EUROPEAN COMMISSION

Brussels, 1.6.2023

SWD(2023) 145 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the document

Proposal for a Directive of the European Parliament and of the Council

amending Directive 2009/18/EC establishing the fundamental principles governing the investigation of accidents in the maritime transport sector

{COM(2023) 270 final} - {SEC(2023) 201 final} - {SWD(2023) 146 final}

Table of Contents

1.Introduction: Political and legal context

2.Problem definition

2.1.What are the problems?

2.2.What are the problem drivers?

2.3.How likely are the problems to persist?

3.Why should the EU act?

3.1.Legal basis

3.2.Subsidiarity: Necessity of EU action

3.3.Subsidiarity: Added value of EU action

4.Objectives: What is to be achieved?

4.1.General objectives

4.2.Specific objectives

5.What are the available policy options?

5.1.What is the baseline from which options are assessed?

5.2.Policy measures and policy options

5.2.1.    Discarded policy measures    

5.2.2.    Retained policy measures    

5.3.Description of the policy options

5.3.1.    Policy option A    

5.3.2.    Policy option B    

5.3.3.    Policy option C    

6.WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?

6.1.Economic impacts

6.1.1.    Impacts on Accident Investigation Bodies    

6.1.2.    Impacts on the ship operators    

6.1.3.    Impact on EMSA    

6.1.4.    Impacts on SMEs    

6.1.5.    Functioning of the internal market and competition    

6.1.6.    Impacts on competitiveness    

6.2.Social impacts

6.2.1.    Maritime safety    

6.2.2.    Impacts on working conditions and skills    

6.2.3.    Impacts on fundamental rights    

6.3.Environmental impacts

7.HOW DO THE OPTIONS COMPARE?

7.1.Effectiveness

7.2.Efficiency

7.3.Coherence

7.4.Subsidiarity and proportionality

8.Preferred policy option

8.1.Identification of the preferred policy option and stakeholders views

8.2.REFIT

8.3.Application of the ‘one in, one out’ approach

9.How will actual impacts be monitored and evaluated?

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

2.Organisation and timing

3.Consultation of the RSB

4.Evidence, sources and quality

Annex 2: Stakeholder consultation (Synopsis report)

Annex 3: Who is affected and how?

1.Practical implications of the initiative

2.Summary of costs and benefits

3.Relevant sustainable development goals

Annex 4: Analytical methods

Annex 5: EU/EEA accident investigation bodies falling within the scope of Directive 2009/18/EC by type and year of establishment

Annex 6: IMO superseded or out of date references

Annex 7: Discarded problem drivers and policy measures

Annex 8: Retained policy measures

Annex 9: Effectiveness of the different policy options

Glossary

1.Introduction: Political and legal context

This Impact Assessment accompanies a legislative proposal for a revision of Directive 2009/18/EC establishing the fundamental principles governing the investigation of accidents in the maritime transport sector (hereinafter “the AI Directive” or “the Directive”) 1 .

The AI Directive establishes the fundamental principles governing the investigation of accidents in the maritime transport sector. Under the Directive, maritime accidents falling within the scope of the Directive are investigated by independent national accident investigation bodies determine the cause of the accident so as to improve maritime safety, as well to protect the marine environment, by learning lessons from accidents to prevent their reoccurrence. These inquiries are called safety investigations.

Maritime transport is a key sector for the EU economy as it embodies the main transport mode for European imports and exports to the rest of the world. Maritime transport is estimated 2 to represent around 80% of worldwide goods transported and around 30% of intra-EU transport activity. In 2019, 1.9 billion tonnes were transported by short sea shipping 3 to/from the main EU ports. In addition, 418 million passengers aboard ferries and cruise vessels embarked and disembarked in EU ports in 2019. 

At the same time, an average of 2,239 marine accidents were reported on average per year between 2014 and 2020 for EU Member States, of which 33% were categorised as very serious or serious casualties. In addition, 370 cases of marine pollution have been reported in total during 2014-2020 4 .   

The revision of the AI Directive has to be seen in the context of the Commission’s Communication on a Sustainable and Smart Mobility Strategy (SSMS) which sets out the EU vision for the transport system of the future. The strategy announced that the Commission is planning to initiate a major review of existing legislation on flag State responsibilities, port State control and accident investigation in 2021. According to the SSMS, the overall objective of this review should be to enable safe, secure and efficient maritime transport and further stresses that “safety and security of the transport system is paramount and should never be compromised and the EU should remain a world leader in this field. Continuous efforts with international, national and local authorities, stakeholders, and citizens is key […] 5 .” The strategy sets as one of the goals that by 2050, the death toll for all modes of transport in the EU to be close to zero. The objective of the EU and its Member States to ensure a high and uniform level of maritime safety and environmental protection has also been underlined in several Council conclusions and in particular those from 2017 6  and 2020 7 .

In the context of the AI Directive, maritime accidents are investigated to improve maritime safety, as well to protect the marine environment, by learning lessons from accidents to prevent their reoccurrence. Safety investigations of this type are separate from and independent of administrative and/or criminal investigations and do not seek to determine or assign any civil or criminal liability 8 .

In relation to the protection of the environment from pollution caused by accidents, the initiative should deliver the zero pollution ambitions of the European Green Deal (EGD) 9 and the SSMS. The initiative contributes towards Sustainable Development Goal (SDG) 3 (“Ensure healthy lives and promote well-being for all at all ages”) and SDG 14 (“Conserve and sustainably use the oceans, seas and marine resources for sustainable development”).

International context

Due to its history and international nature, maritime transport has developed a specific regulatory structure. At the global level maritime safety and protection of the environment are promoted through an international legal framework adopted under the auspices of the United Nations’ specialised agency responsible for regulating shipping, the International Maritime Organisation (IMO) 10 .

The obligation to investigate marine casualties or navigation incidents finds its origin in the United Nations Convention on the Law of the Sea 11 (UNCLOS) which imposes an obligation on the flag State 12 to conduct an investigation. However, the scope of this obligation is limited to a small set of requirements: (i) accidents happening on the high seas 13 , (ii) accidents which result in loss of life or serious injury to nationals of another State and/or (iii) accidents involving a ship or installation of another State 14 .

The IMO, which plays a role in the implementation of UNCLOS, subsequently developed its Casualty Investigation Code 15 . The Code requires a safety investigation 16 to be conducted into casualties involving the total loss of the ship or a death or severe damage to the environment. It also recommends that investigations be carried out into other marine casualties and incidents, by the flag State of a ship involved, if it is considered likely that it would provide information that could be used to prevent future accidents. In addition, the IMO is progressing work under the International Convention for the Prevention of Pollution from Ships (MARPOL) including the Action Plan to address marine plastic litter from ships, including as a result of an accident.

EU context

EU action in the field of maritime safety both complements and implements the international framework as defined within IMO. The transposition of IMO rules into the EU legal system makes these provisions actionable before the European Court of Justice thereby ensuring their uniform enforcement across the Union.

The bulk of today’s EU maritime safety policy was developed in the early 1990s, and was further worked upon between 2000-2009, in the wake of major maritime accidents causing substantial oil spills 17 and loss of life 18 . The overall objective of the EU has therefore been – and continues to be – a Union policy on safe seas 19 , and continuous work to improve safety and thereby reduce accidents and pollution into the sea.

Although the colloquial term accident is used to describe these unplanned, undesired events, both the IMO Code and the AI Directive refer to the more precise legal terms “marine casualties” 20 and “incidents” 21 .  

The AI Directive goes further than the IMO Casualty Code. First, it obliges Member States to establish independent accident investigation bodies (“AIBs”) and to provide in their legal order that these AIBs are notified of marine casualties and incidents 22 . The AIBs should investigate accidents depending upon their severity. For “very serious marine casualties” 23  (VSMC) the Directive, as does the IMO Casualty Code, provide that there must always be an investigation, so there is no discretion as to whether this accident type has to be investigated. The Directive also extends this obligation in fishing vessels of 15 metres in length and over.

For “serious casualties” 24 a preliminary assessment in order to decide whether or not to undertake a safety investigation should take place. For less serious casualties and incidents, it is for the AIB to decide to investigate when it considers that lessons to improve safety at sea can be learnt. Independence in the organisation, legal structure and decision-making means that the AIB can investigate and address safety recommendations without any appearance of bias or conflict of interest. Second, the types of vessels covered by the AI Directive is also different. The IMO rules mainly apply to commercial passenger or cargo craft while the AI Directive, is broader including fishing vessels of 15 metres in length and over and also includes pleasure yachts and pleasure crafts which can carry more than 12 passengers for commercial purposes. 

Procedures on how to conduct investigations in the EU are laid down in a so-called “Common Methodology”. 25  AIBs should publish reports on the basis of the investigations carried out and they should also notify the Commission of marine casualties and incidents via a database (the European Marine Casualty Information Platform – EMCIP), established and maintained by the European Maritime Safety Agency 26 (EMSA) for this purpose. The Directive allows for the collection, collation and sharing of safety data through EMCIP. This facilitates national administrations in responding to accidents, allowing for digitally sharing information both regarding the circumstances of occurrences and the safety reports resulting from the investigations. The use of digital tools optimises processes and means that the Directive is digital-ready. The carrying out of harmonised investigations and the electronic sharing of information and lessons learnt allows for the development of more targeted, evidence-based policy responses at international, EU and Member State level to accidents and incidents, and thereby it is part of continuous improvement. 

Safety investigation reports set out the facts of the incident outlining the particulars of the ship and its voyage, its crew, the circumstances of the casualty and the sequence of events setting out the consequences in terms of impact on persons, vessels and the environment. The investigation must analyse the key events in the build up to the casualty and its aftermath to discover contributing factors and ultimately safety issues. The investigation then identifies findings, contributing factors and safety issues so as to be able to make safety recommendations. The Directive provides that Member States shall ensure that safety recommendations made by the investigative bodies are duly taken into account by the addressees and, where appropriate, are given an adequate follow-up in accordance with Union and international law. Furthermore, where appropriate, an investigative body or the Commission shall make safety recommendations on the basis of results of multiple investigations (so-called “abstract data analysis” according to Article 15(2) of the AI Directive) and of the overall results of safety investigations carried out.

The safety recommendations are addressed to ship owners/operators, to national authorities, to the Union or to international bodies such as the IMO. While the safety recommendations are recommendations and are therefore not legally enforceable, the status of a recommendation made by an independent AIB and addressed to ship operators, to national authorities or for example to the IMO means that these recommendations have considerable persuasive authority based as they are on a safety investigation. EMSA collates accident data and publishes an Annual Overview 27 and a Maritime Transport Environmental Report 28 .

Since June 2011 when the Directive entered into effect some 2000 individual Safety Recommendations have been issued following safety investigations. 67% of them have been accepted, meaning that actions have been taken to improve safety of ships and pollution prevention. These are usually focused on a specific accident, as under the Directive and the IMO Casualty Code all very serious have to be investigated but there can be some repetitions (as accidents can inevitably resemble each other) but the recommendations and addressees remain different.

It is important to note that improvements in safety and the measurable perception thereof takes time. Changes to legislation or regulation are slow and the reason for this as there is a need to provide predictability and legitimate expectations, in view of the implementation and application. When changes to international conventions are involved it typically takes time for the international community (and in maritime this involves global action) and industry to recognise that there is a problem and to identify the root cause, to agree on the measures to be taken to address the problem; and for improvements to be observed in terms of reduced number of accidents. Changes involving the so-called “human element” where the safety recommendations are addressed to individual companies, to sectors or to administrations and which involve changes to practices, procedures, manning or training can however be quicker. This all forms part of continuous improvement in maritime safety.

Safety recommendations can result in the better enforcement of already existing IMO legislation. An example relates to the grounding of Hoegh Osaka which led to better and enhanced enforcement of the IMO Resolution A.582(14) related to the maximum securing load of lashings to be used when securing road vehicles on cargo decks. They can also lead to the improvement of EU legislation: the safety investigation into the fire on the MSC Flaminia concluded that the EU implementation related to Places of Refuge should be improved by virtue of further guidelines to better assist EU Member States. The same investigation also led to better implementation of IMO Resolution A.950(23) at the EU Member State level regarding the establishment and operation of maritime assistance services.

The conclusions of several safety investigations, as these relate to similar or repeated accidents, can also bring to light safety issues that might be individually disregarded, for example, the highest number of fire on-board ships is in the Engine Room. Most of these do not result in serious damages and unless they are catastrophic would not be regarded as very serious marine casualties (VSMCs) but such fires represent a regular threat to the ship. In that sense and drawing on accident investigations reports the International Association of Classification Societies 29 revised its unified interpretation related to hot surfaces in engine room in order to reduce the number of fires, whatever the ship type.

The Directive also provides that the Member States shall, in close cooperation with the Commission, establish a permanent cooperation framework (PCF) 30 to enable the AIBs to cooperate among themselves to the extent necessary to attain the objectives of the Directive.

Synergies with other EU policy instruments

The responsibility for maritime safety involves three overlapping lines of State intervention both internationally and at the EU level (see Figure 1). States have differing but complementary responsibilities either as a vessel’s state of registration, a state which is being visited by a foreign flagged vessel or as a coastal state by which a vessel is travelling without calling. The first “line of defence” is provided by the flag State. However, as flag State rules only applies to vessels that fly that flag, many of the IMO's most important technical conventions contain provisions for ships to be inspected when the vessels visit foreign ports to ensure that they meet the international requirements. This control by port States is regarded as the second line of defence.

Although both the flag State control and the port State control have worked to improve maritime safety and improve the marine environment, accidents can still occur. Once this happens, it is important to investigate what went wrong and how a similar accident can be avoided in the future. To achieve this the third line of defence of accident investigation was created.

This impact assessment has been initiated in parallel to the related impact assessments of the Port State control Directive (Directive 2009/16/EC 31 ) and the Flag State Directive (Directive 2009/21/EC 32 ). All three EU maritime safety Directives are based on the rules and standards established by the IMO at the international level and while they each reflect the differing responsibilities of the EU Member States in their various roles as flag, port and coastal states they have to be coherent with each other and any proposed change to one has to take the other two Directives and the broader international regulatory environment into account.

