctic Bridge,” anti-icing solutions and safe Arctic voyage optimization (INEA, 2022). The Canadian government funded a multi-partner research initiative in 2019 with a budget of CAD 523,000 to test new methods and advanced technology for locating and cleaning up oil spills in the Arctic with greater efficiency.

This category includes the safety management system costs. It encompasses both ship and shore operations and includes regular safety inspections, internal audits, management reviews and reporting/follow-up of accidents, incidents and deficiencies.

• (1)

  Ice Class notation, the Polar Code compliance and other inspections conducted by multiple parties require safety inspections and audits. In the maritime literature, (Knapp et al., 2011) estimated inspection costs as follows: $10,879 for class annual survey, $1,238 for port state control (PSC) inspection in Australia, $1,188 for flag states inspection and P&I Club inspection.

• (2)

  Safety awards can promote enhanced health and safety management systems (Tait and Walker, 2000). Safety incentive programs for crew and shore staff can take the form of bonuses, time off, certificates of recognition and merchandise.

• (3)

  Safety committee meetings and drills form core activities for safety onboard a ship. They should take place regularly, taking into account the pattern of operation of the vessel and the arrangement for manning and with sufficient frequency to ensure continuous improvement (ILO, 2015).

• (4)

  Arctic coastal states, port states and flag states may face extra administrative costs in implementing specific regulations and requirements. Examples are the Arctic Coast Guard Forum established in 2015, the Joint Arctic SAR Arena (JASA) established by the Association of Arctic Expedition Cruise Operators (AECO) in 2023, and a possible Arctic Port States MoU in the future.

Accessing Arctic accident cost data is a difficult task, as the statistics on marine accidents in Arctic waters are still limited. The major issues are the lack of common standards (Fedi et al., 2024) and limited reliability (Milaković et al., 2018; Nevalainen et al., 2017). For example, data on insurance claims in the Russian Arctic are rather sketchy and not updated (Fedi et al., 2020, 2024), and various published studies have concluded that the existing Arctic accident data are currently discontinued, fragmented and not fully standardized per the IMO rules.

The major costs in this category are medical expenses, costs due to absenteeism (lost time), and compensation costs (Leopold and Leonard, 1987). Indirect expenses paid by the shipowner may also include wages for co-worker overtime, fringe benefits and the costs associated with employee replacement (Dorman, 2000; Rikhardsson and Impgaard, 2004).

• (1)

  Medical costs – Medical charges for treatment, hospitalization, transfer of injured crew to clinic or hospital and future medical treatment.

• (2)

  Compensation costs – Expenses paid by the shipowner over and above workers’ compensation insurance and the insurance coverage limits.

• (3)

  Costs due to absenteeism – The value of the work time lost due to absenteeism and associated expenses.

• (4)

  Non-medical rehabilitation – Expenses spent to facilitate the injured crew to return to work.

• (5)

  Employee replacement costs – Replacement costs for the crew due to disability, early retirement and resignation as a result of the accident.

As early as 1980, O’Rathaille and Wiedemann (1980) estimated a seaman’s life value to be between £100,000 and £120,000 (1978 prices). This value can be converted into between £763,090 and £915,708 (2017 prices). However, this estimation is commercially based and does not include many noneconomic costs (costs of society). For injury costs SAFEDOR (2007), suggests a range of $20,000 to $70,000 per injury in the maritime industry.

Collisions, groundings, fires, explosions and other occurrences can result in damage to vessels and public facilities such as quays, ranging from minor losses to total losses (U.S. Coast Guard, 1997). This category covers the costs of repairing the damaged ship, compensation for the damage to other vessels or facilities and expenses paid to deal with cargo loss or damage (Ellis et al., 2007; Finnish Maritime Administration, 2008).

• (1)

  Ship damage – Costs of repairing the damaged ship.

• (2)

  Cargo loss or damage – The expenses related to cargo loss or damage and compensations to the cargo owner in the case where the shipowner has incurred liability.

• (3)

  Damage to other parties – Costs of damage to other ships or facilities in the accident.

Early in 1997, the U.S. Coast Guard (1997), provided a $2.5 m damage average cost for a cargo vessel accident based on the USCG dataset, and the Canadian Marine Pilots’ Association (2017) used an inflation multiplier (1.488) to bring the cost up to $3.7 m per accident (2017 price). For ocean vessels, the average claim costs of a fire accident and a collision in 2023 are around $3.5 m and $1.1 m, respectively, according to Cefor’s ocean hull portfolios (Cefor, 2024).

The most common environmental damage is oil spills. Environmental damage costs include clear-up costs, recovery costs, and compensation to local communities (Grigalunas et al., 1986; Afenyo et al., 2019).

• (1)

  Clean-up costs – Expenses paid to clean-up the pollution and remove debris.

• (2)

  Recovery costs – Expenses paid to restore the damaged environment after the accident.

• (3)

  Compensation to local communities – Compensation is paid to local communities due to the accident’s adverse effect on local people or industries.

For oil spill costs, the SAFEDOR project (Skjong et al., 2005) suggests a global average cost of USD 60,000 per tonne, which does not count for socioeconomic costs (Browne et al., 2022). Currently, there is no consensus on the global average oil spill cost at the international level. It resulted from a significant regional variation in clean-up costs (normally 60% of the total cost), ranging from a minimum of $1,300 per tonne in the Middle East to a maximum of $33,300 per tonne in Asia (Browne et al., 2022).

Determining oil spill-related loss and compensation after an oil spill is challenging and requires modeling for more accurate estimation. For example, Afenyo et al. (2022) developed a multiperiod model for assessing the socioeconomic impact of an oil spill incident in the Canadian Arctic (Afenyo et al., 2023) developed a hybrid Bayesian loss function-based method to accurately assess ship oil spill-related loss.

In some cases, such as pollution, penalty or fines (Loughnane et al., 1995), ship operation disturbance (Walker, 2000), increasing insurance (Österman and Rose, 2015), disputes, and lawsuits (Ornitz, 2001) may occur and raise the costs. The following needs consideration:

• (1)

  Operation disturbance costs – Costs associated with operation disruption due to the accident.

• (2)

  Accident response operation – Costs associated with accident response operation.

• (3)

  Administration costs – Costs of administrative work after the accident.

• (4)

  Insurance costs – The increases in the future in insurance premiums for the crew and the vessel.

• (5)

  Penalties and fines – Costs of punishment imposed on the shipowner by the authorities for activities against regulations and law or due to consequences of the accident (e.g. oil pollution and chemical pollution).

• (6)

  Legal – Expenses for legal consultation and services for possible labor disputes and lawsuits with other parties (e.g. cargo owners, other shipowners).

• (7)

  Accident investigation costs – The value of the time and expenses spent by the shipowner on internal and external investigation activities.

• (8)

  Opportunity costs – The value of sales (orders) lost, the potential reduction of market shares, and the non-realized profit due to accident costs, which could be invested instead in a profitable activity (e.g. production, stock market, and saving) generating interests.

Thus, the classification of safety costs in Arctic shipping can be demonstrated in Figure 2, which comprises categories of SPC and MAC.

The sub-categories of safety costs for shipowners can be derived from Figure 2 and can be further broken down into cost items of prevention costs as well as (directly and indirectly related) activities following the accidents (see Tables 2 and 3). Those items and activities are identified from the examples found in the literature and should cover related expenditures as much as possible.

The next step is to assign specific pricing techniques and information sources to give each cost item and each activity a monetary value. Most of these costs for shipowners are direct costs in the form of resources spent (e.g. capital, labor and external services). Therefore, the pricing techniques include a review of purchase prices, the sum of expenses paid and the sum of wages paid for the time. Corresponding to these pricing techniques, direct information sources related to the resources spent can be found from the shipping company, including:

• (1)

  Capital/external services – Accounting records, purchase orders, receipts, invoices, contracts, compensation agreements, supplement pay, orders of fines, orders of penalties and insurance claims.