The Directives work together to contribute to a higher level of maritime safety and maritime transport efficiency as well as a stronger level playing field between Member States. The counter factual situation would be international obligations transposed into national legislation with no means for supervision at Union level or to cooperate with the support of EMSA. This would lead to less protection for EU citizens.

Figure 1: The flag State, port State and accident investigation responsibilities of EU Member States

The impact assessment of the port State control Directive (PSC) looks at extending the scope of port State control to fishing vessels, which are currently not covered by PSC. It also seeks to update the Directive to take account of changes to the international legal and regulatory framework since 2009. It seeks to address issues related to the targeting of vessels for inspection, including the use of electronic data and certificates and seeking to make PSC more environmentally focused. It also looks at issue encountered by Member States’ authorities in their implementation of the PSC regime.

The impact assessment of the Flag State Directive looks at updating and aligning the EU acquis with international rules maintaining the IMO-Audit provision and follow-up, to clarify and strengthen the flag States to perform their obligations, including monitoring of EU Recognised Organisations (otherwise known as Classification Societies). It also looks at digitalising the Flag registers, to revising the flag State performance measurement and rewarding good quality.

These three Directives are also part of and have significant interaction with the larger maritime safety acquis which includes elements such as the EU vessel traffic monitoring and information system (SafeSeaNet) 33 , the EMSA founding Regulation 34 , the fishing vessel safety Directive 35 , the EU legilslation relating to Recognised Organisations 36 and other EU environmental legislation 37 .  

The interaction of the AI Directive with the EMSA oil pollution response capacity, is however limited as the AI Directive is designed to protect the marine environment by preventing accidents by finding out their root causes while the EMSA oil pollution response is a damage clean up and mitigation measure. EMSA had a role in oil pollution response since 2006 and this was expanded to cover fixed spills from oil and gas installations in 2013.

In this regard, EMSA provides different services to ensure that EU coastal states respond quickly and effectively to marine pollution from vessels, oil and gas facilities. This includes a network of stand-by oil spill response vessels, a satellite-based oil spill monitoring service known as “CleanSeaNet”.

In addition to the oil spill response vessels EMSA has pre-positioned oil spill response “Dispersant Stockpiles” in the Member States for fast response and this combined with the “Dispersant Usage Evaluation Tool ” an e-dimensional numeric model to simulate oil spills with dispersant applications, which provides a quantitative comparison of response options according different levels of effectiveness depending on the dispersant and timing of its application can mitigate the damage cause by oil spills.

EMSA also provides other services to the Member States within the scope of emergency response to chemical spill into the sea, namely, (MAR-ICE) a network of chemical experts to strengthen information transfer on chemical substances involved in maritime pollution emergencies in EU waters and the “Marine Intervention in Chemical Emergencies Network (MAR-CIS) “Marine Chemical Information Sheets” which provide in a concise way substance-specific and maritime relevant information on chemicals aiming to assist the competent authorities during the initial stage of the response to maritime incidents involving such substances.

The linkage between the AI Directive and the Marine Strategy Framework Directive (MSFD) 38 is also limited as the latter provides for a framework within which Member States must take the necessary measures to achieve or maintain good environmental status in the marine environment by the year 2020 at the latest. By virtue of the MSFD, EU Member States are required to implement strategies aimed at (i) the protection and preservation of the marine environment, the prevention of its deterioration or, where practicable, the restoration of marine ecosystems in areas where they have been adversely affected; (ii) the prevention and reduction of inputs in the marine environment, in order to ensure that there are no significant impacts on or risks to marine biodiversity, marine ecosystems, human health or legitimate uses of the sea. This means that the MSFD is proactive in that it obliges Member States to have a plan in place to achieve good marine environmental status. Therefore and while the AI Directive has a role in environmental protection by seeking to prevent pollution and environmental damage by preventing accidents, it does not put any responsibilities on the MS to address marine damage or mitigate such damage nor does it require that damaged environments are restored. The AI Directive does not have the same overarching environmental objectives as the Marine Strategy Framework, the two pieces of legislation are parallel but separate in their scope of mechanism of action.

Linkage with the Green Deal

The Green Deal and more specifically the FuelEU maritime initiative will result in significantly higher uptake of renewable and low-carbon fuels and low and zero-emission propulsion systems over time. The change of type of fuels used can affect the reliability and performance of engines, calibrated for certain types of fuels, and the issue with this is the risk of loss of propulsion which is an obvious safety risk. Another risk relates to fire and explosion. At this stage however, in the context of no significant uptake of renewable and low-carbon fuels and low and zero-emission propulsion systems in the maritime sector it is difficult to predict what type of accidents may likely occur in the future. Nevertheless, a higher uptake of electric vessels in the long term is expected to result in lower pollution due to fuel lost.

Ex-post evaluation of the Directive and the Maritime Transport Fitness Check

The 2018 REFIT ex-post evaluation 39  and Maritime Transport Fitness Check 40  concluded that in the absence of the AI Directive, it is likely that the 16 AIBs established since 2009 would not have been put in place. The majority of the States without an AIB prior to 2009 primarily conducted investigations for criminal prosecution purposes if they did so at all. The creation of AIBs has given a boost to accident investigations for safety reasons, with an emphasis on independence and the development of safety recommendations for accident prevention purposes. In addition, the evaluation concluded that the Directive provides a consistent framework for conducting maritime accident investigations and thus ensures that accident investigations are conducted in a uniform and harmonised way throughout the EU.

While the IMO had developed its Investigation Code, the obligatory nature of the AI Directive has led to a harmonised reporting of accidents and incidents, as a standard set of requirements has to be met. The overall conclusion of the evaluation based on the assessment of relevance, effectiveness, efficiency, coherence and EU added-value was that AI Directive has largely met expectations, achieving EU-wide benefits.

However, shortcomings of the current policy framework with issues of resources, staffing and expertise widely reported as problematic. Resource constraints limit some Member State AIBs in discharging their responsibilities in a correct and timely manner. This is particularly the case for those Member States with small fleets and limited maritime transport activity where the establishment of a permanent dedicated AIB is seen as a disproportionate burden. In similar vein the Maritime Transport Fitness Check concluded that the capacity of Member States to fulfil their international obligations as a flag, port or coastal State in relation to the various Directives is under strain. Resources, staffing and expertise issues were widely reported as problematic for AIBs. The Fitness Check made a number of recommendations relevant to accident investigation including the better use of resources at national and EU level, including the pooling of resources building on the role and support provided by EMSA and exploiting the efficiency gains offered by digitalisation and information sharing. The results of the ex-post evaluation are reflected in this impact assessment as summarised in Table 1.

Table 1: Links between conclusions of the ex-post evaluation and the impact assessment

2.Problem definition

Significant accidents involving certain categories of vessels are currently not reported and investigated across the Union in a coordinated and harmonised manner and the necessary safety improvements to prevent loss of life and protect the marine environment are unlikely to be made. The investigation level also varies greatly between Member States. Moreover, some of the definitions necessary for the correct functioning of the Directive are either not provided or are not specific enough while others essential to keep the Directive aligned with the relevant international legal framework have become outdated. At present, many AIBs lack sufficient resources to be able to carry out the responsibilities of reporting on investigating marine casualties and incidents in a timely and effective manner. This impacts on the proper functioning of the Directive and the harmonised safety level it was designed to underpin. The underlying drivers, problems and implications that are relevant for this revision are presented in Figure 2.

Figure 2: Problem tree

2.1.What are the problems?

Problem 1 – Accidents involving some categories of vessels are not investigated or are not investigated in a harmonised manner 

During 2014-2020, an average of 2,239 marine casualties and incidents were reported on average per year showing a relatively stable evolution over time (see Figure 3). VSMCs represented 3.1% of the total while serious casualties an additional 30.2%, resulting in an average of 69 fatalities. Crew is the most affected category of victims, representing around 89% of fatalities. Furthermore, a total of 145 ships were lost over the 2014-2020 period (21 per year on average)  41 .

A total of 370 cases of marine pollution have been reported during 2014-2020, showing a decreasing trend over time. Marine pollution in the form of ship bunkers (fuel) and other pollutants (e.g. cargo residues, lubricating or hydraulic oils) represented 68% and 18%, respectively, of the total number of cases of pollution (Figure 4) 42 .

However, not all maritime accidents in European waters or involving European vessels with severe consequences such as loss of life or loss of the ship are currently within the scope of the Directive. Several ships types are excluded from the scope of the Directive, namely: a) ships of war and troop ships and other ships owned or operated by a Member State and used only on government non-commercial service; b) ships not propelled by mechanical means, wooden ships of primitive build, pleasure yachts and pleasure craft not engaged in trade, unless they are or will be crewed and carrying more than 12 passengers for commercial purposes; c) inland waterway vessels operating in inland waterways; d) fishing vessels with a length of less than 15 metres; e) fixed offshore drilling units. The reasons why these vessel types were excluded from the scope of the Directive are various 43 .

Figure 3: Number of reported marine casualties and incidents by type (left side) and number of fatalities (right side)

Source: European Maritime Safety Agency (2021)

Figure 4: Cases of marine pollution by type

Source: European Maritime Safety Agency (2021)

During 2014-2020, 810 investigations were finalised (116 investigations per year on average), 48% of these being related to VSMCs and 42% to serious casualties 44 . 

The need to further assess the inclusion of fishing vessels with a length of less than 15 metres in the scope of the Directive has been identified based on the feedback to the inception impact assessment and the targeted stakeholders’ consultations. These maritime accidents are neither reported nor investigated in a systematic or harmonised manner, despite the fact that they represent around 90% of the EU fishing fleet 45 . There is therefore a gap in safety and the possibilities of learning from these accidents and preventing such accidents in the future are not being followed up, with negative consequences on their crews and the environment 46 . No issues resulting from the exclusion from the scope of the Directive were identified by stakeholders for the other categories of vessels. 

Problem 2 – Some accidents are not investigated as AIBs lack clarity regarding the circumstances in which they must act and how they must act

The Directive contains a number of definitions which have given rise to questions of interpretation and to differences of approach between AIBs. This has also created a lack of legal certainty and an absence of harmonisation.

In 2015, PCF members were asked to list issues related to the implementation of the Directive and the lessons learned thus far. During the course of this exercise several Member States reported that a number of definitions (i.e. definition of length of fishing vessel, definition of pleasure yachts and pleasure craft, definition of ‘directly in connection with the operations of a ship’, definition of a deadline for fatal/non-fatal injuries, conditions for a simplified investigation report) should be further clarified as there are different interpretations between Member States and between Member States and EMSA, when entering data in the EMCIP database.

While not all AIBs agree that all of the definitions mentioned above are problematic, the lack of a common understanding about these topics can have an impact on the day-to-day activities of the AIBs in the case of a joint investigation, and may affect the quality of the reports submitted to EMCIP.

In addition, the AI Directive refers to IMO definitions as well as other EU law. Since the adoption of the Directive, the IMO legislation impacting on the Directive has changed (see Annex 6). As a result, the Directive does not always refer to the up-to-date legislation. This creates a lack of legal certainty and harmonisation. This prevents cooperation with AIBs in third countries (which follow the more up to date IMO provisions).

It has not proved possible to estimate the extent to which less clear, ambiguous or outdated definitions have impacted on the number of investigations that have been carried out. This is because there is no information available on the number of investigations that have not been carried out due to definitions being unclear or un-harmonised or which do not correspond to the IMO standard. That being said, the issue of problematic and outdated definitions has been extensively discussed within the PCF. While agreement on some was reached (for example accidents involving the operations of a ship – while in port) these agreed interpretations have not been incorporated into a legislative text and are therefore not mandatory or enforceable.

Problem 3 – Some accident investigation bodies are not able to effectively and consistently report and investigate marine casualties

This problem was highlighted in the 2018 Maritime Transport Fitness Check. There are multiple reasons for this: it can be an internal resource issue if the number of employed investigators is limited, the AIBs may also not have sufficient budget available to hire external experts, even when additional personnel or technical or scientific expertise are required. Resource constraints can also be the result of unexpected events, such as multiple accidents occurring during a short period, or for example during the COVID-19 pandemic when investigators possibilities to travel to the scene of accidents were restricted. The evidence collected from the stakeholders, particularly the AIBs, indicates that the severity of the problem differs between Member States with the shortage of human and/or financial resources being a bigger issue for some AIBs than others.

2.2.What are the problem drivers?

Problem Driver 1: The Directive does not apply to certain vessels 

This problem driver links to Problem 1. Fishing vessels in general show a relatively high incidence of serious casualties and VSMCs. For 2014-2020, data reported to EMCIP for fishing vessels of more than 15 metres in length show a share of VSMCs in total number of marine casualties of 6.8% and of serious casualties of 59%, well above the average for other ship types (2.6% and 24.3%, respectively). In terms of share in the number of VSMCs and serious casualties they come second after cargo ships, with 32% and 31% respectively (see Figure 5). In addition, fishing vessel crews constitute 34% of all fatalities 47 .

Figure 5: Distribution of severity per ship type for 2014-2020 (cumulative number of accidents – left side; share in the total number of accidents – right side)

Source: European Maritime Safety Agency (2022)

The occurrence indicators, calculated by EMSA as the ratio between the number of reported accidents for a given ship type and the corresponding fleet size, also shows that safety level related to fishing vessels has negatively evolved during the past years. While the overall average occurrence indicator has decreased over time, the one for fishing vessels has continuously increased between 2014 and 2020. In addition, fishing vessels represented 61% of the total number of ships lost during the 2014-2020 period (13 fishing vessels of the 21 total ships lost per year on average) 48 . In contrast to larger vessels, EU-wide data is not systematically collected on accidents involving fishing vessels smaller than 15 metres. The data reported to EMCIP relating to fishing vessels in this category either relates to information voluntarily reported by the AIBs or because the accident in question also involves a vessel falling within the scope of the Directive. 

Eleven AIBs stated that accidents involving fishing vessels below 15 metres in length are already investigated in their respective countries under national law. This activity generally only covers fatalities and/or serious accidents. Data concerning 2012-2019 from EMCIP validated through interviews with AIBs, shows that Denmark, France, Portugal, Sweden and Spain have uploaded accident reports concerning VSMCs with fishing vessels that are smaller than 15 meters (these Member States investigate these accidents based on national legislation even though there is no obligation to do so under EU law). Other AIBs (Germany, Greece, Malta, and Poland) have only uploaded accident reports for some years. While some other AIBs (e.g. Norway) have indicated that they also investigate these cases, but the reports are not uploaded in EMCIP.

As small fishing vessels are outside the scope of the Directive, EMSA does not follow up on this work and therefore it is not clear if these investigations are carried out in a systemic or harmonised manner. Furthermore, there is no broad or systematic dissemination of information on the results of these investigations – there is therefore no indication if the sporadic and uncoordinated non-harmonised investigation of these smaller fishing vessels accidents has a positive impact on safety in those Member States in which they are carried out. In addition, AIBs cannot benefit of lessons learned through investigations carried out and safety recommendations made in other Member States.

The Eurostat European Safety and Work data relating to accidents involving sea fisheries was also looked into. However, the Eurostat statistics only cover eight Member States and some of the Member States reporting to EMCIP do not appear in the Eurostat statistics, while some of the Member States who do report to EMCIP and appear in Eurostat statistics show conflicting numbers. Therefore, as these statistics cannot be reconciled, it has been decided to rely on the EMCIP data. The issue of fishing vessels safety has also been the subject of discussion by the Committee on Safe Seas and the Prevention of Pollution from Ships 49 where Member States agreed that there was a need to do more on this matter at EU level, including for smaller fishing vessels.

All stakeholders generally acknowledge the inherent risks involved in the operation of smaller fishing vessels. The AIBs interviewed noted that small fishing vessels are particularly prone to capsizing; falls overboard are also relatively common. It was also highlighted by multiple AIBs that often the operations of small fishing vessels are rather informal, with possibly only one (often elderly/retired) operator fishing alone, without an affiliation with a larger company. Consequently, there may be a lack of awareness, attention, or safety equipment at play, and in some cases, alcohol may be involved.

The exclusion of the mentioned category of vessels reduces the effectiveness of the Directive, as there is no obligation at the EU level for the AIBs to conduct a safety investigation. This in turn means that no harmonised EU-wide safety recommendations can be adopted to improve the safety of the EU fishing fleet. Consequently, the overall safety performance of smaller fishing vessels in the European fishing fleet cannot be improved on the basis of harmonised accident data. It appears however that the challenge with fishing vessels below 15 metres is that the incidents are often difficult to investigate particularly due to lack of witnesses, evidence (especially in the case of loss of one-person crew and/or vessel), and the causes of incidents are often difficult to address by recommendations. A balance is therefore needed between resources spent on and safety benefits expected from the investigation.

Problem Driver 2: Some definitions are not provided in the Directive or are not specific enough

This problem driver links to Problems 1 and 2. Three problematic definitions have been identified in this context.

Definition of length of fishing vessel. Marine casualties and incidents that involve only fishing vessels with a length of less than 15 metres are outside the scope of the Directive. There are a number of different methods by which the length of a vessel can be established with the two most common methods used globally being ‘length overall’ (LOA) 50 and ‘length between perpendiculars’ (LBP). This means that depending on the method used the same ship can have different lengths and could be within or outside the scope of the Directive. This ambiguity can create confusion and a non-harmonised approach between Member States regarding investigations involving this vessel type.

A relevant consideration in this regard is why the threshold of 15 metres for fishing vessels in the Directive was chosen. This was linked with a requirement in Directive 2002/59/EC as amended by Directive 2009/17/EC on vessel traffic monitoring which was adopted at the same time as the AI Directive. This amendment provides that any fishing vessel flying the flag of a Member State, or operating in EU waters, or landing its catch in the port of a Member State and with a length of 15 metres LOA had to be fitted with an Automatic Identification of Ships - Class A automatic tracking system. It is therefore clear that LOA was the intention of the co-legislators at the time the AI Directive was drafted.

Definition of ‘directly in connection with the operations of a ship’. For the purposes of the Directive, the terms “marine casualties” and “marine incidents” shall be understood in accordance with the IMO Casualty Code providing that these events have “occurred directly in connection with the operations of a ship”. There is significant divergence of approach as regards what constitutes the “operations of a ship” and these differences in approach can have a considerable operational impact, particularly as regards what could be a joint investigation and as regards the gathering of meaningful accident and incident data. The question whether the accident or incident is directly connected to the operation of a ship is pivotal for assessing if the occurrence should be considered, or not, under the scope of the Directive and therefore has to be investigated. This is a particular issue with regard to accidents on board ships in ports involving stevedores/dock workers.

The matter has been extensively discussed between Member States AIBs in the context of PCF and there is agreement that the issue of what constitutes the operations of a ship particularly when that ship is in a port needs to be clarified.

Definition of a deadline for fatal/non-fatal injuries. The IMO Casualty Code defines a VSMC as “involving the total loss of the ship or a death or severe damage to the environment”. This means a fatality will count as a VSMC triggering an obligation on the AIB to carry out an investigation. However, neither the relevant IMO nor the Directive provide any guidance on the time within which a death has to take place following the incident for it to qualify as a fatal incident.

A similar problem was encountered in the past in the aviation sector, but EU legislation 51 now provides, a clear time frame for a fatal injury. The absence of a unified single definition of fatal injury in the AI Directive may mean that maritime incidents which result in death are not being investigated as there is no harmonised approach to the time-frame within which death must occur for a fatality to be counted as a VSMC across the Union. The matter has been extensively discussed in the context of the PCF and there is agreement that a clarification is needed.

Problem Driver 3: References to outdated international and EU rules

This problem driver links to Problem 2. The Directive refers to a number of IMO legal texts which have been abrogated, amended or revised since the entry into force of the Directive (see Annex 6). As the Directive is built upon the IMO texts and as it is necessary that Member States fulfil both their international obligations as well as their EU obligations, it is important that the Directive be kept aligned with the relevant international instruments and that there is no contradiction in their obligations.

The Directive refers to the “IMO Code for the Investigation of Marine Casualties and Incidents 52 ” adopted in 1997. In May 2008 IMO adopted another text 53 to “incorporate and build upon” the 1997 IMO Code referred to in the Directive. However, in 2013 a further resolution was adopted 54 which revoked the 1997 IMO text. The Directive also refers to Circular MSC/MEPC. 3/Circ.3 55  of 2008 which provides the definition of a “serious casualty”; this IMO text was superseded by another IMO Circular in 2014 56 which no longer refers to “serious casualty”.  

If the Directive is not updated to take account of alignment with the international legal requirements, the Member States may face different obligations leading to confusion and extra effort for the AIBs, leading to an inefficient accident investigation process. Alignment between the EU acquis and the IMO regulatory environment is particularly important having regard to joint investigations carried out with third countries. This has not yet been done, as it requires that the Directive is updated. It is important to ensure that the Member States do not have contradictory obligations as regards the EU and IMO regulatory requirements.

Problem Driver 4 – Some AIBs lack sufficient resources and/or expertise to be able to effectively and consistently report on and investigate marine casualties

This problem driver links to Problem 3. A number of Member States already had an AIB before the adoption of the AI Directive and therefore the various AIBs have developed differently depending on their national context. The AI Directive is not prescriptive as to the form or structure the AIB had to take other than as to its independence. All 27 Member States now have AIBs with fifteen having established maritime-only AIBs, while twelve have multi-modal AIBs. Comparing the size and structure of different AIBs in different Member States is not useful as it ignores the fact that different Member States have different sized fleets, the flagged fleets are different in terms of composition, they vary as to the number and size of ports in their territory as well as their length of coastline. Moreover, some AIBs are multi-modal.

Accidents by their nature are not predictable and the issue of resource constraints becomes particularly pressing when there is a period with several VSMCs or serious marine casualties to investigate. Both the 2018 ex-post evaluation and the Maritime Transport Fitness Check showed that resource availability, either as regards available staff or with regard to external staff engaged to carry out a particular investigation can be critical considerations in deciding whether or not to perform an investigation into a less serious accident. Budget and personnel constraints might also influence the preliminary assessment in case of a serious accident. If an investigation is delayed due to insufficient resources, it may not be conducted in the most effective way, as due to time passed between the accident and the start of the investigation evidence might be lost, or the memories of the witnesses may be affected.

Seven of the 17 AIBs who were involved in the targeted consultation stated that they suffer from lack of resources, or that they would carry out more investigations if they had more staff available. The AIBs reported that in principle they have the option to subcontract experts, particularly when specific knowledge is required, but many noted that budget and/or administrative complications make this impractical or impossible. Only four AIBs explicitly stated that resource availability is not currently an issue for them, however one of the four indicated that the obligations to notify marine casualties and incidents to EMCIP place a significant additional stress on their resources as the AIB does not deal with statistical data otherwise.

Issues relating to the obligation of the AI Directive to report accident and incidents to the EMCIP database had been identified in the ex-post evaluation. Underreporting of casualties and incidents is a significant issue – the difficulty being that it is hard to estimate what is not known. Nevertheless, EMSA has carried out an estimation of the size of the underreporting issue comparing reported accident data with publicly available information. On the basis of its research, EMSA concluded that the underreporting rate may be as high as 50%. This analysis is based on work carried out by EMSA as well as on academic theories of industrial accident prevention 57  whereby there is a numerical relationship between the number of accidents resulting in serious injuries, minor injuries or no injuries with the number of each increasing as their severity decreases.  

On this basis the underreported accidents seem to relate more particularly to less serious incidents which are not always reported to AIBs by the shipping companies and therefore are not recorded in the EMCIP database. These are typically cuts, slips and trips usually involving crew members which do not result in lost days of work. Because of this, underreporting is not further considered in the assessment of impacts. At the same time, a number of AIBs see the reporting obligation as burdensome and a resource-intense obligation which could be better discharged by other competent authorities, allowing them to concentrate on investigation. The issue is also problematic because it means that there is no truly accurate picture of the accident situation.

Another aspect of this problem is the question of suitably trained staff with sufficient expertise in accident investigation and maritime transport. EMSA currently provides extensive training to AIB staff but it is the AIBs which have the responsibility for the recruitment and training of appropriately qualified investigators. The EGD and in particular the “Fit for 55” package will require boosting multimodal transport and ramping up the deployment of sustainable alternative transport fuels (for example through the “FuelEU maritime” initiative). This shift to renewable and low carbon fuels as well as to new technologies is expected to lead to a radical change in the maritime transport sector over the coming decades and will inevitably have an impact on the types of accidents that occur as well as impacting on the investigation thereof. This will present new challenges for the AIBs.

Other potential drivers considered in the impact assessment and discarded

A number of possible problem drivers (in relation to Member State acting as Substantially Interested State (SIS), the definition of pleasure yachts and pleasure craft and the COVID-19 pandemic) were considered during the impact assessment process but were discarded either because the identified problem driver was not validated by the stakeholders consulted, because the problem was not susceptible to a solution by means of EU legislation or because proposing an action to address the issue at EU level will not yield additional results. Further details on the discarded problem drivers and the reasons are set out in Annex 7.

2.3.How likely are the problems to persist?

Problem 1 - Accidents involving some categories of vessels are not investigated or are not investigated in a harmonised manner. As long as smaller fishing vessels are not included in the scope of the Directive, those Member States who do not currently investigate these accidents under national law will not do so. Those Member States that do investigate (under national law) smaller fishing vessels accidents do so in an uncoordinated way, and are not obliged to upload their reports and recommendations to EMCIP. Consequently, no lessons will be learnt with regards to safety on fishing vessels and no standardised EU wide data on this subject will be collected. There is no evidence to suggest that the situation will change in the future without EU level intervention. The lack of accident investigation reports, safety recommendations and lessons learnt could lead to a decline in safety in the fishing sector operating with ships below 15 meters in length. Without reminders and information about risks to avoid and safe practices to follow, safety standards may decline. The small size of the businesses in question would make it highly unlikely that they would promulgate such lessons within the industry themselves. As explained in section 2.2, the accident occurrence indicator 58 for fishing vessels has continuously increased between 2014 and 2020 showing that the safety level related to fishing vessels has negatively evolved during the past years 59 . 

Problem 2 – Some accidents are not investigated as AIBs lack clarity regarding the circumstances in which and how they must act. Without further EU level intervention, the unclear, ambiguous, or outdated definitions of the AI Directive will persist. This means that that AIBs will continue to lack clarity regarding the circumstances in which they must act (i.e. in relation to the length of fishing vessel within the scope, of the definition of ‘directly in connection with the operations of a ship’, and of the definition of a deadline for fatal/non-fatal injuries) and how they must act (i.e. in relation to conditions for a simplified investigation report). AIBs will be placed in the situation where they have to find ad hoc workable solutions not coordinated or harmonised across the EU.

Problem 3 – Some AIBs lack sufficient resources and/or expertise to be able to effectively and consistently report on and investigate marine casualties. As explained in section 2.2, some AIBs already lack sufficient human and financial resources and/or expertise to effectively and consistently report on and investigate marine casualties. Maritime traffic and ship calls are projected to increase over time. This in turn is expected to lead to an increase in the number of maritime accident investigations and put additional pressure on resources, despite the fact that improved technology may offer better means for investigators to investigate accidents. For example, most (larger) vessels that are built after 2002 60  are equipped with Voyage Data Recorders (VDR) 61 which should make the work of the investigators easier. On the other hand, the uptake of renewable and low carbon fuels in the maritime sector, in line with the “Fit for 55” package and in particular the “FuelEU maritime” initiative, will require resources to continuously update the staff skills. Problem 3 is thus expected to persist over time without further EU level intervention, although technological innovations are expected to facilitate to some extent the work of investigators. At the same time, an upgrade of staff skills is expected to be needed in the future.  

3.Why should the EU act?