• (2)

  Labor – Wage, timecards and payroll records.

• (1)

  Obtaining data and information is difficult.

Collecting data from direct sources requires access to the datasets of various departments (e.g. accounting, procurement, ship management, health and safety and human resources), which makes it a challenging task.

• (2)

  Some cost items/activities rely on company or expert estimations.

For some items and activities, cost estimation must be conducted by internal or external experts, as special knowledge is needed. For example, additional capital investment in an ice-class ship (PC11) is deemed an additional cost for the vessel to obtain Ice Class notation or the cost of the ship conversion/upgrading to a higher Ice Class. For acquisition, this cost is considered the added cost of an ice-class ship compared with ships with the same basic parameters but without Ice Class notations, called the “Cost of Ice.” It can be expressed in Equations (2 and 3).

• Or

The costs can be amortized into the expected years of service, which requires the cost differential, costs of conversion/upgrading, and the years of service to be determined by experts from the company or external experts. Similar to PC11, other items and activities involving differential (e.g. PC33), depreciation (e.g. AC21), vessel charting rate (e.g. AC41) or time spent (PC53) may also need to be estimated by experts.

• (3)

  Direct information sources may not be available.

One of the major problems in occupational safety and health (OSH) economic assessment is that companies may not keep track of OSH costs (EU-OSHA, 2002), especially accident costs, which may not even be recorded by the company (Targoutzidis et al., 2014). If the direct information sources (e.g. insurance claims) for accident costs are not available, simplified approaches may serve as alternatives to obtain the cost values. Such approaches to obtain the cost values of the main marine accident consequences are given below:

• (1)

  Total loss – The best loss estimation for the total loss vessels is the ship’s second-hand market value (Giziakis, 1982; Knapp et al., 2011). A simplified approach to obtain this value is the second-hand price per tonne provided by ship brokering companies (e.g. Clarksons).

• (2)

  Ship damage – Average claim costs (excluding total loss) given by marine insurance statistics (e.g. Cefor ocean hull portfolio) are often summarized for vessel type and size (gross ton size group), accident type, vessel age or other parameters (Ellis et al., 2007). Thus, the average claim cost in the year per corresponding accident type can be used as the damage cost.

• (3)

  Cargo damage/loss – The best approach to estimate the value of cargo damage/loss is to use the average international yearly market price for the cargo lost (transported commodity) (Giziakis, 1982).

• (4)

  Death – The IMO FSA guidelines adopt a cost of averting a fatality (CAF) value of $3 m (IMO, 2018). Browne et al. (2022) adopted this CAF value as the cost of an Arctic maritime fatality in the study.

• (5)

  Injury – Estimators given by the corresponding government agencies can be used to estimate direct and indirect costs according to the injury types, such as the OSHA’s Safety Pays Individual Injury Estimator given by the US Department of Labor (OSHA, 2023). Another approach uses a value for each injury. For example, the IMO FSA recommends $42,000 per injury.

• (6)

  Environmental damage – (IMO, 2018) provides regression formulae derived from the consolidated oil spill database (all spills) are expressed in Equation (4), where $f(V)$ is the total spill cost (2009 US dollars), and $V$ is spill size in tonnes.

To facilitate the evaluation of safety costs for shipping companies operating in the Arctic region, estimation checklists for prevention costs and MAC (see Tables 2 and 3) have been developed to combine the categories identified through the literature.

The classification and the estimation approach provide a guide for decision-makers to assess safety costs in Arctic shipping. The application requires the computation and analysis of a case study. The case study allows testing the tool with actual data to confirm the applicability and effectiveness of the classification system and estimation approach.