3.1.Legal basis

Title VI (Articles 90-100) of the Treaty on the Functioning of the EU (TFEU) establishes the EU’s prerogative to make provisions for the Common Transport Policy. Pursuant to Article 100(2) TFEU, the Union legislator may lay down appropriate provisions for sea transport. Article 91(1)(c) of the TFEU provides that the Union has competence in the field of transport to lay down measures to improve transport safety.

Within this legal framework, the EU provides for a coordinated and harmonised safety standard, protecting life and the marine environment across the Union, instead of relying on the uncoordinated action of individual Member States only. Travellers, workers and citizens in general can be reassured that accidents will be investigated to the same high standard across the Union and that the lessons learned from such accidents can be disseminated, allowing for better evidence based policy responses at national and EU level.

3.2.Subsidiarity: Necessity of EU action

Under the principle of subsidiarity, in areas which do not fall within its exclusive competence, the Union shall act only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States. To the extent that international instruments in the field of Accidents Investigation in the maritime transport sector are an exclusive Union competence pursuant to Article 3(2) TFEU, the subsidiarity principle does not apply, either to those instruments or to Union rules implementing those agreements.

In any case, in the absence of the AI Directive, 16 of the Member States’ AIBs are unlikely to have been established. Furthermore, the majority of the States which did not have an AIB prior to 2009 used to conduct investigations primarily for criminal prosecution purposes. The Directive has therefore addressed a safety need.

In the absence of EU action, EU Member States would work in an uncoordinated and non-harmonised way. Given the international nature of maritime transport, there is therefore a “Union relevance” of improving the Directive to address the identified problems. It is essential that all Member States investigate similar accidents in the same way to ensure an appropriate safety net across the Union to protect life and the marine environment.

The identified problems apply across the entire Union and have the same underlying causes. Only the issue of inadequate resources to discharge the Directive’s obligations varies between Member States as some AIBs have sufficient human, financial and technical resources to be able to report or investigate all casualties that they would like to and others do not. Nevertheless, the fact that some AIBs have sufficient resources should not prevent those who do not from receiving the necessary assistance.

3.3.Subsidiarity: Added value of EU action

Shipping is an international sector, operating in different EU and international waters and regulated at the global as well as regional and national instances. Therefore, it has by nature a strong cross-border dimension. The 2018 ex-post evaluation of the AI Directive and the Maritime Fitness Check also underlined its EU added value in terms of its effectiveness, efficiency and synergies that it brings. The evaluation found that the obligation to create an independent AIB was regarded by the majority of stakeholders as the main added value of the Directive. This was done to avoid the situation that an AI body which is part of the maritime administration would in effect be investigating itself with the attendant risk of conflict of interest. Given the role and responsibilities of maritime administrations as regards certification of vessels, of equipment, of seafarer training as well as the control exercised over vessel traffic monitoring as well as of coast guards and other bodies, the EU co-legislators felt that it was important that the AIB was separated from the maritime administration. The Commission has been obliged to take a number of infringement actions in this regard 62 . The Directive also brings EU legislation in line with IMO regulation. EU action ensures a level playing field for shipowners, ports and between Member States. Harmonisation of investigations at EU level is essential in this respect.

Given the international nature of maritime transport and the need for a coordinated and harmonised approach to maritime safety, a multiplicity of Member State responses to the extent allowed, issues identified is not appropriate, therefore the issues previously identified are only susceptible to EU-wide solutions. Failure to adapt the Directive would remove the synergistic benefits gained through its implementation.

4.Objectives: What is to be achieved?

4.1.General objectives

In view of the problems identified in section 2.1, the review of the AI Directive should improve maritime safety and improve the protection of the marine environment. It should thus also contribute towards Sustainable Development Goal (SDG) 3 (“Ensure healthy lives and promote well-being for all at all ages”) and SDG 14 (“Conserve and sustainably use the oceans, seas and marine resources for sustainable development”).

4.2.Specific objectives

This initiative is designed to effectively address the identified problems that hamper the coordinated and harmonised investigation of maritime accidents in the European Union. The specific objectives (SOs) and their correspondence with the problem drivers are presented in Figure 6.

SO1: Protect fishing vessels, their crew and the environment. VSMCs of smaller fishing vessels of less than 15 meters in length including fatalities and loss of vessels are not investigated in a systematic and harmonised way across the European Union. Including VSMCs involving this vessel type within the Directive would require reporting and allow for more investigations and a better understanding of the safety problems that this segment faces. Systematic safety investigations of these accidents would lead to safety recommendations to the vessel operators, the sector, national administrations and other interested parties which should improve the safety picture. The conduct of systematic safety investigations will also provide policy makers at national and EU level with the information to prepare well-developed and evidence-based policy responses.

Figure 6: Correspondence between the specific objectives and the problem drivers

SO2: Member States accident investigation bodies to have clarity and precision regarding definitions to ensure that all accidents that need to be investigated are investigated in a timely and harmonised manner. AIBs need clarity on the circumstances in which they should carry out investigations. An absence of adequate harmonisation and a common understanding of definitions relating to the legislation means that similar accident are subject to different treatment as to whether they are investigated or not in different Member States. The Directive should also be aligned with the relevant international instruments so as to ensure better consistency between Member States obligations under the Directive and their international obligations – this is particularly an issue having regard to joint investigations with third countries.

SO3: Enhance the capacity of AIBs to conduct (and report on) accident investigations in a timely, expert and independent manner - including as regards renewable and low carbon fuels and technologies. Member States AIBs should be able to respond in a timely manner, report to EMCIP and to carry out appropriate safety investigations. Adequate resources are relevant in this context – if the AIBs need specialised training, operational or technical assistance there should be a mechanism within the Directive that provides for such a possibility. The uptake of renewable and low carbon fuels in the maritime sector, in line with the “Fit for 55” package and in particular the “FuelEU maritime” initiative, will require resources to continuously update the staff skills.

5.What are the available policy options?

5.1.What is the baseline from which options are assessed?

The EU Reference scenario 2020 (REF2020) is the starting point for the impact assessment of this initiative. The REF2020 takes into account the impacts of the COVID-19 pandemic that had a significant impact on the transport sector. More detailed information about the preparation process, assumptions and results are included in the Reference scenario publication 63 . Building on REF2020, the baseline scenario for this impact assessment has been designed to include the initiatives of the ‘Fit for 55’ package proposed by the Commission on 14 July 2021. A common baseline was developed for this impact assessment, as well as for the Flag State and Port State control impact assessments, to ensure consistency. More details on the baseline scenario are provided in Annex 4.

The baseline scenario assumes no further EU level intervention beyond the current AI Directive and the continuation of the application of the IMO Casualty Code as it is. Smaller fishing vessels would continue to be outside the scope of the Directive and therefore Member States would adopt different approaches to similar accidents. Member States with inadequate financial, technical or operational resources to be able to report on and investigate accidents in a timely and efficient manner would have to make choices as to what accident to investigate and how. As previously stated, the role of EMSA in the implementation of the Directive is central. EMSA operates the EMCIP database, provides training to investigators on the request of AIBs and assists in the management of the PCF. The Commission has launched an impact assessment on the possible review of EMSA founding Regulation. 64  However, the outcome of this impact assessment cannot be prejudged and thus the baseline scenario does not account for changes in the EMSA founding Regulation.

The COVID-19 pandemic had a major impact on global shipping, affecting all its segments from passenger ships to container ships and oil tankers. 65 In the baseline scenario, international maritime freight transport activity (intra and extra-EU) is projected to be 21% lower in 2020 relative to 2015. From 2021 onwards however it is projected to start recovering and grow strongly by 2025 and beyond (i.e. 19% growth for 2015-2030 and 48% for 2015-2050), due to the rising demand for primary resources and container shipping. Relative to 2019, this is equivalent to 8% increase in transport activity by 2030 and 33% growth by 2050.

The number of port calls for 2025-2050 is projected to grow at lower rate than transport activity, following similar evolution over the historical period 66 . This reflects the fact that transport activity is also driven by other factors such as the increase in the size of vessels over time, and of the distance travelled. In the baseline scenario the number of port calls is projected to go up by 14% by 2030 relative to 2015 and by 36% by 2050 (equivalent to 6% growth by 2030 relative to 2019 and 26% increase by 2050), following the recovery from the COVID-19 pandemic.

Driven by the increase in the transport activity and the number of vessels, the number of marine casualties is projected to increase over time in the baseline scenario. Considering only the scope of the AI Directive (i.e. excluding fishing vessels below 15 meters) the number of casualties is projected to increase by 14% by 2030 relative to 2019 and by 45% by 2050 without further EU level action. At the same time, the degree of severity of marine casualties is projected to decrease, leading to a decrease in the number of vessels lost (by 5% for 2019-2030 and for 2019-2050) and a relative stabilisation of the number of fatalities by 2050 (11% decrease for 2019-2030 and 3% increase for 2019-2050).

Accounting also for the fishing vessels below 15 meters, which are currently outside the scope of AI Directive, the number of marine casualties is projected to increase by 9% by 2030 relative to 2019 and by 27% by 2050 in the baseline scenario. The lower growth in the number of marine casualties for fishing vessels below 15 meters (6% for 2019-2030 and 15% for 2019-2050) is explained however by the projected reduction in the number of fishing vessels by 2050, while the marine casualties’ rate would still increase over time. The degree of severity of marine casualties including fishing vessels below 15 meters is projected to decrease over time but 210 fatalities are still projected in 2030 and 203 in 2050 in the baseline scenario, while 141 vessels are projected to be lost in 2030 and 129 in 2050. By 2050, the number of fatalities involving fishing vessels below 15 meters would represent 67% of the total number of fatalities while the vessels lost 85% of the total. In addition, the estimated number of fatalities and injuries where "other persons are involved" is projected to remain stable up to 2030 relative to 2019 but increase by 25% and 23%, respectively, up to 2050 (15 fatalities and 103 injuries). This is particularly relevant with regard to accidents on board ships in ports involving stevedores/dock workers. This is still far from the goal of the Sustainable and Smart Mobility Strategy of close to zero death toll for all modes of transport in the EU.

The tonnes of bunker fuel lost at sea due to VSMCs involving all vessels, excluding fishing vessels below 15 meters, is estimated to go up from around 650 tonnes in 2019 to 740 tonnes in 2030 and 890 tonnes in 2050. When also considering fishing vessels below 15 meters the bunker fuels lost at sea is estimated to increase from 1,178 tonnes in 2019 to 1,301 tonnes in 2030 and 1,497 tonnes in 2050 67 .

Driven by the projected number of marine casualties, the number of safety investigations in the scope of the AI Directive is projected to increase from 95 in 2019 to 111 in 2030 and 138 in 2050 in the baseline scenario. The baseline projections also account for the impact of Brexit 68 , which is likely to be felt more significantly by some Member States (e.g. Ireland, France, Belgium, Cyprus or the Netherlands) due to their links, either geographical or commercial, with the UK. The impact of the Russian invasion of Ukraine on maritime traffic, on maritime trade flows or safety and thereby indirectly on maritime accident investigation has not as yet been possible to quantify.

As explained, safety investigations for fishing vessels below 15 are not in the scope of the AI Directive and Member States that perform them do so in line with national legislation. In the baseline scenario, the number of safety investigations for VSMCs involving fishing vessels below 15 meters is projected to go up from a range between 53 and 156 investigations in 2019 to 58 to 171 investigations in 2030 and 63 to 186 investigations in 2050. These projections also take into account the effect of Brexit.

The costs for all EU AIBs for conducting investigations according to AI Directive (including administrative tasks) are projected to increase from a range between EUR 1.0 and 6.2 million in 2019 to EUR 1.2 to 7.2 million in 2030 and EUR 1.5 to 8.9 million in 2050. The investigations of VSMCs involving fishing vessels below 15 meters, performed by EU AIBs in line with the requirements of national law may add between EUR 0.6 to 10.1 million in 2019, to EUR 0.6 and 11.1 million in 2030 and to EUR 0.7 to 12.1 million in 2050.

5.2.Policy measures and policy options

As a first step, a comprehensive list of possible policy measures was established after extensive consultations with stakeholders, expert meetings, independent research and the Commission’s own analysis. This list was subsequently screened based on the likely effectiveness, efficiency and proportionality of the proposed measures in relation to the given objectives, as well as their legal, political and technical feasibility.

5.2.1.Discarded policy measures

A number of possible problem drivers and policy measures were considered during the impact assessment process but were discarded either because the identified problem driver was not validated by the stakeholders consulted, because the problem was not susceptible to a solution by means of EU legislation or because proposing an action to address the issue at EU level will not yield additional results. Further details on the discarded problem drivers and policy measures and the reasons are set out in Annex 7.

5.2.2.Retained policy measures 

The retained policy measures to address the problems identified in section 2 are provided in Table 2. A more detailed description of the policy measures is included in Annex 8.

5.3.Description of the policy options 

The retained policy measures have been grouped in 3 policy options: policy option A (PO A), policy option B (PO B) and policy option C (PO C). Table 2 presents the links of policy measures included in the policy options with the problem drivers and specific objectives.

Table 2: Overview of policy measures and policy options

All three policy options contain four common policy measures, namely: PM3 to clarify the length of fishing vessels, PM6 to align the Directive with the most up to date version of the IMO Casualty Code to avoid the situation where Member States have divergent responsibilities to the EU and IMO, PM9 on EMSA providing training on new developments and technologies as well as on the GDPR rules and PM10 requiring that EU AI bodies to have a quality management system.

5.3.1.Policy option A 

This policy option proposes a number of changes to the Directive to better deliver on the goal of improving maritime safety and pollution prevention. Policy option A will leave the scope of the Directive as it is and the AIBs would continue to have the flexibility they currently have. The changes envisaged would add little in terms of new requirements in the AI Directive while EMSA assistance would offer some benefits.

On Specific Objective 1 “Protect fishing vessels, their crew and the environment” - a non-regulatory measure is proposed where Member States will be recommended to report on VSMCs (crew fatalities or loss of vessel) and in so doing should provide more insight into the number of accidents as well as some of the reasons behind them (PM 1a).

With regard to Specific Objective 2 “Member States AIBs to have clarity and precision regarding definitions to ensure that all accidents that need to be are being investigated in a harmonised manner” the amendments are focused both on aligning the Directive with new references in IMO legislation (PM 6) and on providing more guidance on how the effectiveness of accident investigation can be improved to ultimately positively impact maritime safety. Problems with definitions will be clarified by legislation (PM 3) or by using non-regulatory measures such as interpretative guidelines (PM 4a and PM 5a).

In relation to Specific Objective 3 “Enhance the capacity of AIBs to conduct (and report on) accident investigations in a timely, expert and independent manner - including as regards renewable and low carbon fuels and technologies” the policy option requires a quality management system (QMS) be put in place to assure each AIBs’ effective management of resources and of inspection reports (PM 10). On the issue of resources, an enhanced role is provided for EMSA in support of the Member States by providing highly specialised analytical support (soft skills) during an investigation (PM 7), the supply of tools and equipment (hardware) to AIBs (PM 8) and the sharing of knowledge or the organisation of training on specific techniques/tools on issues such as renewable and low carbon fuels which are expected to have significant impacts on the way in which maritime transport operates and which will be relevant for accident investigation in the future as well as on issues relating to the General Data Protection Regulation “GDPR” (PM 9) and the sharing of information with third country authorities.

5.3.2.Policy option B 

This policy option seeks to balance enhanced harmonisation by means of more clearly stated definitions with a limited additional requirement as regards smaller fishing vessels. In this policy option, flexibility for the AIBs is limited while harmonisation is highlighted as much as possible.

On Specific Objective 1 “Protect fishing vessels, their crew and the environment” the Directive would be amended so that Member States would be required to report all fatalities and vessel losses involving fishing vessels of less than 15 metres to the EMCIP (PM 1b). This should provide the Commission and Member States with empirical data for future policy making in this area.

With regard to Specific Objective 2 “Member States AIBs to have clarity and precision regarding definitions to ensure that all accidents that need to be are being investigated in a harmonised manner”, the same alignment as in PO A on IMO (PM 6) is provided for.

On the possibly ambiguous terms and definitions which leave room for interpretation specifically how to measure the length of fishing vessels, how the phrase ‘directly in connection with the operations of a ship’ should be defined and the fixing of a time limit for situations when person does not die immediately after the accident are explicitly defined in the Directive to harmonise accident investigation further (PM3, PM 4b and PM 5b). Limiting the flexibility of AIBs as regards definitions will ensure a more harmonised approach to investigations and more coherent inspection results in terms of reports and safety recommendations.

In relation to Specific Objective 3 “Enhance the capacity of AIBs to conduct (and report on) accident investigations in a timely, expert and independent manner - including as regards renewable and low carbon fuels and technologies” there continues to be the requirement for a QMS (PM 10) like in PO A. The role of EMSA in support of the Member States is slightly enlarged beyond what is currently provided for in existing legislation by sharing of knowledge or the organisation of training on specific techniques/tools on issues such as renewable and low carbon fuels which are expected to have significant impacts on the way in which maritime transport operates and which will be relevant for accident investigation in the future as well as on issues relating to the General Data Protection Regulation “GDPR” (PM 9). However, the responsibility of AIBs to report accident information to EMCIP is reduced by broadening the scope of authorities who can input such data to include any competent authorities nominated by the Member States as well as the Commission and EMSA (PM 11). The EMCIP reporting burden on AIBs is somewhat lessened by broadening the range of competent authorities that can report to the database.

5.3.3.Policy option C 

This policy option revises the Directive, increasing the level of policy intervention and focuses on the harmonisation. This policy option is the most ambitious in terms of increased administrative and investigatory requirements for the AIBs.

With respect to Specific Objective 1 “Protect fishing vessels, their crew and the environment” this option will oblige AIBs not only to report on fatalities and vessels losses of smaller fishing vessels (PM 1b), but also to carry out at least a preliminary assessment to determine if lessons can be learned and if a full investigation needs to be carried out (PM 2). At the moment, AIBs are not required to provide information on the safety record if smaller fishing vessels or their susceptibility to accidents at EU level. However, smaller fishing vessels are generally older, not typically appropriately invested in or operated by their owners and are frequently crewed by only one or two crewmembers. Therefore, if they are involved in a fatal incident there may not be any crew to interview. This does not mean however that there will be no evidence and that the lessons learned cannot be shared. Moreover, as they fall outside the scope of the Directive, accidents are not reported and therefore, there are no statistics that allow monitoring the evolution of accidents (even if they are not investigated).

All stakeholders generally acknowledged the inherent risks involved in the operation of smaller fishing vessels. The AIBs interviewed noted that small fishing vessels are particularly prone to capsizing; falls overboard are also relatively common.

However, there is mixed opinion among AIBs as to whether the scope should be extended. AIBs already investigating casualties involving smaller fishing vessels are in favour of extending the scope. AIBs not yet investigating casualties involving smaller fishing vessels have mixed opinions. Some of them, especially the small ones, are against a scope extension as they fear that it will increase their workload substantially. The larger AIBs seem to be more in favour as they see advantages of investigating the accidents as in the long run it could improve safety. Social partners in the sea fisheries sector (employers and employees) on the other hand have expressed support for this measure, underlining the importance of gathering solid, detailed information on the causes of accidents.

At the same time, IMO has removed the reference to serious casualty and as part of alignment the Directive will be amended to abolish the requirement that a preliminary assessment is carried out in respect of all serious casualties. This means that apart from VSMCs (which currently make up 48% of European AIBs completed safety investigations and following the extension to smaller fishing vessels proposed in this revision whereby the AIB will only be obliged to carry out a preliminary assessment) all other safety investigations (currently just more than half of all investigations) will be at the discretion of the AI bodies. In other words, besides investigating VSMCs, AIBs already focus their work on the accidents where they consider that, a priori, some safety lessons can be learned.

With regard to Specific Objective 2 “Member States AIBs to have clarity and precision regarding definitions to ensure that all accidents that need to be are being investigated in a harmonised manner” the same alignment as in PO A and B on IMO (PM 6) is provided for. Potentially problematic definitions will be laid down in the Directive and clearly addressed. The effect of this policy option is that AIBs will still have some flexibility in some areas, while in other areas the room for interpretation is limited, to improve harmonisation in the EU (PM3, PM 4b and PM 5b).

On the resource issue set out in Specific Objective 3 “Enhance the capacity of AIBs to conduct (and report on) accident investigations in a timely, expert and independent manner - including as regards renewable and low carbon fuels and technologies” the role of EMSA in support of the Member States is enhanced by providing highly specialised analytical support (soft skills) during an investigation (PM 7), the supply of tools and equipment (hardware) to AIBs (PM 8) and the sharing of knowledge or the organisation of training on specific techniques/tools and new developments (in particular renewable and low carbon fuels following the “Fit for 55” package, automation etc.), as well as assistance on EU GDPR which will be relevant for accident investigation in the future (PM 9). EMSA’s assistance both in terms of soft skills and hardware will partially mitigate the additional workload. The authorities which can report data to EMCIP will also be expanded (PM 11). The requirement for a QMS (PM10) is also contained within this option.

6.WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?

This section summarizes the main expected economic, social and environmental impacts of each PO 69 . The proposed measures which involve the amendment of the Directive are assumed to be implemented from 2025 onwards, so that the assessment has been undertaken for the 2025-2050 period and refers to EU27. Costs and benefits are expressed as present value over the 2022-2050 period, using a 3% discount rate. Further details on the methodological approach are provided in Annex 4.

The revision aims at improving how accident investigations are undertaken in the EU. Some measures aim at increasing efficiency and harmonisation through better definitions, to share resources (i.e. through EMSA) and alignment with other pieces of legislation 70 . Better investigation is expected to ultimately lead to enhanced safety and better protecting the environment.

6.1.Economic impacts

This section provides the economic impacts of the policy options on the AIBs, ship operators and EMSA. It also provides an assessment of impacts on SMEs, the functioning of the internal market and competition, and on competitiveness.

6.1.1.Impacts on Accident Investigation Bodies

Impact on the number of investigations. The inclusion of the very serious casualties involving a fishing vessel below 15 meters in PO C is projected to result in 37 to 110 additional investigations relative to the baseline in 2030 and 40 to 119 additional investigations in 2050 (Table 3) at EU level. It should be noted that for fishing vessels below 15 meters, as explained in section 5.1, the investigations in the baseline scenario are conducted in line with the requirements of national legislation in 11 Member States.

The clarification of the length of fishing vessels definition, defined as Length overall (LOA) in all three policy options, is projected to result in a limited increase in the number of investigations that are currently under the scope of the AI Directive (6 additional investigations in all three options relative to the baseline in 2030 and 2050).

As regards the definition of “directly in connection with the operations of a ship” and how this is interpreted having regard to port workers, the impact of a clearer definition explicitly extending the scope of the Directive to situations where there are casualties involving port workers which take place on board a ship in a port is expected to give rise to a limited number of additional investigations per year in PO B and PO C (8 additional investigations relative to the baseline in 2030 and 9 in 2050). The impact would be smaller in PO A whereas only interpretative guidelines are foreseen (4 additional investigations relative to the baseline in 2030 and 5 in 2050).

A number of policy measures further seek to specify or clarify definitions in the Directive. It was not possible to quantify their impact on the additional number of investigations, although their impact is expected to be limited.

On the establishment of a time limit for fatal accidents, this is expected to give rise to a very limited increase in the number of investigations per year in PO B and PO C. The 16 AIBs that responded to the AIB survey indicated that accidents with “delayed” loss of life only occur once a year or less. In many cases, such accidents are investigated anyway. All the AIBs responding to the survey mentioned that it never happened that a fatal accident was not investigated because the death occurred only after the decision not to undertake the investigation. However, several AIBs mentioned that it would be helpful to have an alignment with international standards or guidelines, to facilitate cooperation with non-EU countries.

The alignment with the up to date IMO provisions in PO A, PO B and PO C means that the concept of a serious casualty is removed from the Directive along with the obligation on AIBs to carry out a preliminary assessment of such casualties to determine if a safety investigation has to be carried out. While this alignment with the IMO provision will not necessarily reduce the number of non-mandatory safety investigations carried out it will mean that AIBs are not obliged to carry out preliminary assessments which may lead to a reduction of the costs. Given that all occurrences are subject to some sort of assessment to establish whether it falls within the scope of the Directive and should be investigated, it is not possible to quantify the possible reduction in costs.

Overall, PO A would lead to 10 additional investigations at EU level in 2030 and 11 in 2050 relative to the baseline. PO B is projected to result in slightly higher number of additional investigations than PO A (14 in 2030 and 15 in 2050), while PO C would result in 51 to 124 additional investigations in 2030 and 55 to 134 additional investigations in 2050 (Table 3). As explained above, the number of additional investigations may be slightly higher, in particular in PO B and PO C, which additionally establish a time limit for fatal accidents.

Table 3: Projected number of investigations (EU27)

Source: Ecorys (2022), Impact assessment support study

Impact on the reporting requirements. All policy options include requirements regarding the reporting of very serious marine casualties (VSMC) involving fishing vessels below 15 meters. However, while in PO A Member States are encouraged to report VSMCs involving these vessels (on a voluntary basis), in PO B and PO C Member States are obliged to report them. In 2019, only Denmark, France, Malta, Portugal and Poland reported in total 21 VSMCs in EMCIP, which represent around 5% of the total number of VSMCs estimated for that year 71 . In the baseline scenario, the share of VSMCs involving fishing vessels below 15 meters that are reported in EMCIP is assumed to remain constant over time relative to 2019. In PO A the share is assumed to increase to 50% of the VSMCs involving fishing vessels below 15 meters, which however represents an upper bound and thus leads to conservative cost estimates. In PO B and PO C all the VSMCs would be reported in EMCIP, following the reporting obligation. Thus, PO A would result in 218 additional VSMCs reported in EMCIP in 2030 and 237 in 2050 relative to the baseline, while PO B and PO C would result in 436 additional VSMCs reported in EMCIP in 2030 and 474 in 2050 (Table 4).   

Table 4: Projected number of VSMCs reported in EMCIP

Source: Ecorys (2022), Impact assessment support study

Impacts on costs for AIBs. The increase in the costs for AIBs in the policy options relative to the baseline originate from: 1) an increase in the number of investigations; 2) an increase in the reporting requirements; and 3) from the requirement of establishing a quality management system (QMS) for the AIBs.

The costs related to the additional number of investigations (i.e. adjustment costs) are estimated on the basis of the time needed to complete an investigation. Drawing on stakeholders’ consultations, the time spent on an investigation is assumed to be between 250 and 1,500 hours 72 . It should be noted that PO C only requires a preliminary assessment of VSMCs involving fishing vessels below 15 meters, which may result in lower number of hours and lower workload for AIBs if a full investigation is not pursued. It is however not possible to estimate how many of the preliminary assessments would result in full investigations. For this reason, a conservative assumption is used, namely that all preliminary investigations result in full investigations, which represents the upper bound in terms of costs. PO A is projected to result in additional costs of between EUR 0.2 and 0.8 million in 2030 and EUR 0.3 to 0.8 million in 2050 relative to the baseline, while PO B would result in somewhat higher costs of EUR 0.3 to 1 million in 2030 and EUR 0.3 to 1.1 million in 2050. PO C shows the highest additional costs, estimated at EUR 0.7 to 8.2 million in 2030 relative to the baseline and at EUR 0.7 to 8.8 million in 2050. The much higher cost of PO C, relative to PO A and PO B, is related to the inclusion of the VSMCs involving fishing vessels below 15 meters in the scope of the AI Directive (Table 5).

The time for reporting is estimated by EMSA as being in the order of 15 minutes work per casualty. In any event, AIBs are notified on all casualties and have to sort through them to establish whether or not they will investigate. Any additional costs for AIBs will be small particularly as the possibility to upload data to EMCIP is broadened to allow this to be done by competent national authorities other than the AIBs in PO B and PO C. The additional costs related to reporting of VSMCs (i.e. the administrative costs) in PO A are estimated at EUR 1.6 thousand in 2030 relative to the baseline and EUR 1.8 thousand in 2050. For PO B and PO C the costs would be higher, due to the obligation of reporting all VSMCs involving fishing vessels below 15 meters (EUR 3.3 thousand in 2030 and EUR 3.6 thousand in 2050 relative to the baseline).