The study uses data from a US Arctic ship operator to confirm the study. Indeed, the data represent real figures provided by the company to test the usefulness of the tool. SPC and MAC were calculated and presented in Table 4, based on historical data (2011–2019) taken from the ship operator.

As seen from Table 4 and Figure 3, from 2011 to 2019, PC1 (facilities) and PC2 (safety measures) were the largest portions of the prevention costs, and their numbers are close. PC3 was relatively small, while PC5 was the least. Apparently, no investments in safety-related new technologies (PC4) were made during this period.

Table 4 also reflects that the prevention costs spent by the ship operator have strong effectiveness for AC3 (environmental damage) and AC4 (other costs), which are kept as null from 2011 to 2018. Figure 4 presents the trends of accident costs, which shows that prevention costs are also effective for AC1, AC2 and MAC. These costs gradually decreased during this period, with the highest numbers (e.g. AC2 = $5.4 m) in 2012 and the lowest (e.g. AC2 = $62,000) in 2017. Figure 5 indicates that the ship operator maintained investments in safety prevention at a stable level, and the shipping safety cost (SSC) slightly decreased as the MAC declined significantly.

The accident data were closely examined due to the high accident costs in 2012 and 2015. In 2012, there were three cases of colliding with docks and nine crew injury cases, in which the damage costs and injury costs were $4.6 m and $1.1 m, respectively. In 2015, one vessel hit a submerged object and damaged the propeller blades, which resulted in a $2.6 m damage cost. Besides, three cases of colliding with docks occurred in the year, and the damage costs were $327,680.

Although the data indicates that safety had improved during this period, the accidents involving damages to docks and crew injuries that occurred every year reflect a lack of safety training and training in docking operations. It can be recommended that the ship operator should invest in training for crew members in safety and docking operations and increase the hiring costs and compensation benefits for the crew to retain qualified and experienced crew members, given the fact that those costs belong to PC3 and were relatively smaller in the past than other investments.

Traditional accounting methods are not well-established in safety management (Tappura et al., 2015), and safety costs are rarely registered separately, blurring their identification in the books (Rikhardsson, 2004).

By classifying safety costs in Arctic shipping into nine groups and 45 sub-categories, this paper provides the very possibility to develop such registration. The stakeholders could obtain a transparent vision of the expenses directed toward safety and under which category. This counting would allow us to assess the situation. Additionally, the analysis of casualties and integration of accident costs allow us to effectively verify the relevance of safety investment efforts in terms of volume and focus.

Cost estimation approaches require a combination of realistically informed estimates and access to information from various departments (e.g. procurement, ship management, health and safety and human resources). Simple checklists designed to be used by safety departments that are not always conversant with accounting would facilitate verification of investments and discussion on budget with the top management. Indeed, safety remains seen as a regulatory burden in shipping. Therefore, being not only able to demonstrate the benefits to decision-makers would certainly enhance safety but also cost allocation.

Confined to the financial losses in this study, accident costs are wider, and those difficult to quantify in monetary terms are beyond the scope of this paper, such as damages to the shipping company’s reputation (Ellis et al., 2007), relationships with cargo owners (Österman and Rose, 2015), the stock prices (Tappura et al., 2015), morale (Ornitz, 2001) and job satisfaction of crew members (Österman et al., 2010).

The classification and estimation support the establishment of a method enabling the comparison of investment costs and accident expenses. Equipped with such a tool, shipping companies can monitor their safety engagement, verify investment effectiveness category-by-category and reorient budgets whenever necessary. The tool can become a decision-support system to best allocate safety investments.

This paper explores the topic, and future case studies must be examined to assess the practicability of the method and its relevance over time. Furthermore, the classification and estimation proposed can also be integrated into other economic methods, such as cost-benefits analysis, to compare SPC and safety benefits (e.g. reduced MAC) in the economic assessments.

Finally, other maritime stakeholders, such as the coastal states, may deploy such a method to monitor and orientate their investments.