In addition, under all three POs Member States shall develop, implement and maintain a quality management system (QMS) for their AIBs 73 . This QMS shall be certified in accordance with the applicable international quality standards (ISO 9000 series). The QMS is required for all AIBs to identify and address system problems they may encounter and to thereby improve the quality of their work and in particular of their investigations. According to desk research in the context of the impact assessment support study, the one-off costs for putting in place such a QMS (ISO 9000) ranges between EUR 5,000 and EUR 15,000 per AIB, depending on the size and the system already in place. Annual auditing costs (i.e. ongoing costs) are assessed at EUR 5,000 per annum.

According to the impact assessment support study, only one AIB (i.e. in France) holds an ISO certified system while six other AIBs have indicated that they have a QMS in place, but not ISO certified. On that basis, no one-off costs are assumed for the MS with an ISO certified standard, EUR 5,000 one-off costs are assumed for each of the 6 MS that have an (uncertified) QMS in place and EUR 15,000 one-off costs for each of the 20 MS that have no QMS in place. Ongoing costs of EUR 5,000 per year are assumed for all MS that have an uncertified QMS in place or that do not have any QMS in place. Thus, total one-off costs for all EU MS are estimated at EUR 330,000 in 2025 while ongoing costs at EUR 130,000 per year from 2026 onwards.

The total costs for AIBs are provided in Table 5. In PO A, they are estimated at an additional EUR 0.2 to 0.8 million in 2030 relative to the baseline and EUR 0.3 to 0.8 million in 2050. PO B shows very slightly higher costs relative to PO A due to the reporting requirements for VSMCs and the requirements related to the definition of “directly in connection with the operations of a ship” in the Directive, estimated at an additional EUR 0.3 to 1 million in 2030 and EUR 0.3 to 1.1 million in 2050 relative to the baseline. Finally, PO C shows the highest additional costs for AIBs relative to the baseline due to the extension of the scope of the AI Directive to VSMCs involving fishing vessels below 15 meters (EUR 0.7 to 8.2 million in 2030 and EUR 0.7 to 8.8 million in 2050 relative to the baseline).

Table 5: Costs for AIBs (in thousand EUR, 2020 prices)

Source: Ecorys (2022), Impact assessment support study

EMSA operational assistance to AIBs, included in PO A and PO C would not have any cost implication for AIBs. Rather it should ease their resource burden and allow them to either do all of the activities they would like to but have been unable to do because of resource limitations, or allow them to do their work better or in a more efficient and timely manner.

6.1.2.Impacts on the ship operators

In relation to the reporting requirements for small fishing vessel VSMCs, included in PO A as a non-regulatory action and in PO B and PO C as an obligation, it should be noted that the operators of fishing vessels below 15 meters are already obliged under national law to notify the maritime authorities of a loss of the vessel or of a fatality on-board detailing the circumstances. Therefore, such measure would not add any administrative costs for ship operators relative to the baseline in any of the options. The administrative costs to be considered under the ‘One-in One-out’ are thus zero.  

During investigations, investigators have to interview vessel crew or need to temporarily immobilise a vessel (for example to obtain information from on-board equipment). If the number of investigations of AIBs increases, the frequency that production processes are temporarily stopped (due to unavailability of the vessel and/or its crew) increases, resulting in higher enforcement costs for ship operators relative to the baseline. The policy measures included in the policy options that have such an impact are those involving fishing vessels and clarification of definitions.

The nature of investigations differs however case-by-case. For example, some AIBs have indicated during the consultation that it is not uncommon that, in a VSMC involving a small fishing vessel, the vessel has sunk and there are no witnesses. In these situations, the associated costs of cooperating with the investigation are considered to be zero, as the vessel is no longer operational and the persons on board are deceased.

Overall, as explained in section 6.1.1 the inclusion of the length of fishing vessels in the Directive as Length overall (LOA) in all three policy options, the definition of “directly in connection with the operations of a ship” and the extension of the scope of the directive to VSMCs involving fishing vessels below 15 meters in PO C would lead to additional investigations relative to the baseline, the highest number being estimated for PO C. Other measures clarifying definitions in the AI Directive (as regards delayed fatalities) are expected to have an impact on all shipowners as the number of safety investigations is expected to increase but their impact on costs for shipowners is expected to be small.

For the assessment of additional enforcement costs, a ‘cooperation effort’ of 8 hours per investigation is assumed. Such an investigation will be triggered by either the sinking of the vessel or a fatality. If the vessel has sunk the issue of the vessel being tied up and not able to operate does not arise. In the case of a fatality it is likely that such a case would have to be investigated by other competent authorities such as the police or judicial authorities under national law so the vessel will be immobilised for some period in any event. While safety investigations are separate to and distinct from other types of investigations, it should be possible for AIBs to avail of this window of opportunity when the vessel is not working, thereby limiting the negative economic impact on the enterprise. In 2018 (the latest year for which complete information is available), information from the Commission’s Scientific, Technical and Economic Committee for Fisheries indicates that the (gross) value of landings was EUR 7,994 million in the fishing sector of EU27. The total number of hours worked by vessel crew is estimated at EUR 194 million in 2018. Thus, the gross value of landings per hour worked is estimated at EUR 41. The enforcement costs per investigation are estimated at EUR 328. The total enforcement costs for ship operators in PO A are estimated at EUR 3.3 thousand in 2030 and EUR 3.6 thousand in 2050 relative to the baseline, linked to the higher number of investigations due to the inclusion of the length of fishing vessels in the Directive as Length overall (LOA) and the interpretative guidelines on the definition of “directly in connection with the operations of a ship”. In PO B they are slightly higher than in PO A, estimated at EUR 4.6 thousand in 2030 relative to the baseline and EUR 4.9 thousand in 2050 due to the inclusion of the definition of “directly in connection with the operations of a ship” in the Directive. In PO C enforcement costs are the highest, due to the inclusion of the VSMCs involving fishing vessels below 15 meters in the scope of the Directive, and estimated at EUR 16.7 to 40.7 thousand in 2030 and EUR 18 to 44 thousand in 2050 relative to the baseline at EU level.

On the basis of information from EMCIP, it is estimated that 67% of completed investigations give rise to at least one safety recommendation or define actions to prevent similar accidents from occurring in the future 74 . For PO A and PO B the costs savings for ship operators in terms of avoided vessels lost are estimated at EUR 0.9 million relative to the baseline, expressed as present value over 2022-2050. For PO C the costs savings are estimated at EUR 5.7 to 9.4 million relative to the baseline (corresponding to 18 to 31 vessels lost avoided). It should be noted however that there is high uncertainty regarding these estimates 75 . This is because the impacts of AI Directive on safety are indirect, through the safety recommendations and their follow-up. For this reason, a sensitivity analysis has been performed, assuming 10% and 20% lower value in absolute terms of the elasticity used to derive the impacts. Table 6 shows that even with lower value of the elasticity PO A, PO B and PO C are projected to result in costs savings for ship operators in terms of avoided vessels lost, although the positive impacts on safety would be more limited.

Table 6: Results of the sensitivity analysis on the costs savings for ship operators in terms of avoided vessels lost, expressed as present value over 2022-2050 (in million EUR) relative to the baseline

6.1.3.Impact on EMSA

The changes to the Directive related to reporting of VSMCs for fishing vessels below 15 meters and definitions are not expected to give rise to additional costs for EMSA.

Cost increases would however result from measures that increase the role of EMSA to facilitate the investigation of casualties in PO A and PO C (i.e. provide highly specialised analytical support during an individual investigation, provide analytical tools and equipment during an individual investigation and share knowledge or organise training on specific techniques/tools and new developments) including GDPR in all policy options. Although many AIBs mentioned during the stakeholders’ consultation that they do not have sufficient resources to adequately conduct accident investigations, it often remains unclear what could be done on the EU level to address this issue. Many AIBs simply indicate that they would like to see their staff increased. As the AI Directive has no means to do so, policy measures are defined to increase assistance from EMSA in the conduct of investigations, while of course respecting the independence of AIBs. In practice, this would mean that any assistance provided by EMSA should be requested by the AIBs. The benefit of EMSA assistance would be that it could allow AIBs to investigate all the occurrences they would like to do so in a timely and efficient manner.

EMSA has considerable experience in providing operational support to national authorities in a number of areas. A pilot project involving remotely operated vehicles (remotely operated submersibles) began in 2020. Discussions with AIBs led to an expected need of 10 to 12 operations per year as well as the drawing up of business cases, draft rules of procedures, etc. With regard to the operational support from EMSA it is likely that the support will grow over time with 10-12 support actions in the first year growing to 30 actions per year over the successive years. The number of cases of support will also depend on demand by the AI bodies. This can be reviewed in function of demand.

The highly specialised analytical support during an individual investigation by EMSA in PO A and PO C is foreseen in terms of participation to VSMC investigations of ‘standard’ experts and ‘high level’ experts from EMSA. For ‘standard’ experts, 5 working days are assumed per VSMC at a rate of 1,500 EUR per day. ‘Standard’ experts would participate in around 10% of the VSMCs per year in PO A and PO C. For ‘high level’ experts, 4 working days are assumed per VSMC at a rate of 3,000 EUR per day. ‘High level’ experts would participate in around 2% of the VSMCs per year in PO A and PO C. In PO A the number of VSMCs are projected at 82 in 2030 and 97 in 2050. In PO C, due to the assumed extension of the AI Directive to VSMCs involving fishing vessels below 15 meters, the number of VSMCs are projected at 540 in 2030 and 595 in 2050. As a result, the additional costs for EMSA in PO A are estimated at around EUR 81 thousand in 2030 and EUR 96 thousand in 2050, equivalent to 0.7 full time equivalents (FTE) in 2030 and 0.8 FTEs in 2050 relative to the baseline.. In PO C the additional costs are estimated at around EUR 535 thousand in 2030 and EUR 589 thousand in 2050 equivalent to 4.5 full time equivalents (FTE) in 2030 and 5 full time equivalents in 2050 relative to the baseline. .

The analytical tools and equipment to be provided by EMSA during an individual investigation (hardware) can take various forms like for example:

·On scene tools such as ROV (remotely operator submersible), RPAS (aerial drones), Robots (e.g. small ones to access damaged areas, enclosed spaces, etc.); 

·Specialised tools such as: VDR facilities, software such as MADAS (Marine Accident Data Analysis Suite) 76 ; ad-hoc resources (scientific laboratories, tank testing); ad-hoc services (3D simulation for fire, loss of containers, etc.).

This support will build upon the existing services provided by EMSA in different areas (as seen above aerial drones, submersible drones, satellite imagery training etc.) and will be provided on a first come, first served basis. In the longer term, and depending on experience in implementation, the EMSA resources would be adapted accordingly.

The additional cost estimates per year from 2025 onwards for analytical tools and equipment to be provided by EMSA in PO A and PO C are provided in Annex 4. They are estimated at EUR 1.1 million per year relative to the baseline for all years (EUR 0.9 million for on scene tools and EUR 0.2 million for specialised tools).

In relation to knowledge sharing and organising training on specific techniques/tools and new developments, EMSA currently organises training courses on core skills for accident investigators, advance courses for accident investigators and on use of electronic evidence (VDR/ECDIS). Additional training could be foreseen in all policy options in particular in relation to renewable and low carbon fuels (risks, use on board, ship protections, etc.), in support of the initiatives part of the “Fit for 55” package, but also on: human element, conduct of an automated ship, safety analysis (to encourage Member States to better use casualty data), forensic (photos, sampling, etc.), occurrences (fire, navigational accidents, damage to equipment, occupational activities, etc.) and GDPR rules. In all policy options, 6 additional training courses are foreseen at around EUR 36,000 each (EUR 6,000 estimated for the trainer and EUR 30,000 on average for reimbursement of participants). Thus, the total costs in all policy options are estimated at EUR 216,000 per year starting from 2025.

Total additional costs for EMSA are estimated at around EUR 1.42 million in 2030 and EUR 1.44 million in 2050 in PO A, and at around EUR 1.88 million in 2030 and EUR 1.93 million in 2050 in PO C. In PO B only the training costs to knowledge sharing and organising training are foreseen, estimated at EUR 216,000 per year (Table 7). These additional costs expected for EMSA are not covered by existing budgetary commitments. Part of the costs in PO A and PO C are projected for highly specialised analytical support provided by EMSA during individual investigations (for PO A, equivalent to 0.7 FTE in 2030 and 0.8 FTEs in 2050 relative to the baseline; for PO C, 4.5 FTE in 2030 and 5 FTEs in 2050 relative to the baseline). 77

Table 7: Total additional cost estimates for EMSA relative to the baseline (in thousand EUR, 2020 prices)

Source: EMSA

6.1.4.Impacts on SMEs 

The extension of the AI Directive to VSMCs involving fishing vessels below 15 meters in PO C is ‘relevant’ for SMEs. Due to the non-consolidated nature of this particular segment, it can be assumed that all fishing vessels below 15 meters would qualify as small enterprises or even as micro-businesses 78 . This may be also the case for operators of fishing vessels above 15 meters, for which the length of fishing vessels is to be defined in the Directive in all policy options. Fishermen’s groups have been contacted, as part of the consultation process, and are in favour of the extension of scope.

The analysis in section 6.1.2 has shown that in PO A the additional enforcement costs for fishing vessels operators are estimated at around EUR 3,280 at EU level in 2030 and EUR 3,608 in 2050 relative to the baseline. PO B shows slightly higher costs than PO A (EUR 4,592 in 2030 and EUR 4,920 in 2050). For PO C the additional enforcement costs are estimated at EUR 16.7 to 40.7 thousand in 2030 and at EUR 18 to 44 thousand in 2050. Considering the projected number of fishing vessels in 2030 and 2050 at EU level, the additional enforcement costs relative to the baseline in all policy options are estimated at below 1 EUR per vessel 79 .

In addition, the analysis in section 6.1.2 has shown that no additional administrative costs are expected in relation to the reporting requirements for VSMCs, included in PO A as a non-regulatory action and in PO B and PO C as an obligation. This is because operators of fishing vessels below 15 meters are already obliged under national law to notify the maritime authorities of a loss of the vessel or of a fatality on-board detailing the circumstances. Therefore, no area was identified in the analysis, where significant and disproportionate costs for SMEs, in comparison to all enterprises, would result from the changes under the three policy options.

On the other hand, all policy options will improve safety to different degrees, what will generally benefit all ship operators and in particular fishing vessels operators. PO C in particular will have positive impacts in terms of safety for operators of fishing vessels below 15 meters, which are mainly small enterprises or even micro-businesses. As explained in section 6.1.2, the costs savings for ship operators in terms of avoided vessels lost are estimated at EUR 0.9 million relative to the baseline in PO A and PO B, expressed as present value over 2022-2050, and at EUR 5.7 to 9.4 million relative to the baseline in PO C. These represent benefits for the fishing vessels operators that clearly outweigh the costs in all three policy options, with the highest impact expected in PO C.

6.1.5.Functioning of the internal market and competition

All policy options are expected to have a positive impact on the functioning of the internal market, both by improving overall maritime safety for the benefit of freight customers and passengers throughout the Union as well as by ensuring that the same safety level applies throughout the Union. The positive impacts of PO C are expected to be higher than those of PO A and PO B, because the VSMCs involving fishing vessels below 15 would be included in the scope of the AI Directive resulting into a higher degree of harmonisation between Member States. This is particularly the case as only 11 Member States currently investigate VSMCs involving fishing vessels below 15 metres.

All options provide for a level playing field as all policy options improve safety and the performance of AIBs in the performance of their functions.

6.1.6.Impacts on competitiveness

As explained in section 6.1.2 and in section 6.1.4, the additional enforcement costs for ship operators (i.e. fishing vessels operators) are expected to be very limited in all three policy options and estimated at below 1 EUR per vessel in 2030 and 2050 relative to the baseline. No other additional costs are expected for the industry (i.e. ship operators) in the three policy options. On the other hand all policy options are expected to lead to costs savings for ship operators, in particular fishing vessels operators, in terms of avoided vessels lost. Therefore, it can be concluded that all three policy options may improve the competitiveness of the fishing vessels operators, the impact being higher in PO C relative to PO A and PO B.

6.2.Social impacts

Social impacts are mainly assessed in terms of impacts of the policy options on maritime safety (in terms of lives saved and injuries avoided), working conditions and fundamental rights. Costs impacts for consumers from any of the policy options have not proved possible to be quantified but are expected to be negligible.

6.2.1.Maritime safety 

As explained in section 6.1.2, not all investigations have an impact on the number of marine casualties. To estimate the benefits, a relationship between the number of finalised investigations and the number of marine casualties has been estimated, according to Annex 4 80 . The reduction in the number of casualties is subsequently translated into a reduction in the number of fatalities and injuries by using the ratios between the number of fatalities and injuries and the number of marine casualties projected in the baseline scenario. For PO A and PO B the impact on the number of lives saved and injuries avoided is expected to be limited by 2050 (2 lives saved and 10 injuries avoided relative to the baseline in PO A and 2 lives saved and 14 injuries avoided in PO B). However, PO C would result in higher number of lives saved (28 to 48) and injuries avoided (219 to 379) 81 for 2025-2050 relative to the baseline, mainly due to the assumed extension of the scope of AI Directive to the VSMCs involving fishing vessels below 15 meters. Thus, all policy options contribute towards Sustainable Development Goal 3 (“Ensure healthy lives and promote well-being for all at all ages”), although the positive impact of PO C is the highest.

All policy options are estimated to result in a reduction in the external costs of accidents relative to the baseline (Table 8) although the impact would be highest in PO C (EUR 132.6 to 229.3 million, expressed as present value over 2022-2050) relative to PO A (EUR 6.6 million) and PO B (EUR 9 million) 82 .

Table 8: Reduction in the external costs of accidents, present value over 2022-2050 (in million EUR)

Source: Ecorys (2022), Impact assessment support study 

It should be noted however that there is high uncertainty regarding these estimates. This is because the impacts of AI Directive on safety are indirect, through the safety recommendations and their follow-up. For this reason, a sensitivity analysis has been performed, assuming 10% and 20% lower value in absolute terms of the elasticity used to derive the impacts. Table 9 shows that even with lower value of the elasticity PO A, PO B and PO C are projected to result in lives saved and injuries avoided, although the positive impacts on safety would be more limited.

Table 9: Results of the sensitivity analysis on the reduction in the external costs of accidents, expressed as present value over 2022-2050 (in million EUR) relative to the baseline

It is important to note that while it is possible that police, judicial or administrative investigations might have dissuasive or deterrent effect as regards future accidents, this is not the primary function of such investigations. As these national investigations are carried out under a multiplicity of national procedures, neither they nor their results can easily be compared. Furthermore, the results of the investigations are not communicated or shared between Member States, to enable lessons to be learned. For these reasons, it is not possible to estimate their impact on maritime safety, although the impact on ship operators or AI bodies of Member States other than the investigating State is expected to be very limited.

6.2.2.Impacts on working conditions and skills

The impact of the policy options on working conditions is expected to be positive, although it has not been possible to quantify them. By improving safety, the policy options will result in saved lives (of passengers but in particular of crew), avoid injuries and improve the attractiveness of employment in the sector. The impact is expected to be higher in PO C, relative to PO A and PO B, resulting in the highest number of lives saved and injuries avoided. In addition, the knowledge sharing and training organised by EMSA included in PO A and PO C will improve the skills of investigators in light of new developments which may be relevant for accident investigation in the future, including but not limited to renewable and low carbon fuels, automation and autonomous shipping.

6.2.3.Impacts on fundamental rights 

Given the global nature of the shipping industry and the different jurisdictions with which they may be brought into contact, seafarers need special protection, especially in relation to contacts with public authorities, in line with the human rights protection standards enshrined in EU primary law. Therefore and in the interests of increased maritime safety, seafarers should be able to rely on fair treatment in the event of a maritime accident. The human rights and dignity of seafarers should be preserved at all times and all safety investigations should be conducted in a fair and expeditious manner.

The policy options were assessed to determine if they have an impact on the fundamental rights and/or equal treatment of EU citizens. The starting point of the assessment of the fundamental rights is the Charter of Fundamental Rights of the European Union 83 . The assessment identified two potentially relevant fundamental rights being Article 8 – Protection of personal data and Article 20 – Equality before the law.

Safety investigations and the reports emanating therefrom are not supposed to contain personal data so the question of protection thereof and the implications of the GDPR are somewhat limited. However, the majority of the AIBs noted that GDPR rules do play an important role that should be expressly mentioned in the Directive to guarantee the respect of the seafarers’ rights. Several AIBs have raised the issue of when the AIB has to co-operate with an AIB in a third country when conducting a joint investigation as the GDPR forbids the transfer of the personal data of EU data subjects to non EEA countries — unless appropriate safeguards are imposed, or the third country's data protection regulations are formally considered adequate by the Commission. The EMSA training on GDPR, will address this issue in all policy options.

As regards “equality before the law”, safety investigations are carried out to determine the cause of the accident and are not supposed to attribute blame or responsibility; they are separate from and distinct to police or judicial investigations. Similarly, safety investigation reports are not supposed to be written, in terms of content and style, with the intention of it being used in legal proceedings. All three of the POs were assessed having regard to the relevant EU instrument and the applicable IMO text 84  and it was concluded that they maintain full respect for human and fundamental rights referred to previously and none will have any negative impact thereon.

6.3.Environmental impacts

The impact of the policy options on the environment is also an indirect one, as it depends on the safety recommendations and their follow-up as explained in section 6.2.1. The environmental impact of maritime casualties derives from ships sinking, cargoes lost and oil spills (either as cargo or from bunkers - ship fuels). While there has not been a single significant oil spill similar to that of the Erika (1999) or Prestige (2002) oil spills in EU waters for almost 20 years, the possibility of such an incident is nonetheless present and has to be prepared for and mitigated against. Similarly and in the context of the European Green Deal and the “Fit for 55” package, it is likely that there will be a significant change in the way vessels are powered in the coming decades. This will have implications for the entire shipping and fishing industries including for the safety profile thereof and can be expected to have an impact on AIBs and the cases they investigate.

It is important to note the significance of the safety recommendations. A more recent example is the maritime safety issue related to containers lost overboard (falling off large containerships). Such an accident happened in 2019, where a large container vessel (MSC Zoe) lost a total of 342 containers whilst sailing in bad weather on the North Sea, in the Terschelling-German Bight Traffic Separation Scheme. This resulted in severe pollution of the Wadden Sea area and its islands (NL). The Wadden Sea is designated as a Particularly Sensitive Sea Area and is on the UNESCO World Heritage List.

The investigations by the flag State (Panama), together with affected States (NL and DE), identified safety issues related to various IMO instruments and requirements. Safety recommendations were made in the investigation reports regarding carriage requirements, such as, use of the improved use of the VDR to record such information for the purpose of safety investigations. The administrations involved have submitted proposals to the IMO various Maritime Safety Committees addressing recommendations included in the report for actions to be taken.

This case was also raised and discussed at the Transport, Telecommunications and Energy Council in summer 2021, under the Slovenian Presidency; where Ministers agreed to address the issue of containership safety, including containers lost at sea, fire safety etc., in a holistic manner following a root cause analysis. This work is ongoing, and a first Union submission to IMO has been agreed regarding the issue of container lost at sea and setting up an international mandatory reporting system, based on the legal requirements applicable in EU law already, requiring changes to the conventions (SOLAS and MARPOL).

Furthermore, 23 cases of pollution due to bunker fuel lost were recorded in 2019 for all vessels excluding fishing vessels below 15 meters. Assuming the same ratio of pollution to VSMCs for fishing vessels below 15 meters as for those above 15 meters, the overall number of pollution cases due to bunker fuel lost including for fishing vessels below 15 meters is estimated at 103 for 2019. The level of pollution per very serious accident (resulting in pollution) has been estimated as involving the loss of 30 tonnes of bunker fuel. For small fishing vessels, the level of pollution per very serious accident (resulting in pollution) implied the loss of 6.6 tonnes of bunker fuel while for larger fishing vessels a loss into the sea of 22 tonnes of bunker fuel was indicated.

In the baseline scenario the cumulative number of tonnes of bunker fuels lost between 2025 and 2050 is estimated at around 35.4 thousand. PO A and PO B would result in 7 tonnes of bunker fuel lost avoided during 2025-2050 relative to the baseline, while PO C is estimated to have a more significant impact (101 to 176 tonnes of bunker fuel lost avoided) due to the inclusion of the VSMCs involving fishing vessels below 15 meters in the scope of the Directive (see Table 10). This is expected to have a positive impact on the quality of marine water and biodiversity. It was not possible to quantify the environmental impact from avoided ships sinking and cargoes lost, but this is expected to be positive in all three policy options with the highest impact in PO C 85 .

Table 10: Bunker fuel lost cumulative for 2025-2050 (in tonnes)

Source: Ecorys (2022), Impact assessment support study

It should be noted however that there is high uncertainty regarding these estimates. This is because the impacts of AI Directive on safety are indirect, through the safety recommendations and their follow-up. For this reason, a sensitivity analysis has been performed, assuming 10% and 20% lower value in absolute terms of the elasticity used to derive the impacts. Table 11 shows that even with lower value of the elasticity PO A, PO B and PO C are projected to result in bunker fuel lost at sea avoided, although the positive impacts would be more limited.

Table 11: Results of the sensitivity analysis on the bunker fuel lost at sea, cumulative for 2025-2050 (in tonnes) relative to the baseline

All policy options contribute towards Sustainable Development Goal 14 (“Conserve and sustainably use the oceans, seas and marine resources for sustainable development”), although the positive impact of PO C is the highest.

No significant harm is expected on the environment in any of the three policy options, in particular in the area of sustainable use and protection of water and marine resources to which the initiative relates. On the contrary, as explain above, all three policy options are expected to have small positive impacts on the quality of water and biodiversity – with the highest impact among the three projected in PO C.

7.HOW DO THE OPTIONS COMPARE?

7.1.Effectiveness

The assessment of effectiveness looks at the extent to which the general and specific objectives (SO) of the intervention, as previously described, are met. Table 12 provides the link between policy objectives and assessment criteria.

Table 12: Link between objectives and assessment criteria

Concerning SO1, PO A and PO B should allow for more information to be gathered and thereby allow the AIBs, the Commission and EMSA to get a better appreciation and understanding of the safety picture of fishing vessels below 15 meters and the types of marine casualties they are affected by. However, neither PO A nor PO B requires investigations to be carried out for this type of fishing vessels. PO C aims to extend the scope of the Directive to all fishing vessels irrespective of their length in case they are involved in a VSMC. PO C is fully effective in addressing SO1 in that it requires both the reporting of and the investigation of VSMC involving fishing vessels below 15 meters, with positive effects on safety and protection of marine environment. PO A and PO B address SO1 to a more limited extent by requiring that the reporting of VSMCs is applied in different ways to the smaller fishing vessel segment.

Concerning SO2, as highlighted in the problem definition, some of the definitions used in the Directive are currently not always clear for the AIBs. As a result, AIBs do not always know whether or not an investigation should be started, which may also have negative impacts on the functioning of the internal market due to the different approaches by Member States. By making those definitions clearer, AIBs will better understand whether to investigate or not. Measures aiming to improve (some of) the definitions used are included in all three policy options. In PO A two elements are clarified, namely, how to measure the length of a fishing vessel and how to define the phrase ‘in direct connection with the operations of the ship’, by providing interpretive guidelines. As such, PO A contributes to some extent to achieve SO2. PO B and PO C, in addition to the clarification of how to measure the length of a fishing vessel, go further by including the definition of ‘directly in connection with the operations of a ship’ in the Directive and by including a specific deadline for accidents resulting in a fatality that does not directly occur after the accident took place. Therefore, PO B and PO C are more effective than PO A in achieving SO2. All three options include references to the most up-to-date versions of the IMO Casualty Code. Thus, it can be concluded that in relation to SO2, PO B and PO C are equally effective and more effective than PO A. All policy options contribute to improving the functioning of the internal market. In this respect, PO C is most effective, due to the inclusion of the VSMCs involving fishing vessels below 15 meters within the scope.

With regard to SO3, the problem definition highlighted that AIBs sometimes lack adequate resources (either in the number of staff or available budgets) to effectively conduct all investigations needed. To tackle this problem, several policy measures have been defined. PO B includes several specific policy measures. On the one hand, this option contains a measure on introducing a QMS system that ensures harmonisation of the investigation procedures throughout the EU. A key part of improvements in efficiency is to establish control conditions and a QMS supports this as it is supposed to indicate with more precision where there is a ‘problem’ or systematic ‘fault’ so that the administration can take action. The better management of often scarce resources will identify system problems and allow these to be addressed either by seeking help from EMSA or by making the case for additional resources at the national level. It is to be noted that this includes availability of sufficient and adequate (technical) resources. If the AI function had not been separated from flag State administrations the EU AI bodies as part of the flag State administrations would have already been covered by the QMS requirement in Directive 2009/21/EC 86 and consequently would be audited in accordance with the III-code.

Since the IMO Code on the Implementation of International Instruments (the “III-code”) became mandatory in 2017, it is acknowledged that a QMS ensures better adherence with the III-code requirements. Therefore, it is appropriate for the reasons previously explained as well as to ensure coherence between the different parts of the MS maritime administrations that this requirement is extended to the MS AI and port State control functions, and it is thus included in all policy options. On the other hand, PO B consists of a measure that broadens the group of competent authorities that could include information in EMCIP. At the same time, the role of EMSA is slightly enhanced by only providing sharing knowledge and organising training. By having these dedicated measures, PO B is somewhat effective in achieving SO3. PO A has a different approach in addressing SO3 and includes several measures on support from EMSA to the AIBs. This could either relate to soft skills (such as specific technical support) or hardware (such as specific analytical tools). In addition, EMSA could provide additional training. Besides support from EMSA, PO A also introduces a QMS. By the wider range of support offered, PO A is more effective in achieving SO3 than PO B. PO C is a mix of PO A and PO B and includes all measures proposed. More concretely, not only a QMS is introduced but both the support from EMSA as well as the broadening of the group of competent authorities are included. Thus, PO C and PO A are more effective than PO B in addressing SO3.

All policy options contribute to the general objective of improving maritime safety and the protection of the marine environment. PO C is however more effective than PO A and PO B due to the higher positive impact estimated in terms of lives saved, injuries avoided and pollution due to bunker fuels lost avoided. Annex 9 provides a more detailed overview on the effectiveness of the policy options in relation to the specific objectives.

7.2.Efficiency

Efficiency concerns "the extent to which objectives can be achieved for a given level of resource/at least cost". The combined measures under the three POs have economic, social and environmental impacts. The major costs of the policy options come in the form of investigation costs for AIBs and costs for EMSA related to soft skills (such as specific technical support) or hardware (such as specific analytical tools) and additional training. They are summarised in Table 13.

Table 13: Summary of costs and benefits of policy options – present value for 2022-2050 compared to the baseline (in million EUR)

Source: Ecorys (2022), Impact assessment support study; Note: for the net costs/benefits, negative values indicated net costs and positive values net benefits.

PO C leads to the highest total costs among the three policy options, estimated at EUR 44.2 to 174.17 million in addition to the baseline costs, expressed as present value over 2022-2050. The highest total costs in PO C are linked to the assumed extension of the scope of the AI Directive to VSMCs involving fishing vessels below 15 meters and thus the increased number of investigations. PO A shows lower costs than PO C, estimated at EUR 28.42 to 38.08 million in addition to the baseline costs, expressed as present value over 2022-2050. The highest cost category in PO A is related to costs for EMSA for soft skills (such as specific technical support) or hardware (such as specific analytical tools) and additional training. Finally, PO B shows the lower total costs among the POs, estimated at EUR 8.81 – 22.11 million in addition to the baseline costs.

In terms of total benefits, PO C shows the highest benefits among the three policy options due to the reduction in the external costs of accidents and the avoided vessels lost linked to the extension of the scope to fishing vessels below 15 meters. These benefits are estimated at EUR 138.28 to 238.68 million relative to the baseline, expressed as present value over 2022-2050. PO A and PO B show much lower total benefits of EUR 7.52 million and EUR 9.97 million, respectively. The impact of the avoided pollution due to the tonnes of bunker fuel lost was not possible to be monetised but also in this case the highest benefits are projected in PO C, as shown in section 6.2.1.

Overall, PO A results in net costs relative to the baseline, expressed as present value over 2022-2050, while PO C shows net benefits. For PO B, the range provided shows that it could result either in net costs or net benefits although the possible net costs are lower than in PO A. The net benefits are largest in PO C, estimated at EUR 64.51 to 94.09 million relative to the baseline, expressed as present value over 2022-2050. PO C also shows higher benefits to costs ratio (1.4 to 3.1) relative to PO B and PO A (see Table 13).

Considering the sensitivity analysis on the impacts of the policy options on the costs savings for ship operators in terms of avoided vessels lost and on the external costs of accidents, provided in sections 6.1.2 and 6.2.1 respectively, the net benefits and the benefits to costs ratios have been calculated for each case and are provided in Table 14. The table shows that even with lower values of the elasticity, PO C would result in the highest benefit to cost ratio.

Table 14: Results of the sensitivity analysis on net benefits and benefit to cost ratio of policy options

7.3.Coherence

Internal coherence. The internal coherence concentrates on how the different elements within the Directive itself work together to achieve the objectives. It should be noted that this does not only concern the Directive itself, but also its accompanying secondary legislation (delegated and/or implementing acts) and rules as well as interpretative guidelines. Although all three POs address the identified problems, they do so in different ways. PO A addresses the problems in such a way that room for flexibility remains, meaning that the majority of guidance would be laid down in interpretative guidelines, which are a supporting tool for AIBs. In addition, AIBs are encouraged to undertake certain activities, however, there is no obligation to do so. This notwithstanding, while coherence exists there is still room for improvement. PO B and PO C propose amendments to the Directive itself for all aspects that require further harmonisation and thus ensure a higher degree of internal coherence than PO A.

External coherence. The external coherence concentrates on the compliance of the Directive with key EU policy objectives and international legislation. Revising the Directive and deleting the references to outdated international law will ensure that the Directive and IMO legislation will be aligned. As such, both legal regimes will be consistent. As PO A, PO B and PO C all seek alignment with the current international legal regime, external coherence will be guaranteed in all three policy options. 

7.4.Subsidiarity and proportionality

As highlighted in section 2 there is a clear need for EU action on all three problems identified, and their drivers. The current Directive does not apply to all vessel types involved in VSMCs, contains unclear definitions and refers to outdated legislation. In addition, AIBs are not able to effectively and consistently report and investigate marine casualties. Member States individually are not able to tackle all the problems identified. While it is possible to adopt national legislation to tackle some of the problems highlighted, if 27 Member States would do this and adapt their own legislation and/or rules this would lead to fragmentation of legislation and potential distortion of the internal market. To avoid such a fragmented legal framework, there is a need for EU action. As such, there is a right for the EU to act. As all policy options ensure harmonisation of the legal framework, the subsidiarity requirement is fulfilled. In any event, as stated in Section 3.2, the principle of subsidiarity does not apply to areas subject to EU exclusive competence pursuant to Article 3(2) TFEU.

In relation to proportionality, the proposed revision aims to improve maritime safety. This objective is achieved by investigating casualties and learning from them, to prevent similar accidents from happening in the future. PO A contributes to this objective by strongly focusing on improving the ability of AIBs to investigate accidents. Focus is put on providing more support through soft skills and hardware as well as knowledge sharing and training by EMSA. The additional support offered is considered proportional. PO B contributes to the objective by strongly focusing on clarifying several of the definitions used. As the Member States cannot easily tackle unclear definitions themselves it is important that it is done on an EU-level. During the stakeholder consultation, it became clear that clarifying the definitions is one of the most important aspects for AIBs in the revision of the Directive. The measures proposed under PO B are therefore considered to be proportionate. PO C combines elements of PO A and PO B and adds the scope extension to fishing vessels below 15 meters. The scope extension is considered to be proportionate. The additional costs necessary for AIBs to investigate VSMCs involving fishing vessels below 15 meters seem to be limited, especially, when compared to the potential positive safety impacts it can yield.

8.Preferred policy option

8.1.Identification of the preferred policy option and stakeholders views 

Although each of the options addresses the problems identified, their drivers and the specific objectives, some options are more effective in achieving the specific and general objectives. Based on the assessment done, PO C is regarded as the most effective policy option. With respect to efficiency, PO C has the highest additional costs, followed by PO A. However, when linking costs to benefits, PO C also yields the highest impact in terms of improving maritime safety and thus monetised benefits. As the additional benefits outweigh the additional costs, PO C is seen as the most efficient option proposed. Both the net benefits and the benefits to costs ratio are higher in PO C relative to PO B (when considering net benefits for PO B out of the range provided), while PO A results in net costs. Concerning external coherence, the policy options score more or less the same. PO C is slightly more coherent than PO A and PO B as besides coherence with international legislation, PO C also improves coherence by extending the scope of the Directive to fishing vessels below 15 meters. Concerning internal coherence, PO C is also the most coherent option. In terms of subsidiarity, all options fulfil this principle and there are no real differences among them. The three options are also proportionate. None of the options leads to excessive costs in achieving the objectives set even if PO A results in net costs relative to PO B and PO C.

Based on the analysis above it can be concluded that PO C is the preferred policy option. There is mixed opinion among AIBs as to whether the scope should be extended. AIBs already investigating casualties involving smaller fishing vessels are in favour of extending the scope. AIBs not yet investigating casualties involving smaller fishing vessels held mixed opinions. Some of them, especially the small ones, are against a scope extension as they fear that it will increase their workload substantially. The larger AIBs seem to be more in favour as they see advantages of investigating the accidents as in the long run it could improve safety. The sea-fisheries social partners (employers and employees representatives) have expressed support for this measure, underlining the importance of gathering solid, detailed information on the causes of accidents.

8.2.REFIT

This initiative is part of the Commission Work Programme 2021 under Annex II (REFIT initiatives), under the heading Promoting our European Way of Life 87 . 

The initiative has an important REFIT dimension in terms of alignment and simplification of safety legislation, of improving the safety profile in particular of the small fishing vessel segment and of assisting Member State AIBs to better discharge their reporting and investigative functions. Partially extending the Directive to the smaller fishing fleet while increasing the overall policy ambition, the review also includes some important simplification aspects. AIBs will face higher costs due to the scope extension as they will have to carry out at least a preliminary assessment of VSMCs involving smaller fishing vessels but this has to be seen against the simplification brought about by the removal of the requirement to mandatorily carry out such a preliminary assessment in the case of all serious casualties. It is however difficult to identify any cost savings stemming from this simplification as all casualties are subject to some form of assessment to establish the circumstances, to determine their seriousness and whether they need to be investigated.

In addition, the preferred policy option includes elements of simplification:

·It will increase the efficiency of the national AIBs by providing clarity on when they need to carry out investigations.

·The provision by EMSA of top-up assistance to those AIBs that request operational support in the form of analytical tools, hi-tech hardware such as drones or facilities to deal with VDRs will improve the quality and quantity of investigations carried out as well as increasing their harmonisation and standardisation across the EU.

·EMSA training to AIB staff on technological and regulatory developments as well as on safety issues arising from renewable and low carbon fuels and other developments arising from the European Green Deal as well as issues around autonomous shipping should also provide a boost to AIBs in tackling the significant coming changes in maritime transport.

·EU AIBs are, under the relevant IMO instruments (as is the case for all IMO States) obliged to report accident information data to the IMO Global Integrated Shipping Information System (GISIS) 88 , however once the accident data is reported to EMCIP the possibility is included in EMCIP for this data to be transferred onwards to GISIS. Similarly if a Member State AIB wishes to make EMCIP its national database, EMSA as database manager can customise EMCIP with the required functionalities such that the AIB can report one time to EMCIP, thereby obviating the need for a national reporting as well as EU reporting. As such, the AI Directive already uses interoperable digital solutions and could be considered digital-ready.  

8.3.Application of the ‘one in, one out’ approach 

As explained in sections 6.1.2 and 6.2, the preferred policy option is not expected to result in additional administrative costs or adjustments costs for the private sector, or for the citizens. Additional enforcement costs for the private sector (i.e. ship operators) are estimated at EUR 0.29 to 0.7 million relative to the baseline, expressed as present value over 2022-2050. They would however be largely overcompensated by the benefits in terms of vessels lost avoided, which are estimated at EUR 5.68 to 9.41 million relative to the baseline, expressed as present value over 2022-2050.

9.How will actual impacts be monitored and evaluated?

The Commission services will monitor the implementation and effectiveness of this initiative through a number of actions and a set of core indicators that will measure progress towards achieving the operational objectives. Five years after the end of the implementation date of the legislation, the Commission services should carry out an evaluation to verify to what extent the objectives of the initiative have been reached.

Actions foreseen for verifying implementation include:

-Commission/EMSA monitoring of EMCIP database to verify that investigations are being carried out in a timely and effective manner and that the reports are uploaded to the database, including for fishing vessels below 15 meters for which data and safety gaps are identified. The indicators that will be specifically monitored drawing on EMCIP database will be the number of fatalities and very serious injuries, the number of vessels lost and the number of cases of pollution at sea, and their evolution over time. All these indicators will be monitored by type of vessel involved. The number of safety recommendations published as a result of the investigations and their follow up will also be monitored.

-Visits to Member States to verify operations on the ground, these are carried out by EMSA on behalf of the Commission as part of EMSA’s support role to the Commission 89 . To be translated into visits reports and if shortcomings are identified they are addressed either by means of EMSA training or through the EU pilot and infringement procedure.

-Member States will have to have a QMS to certify its organisation, policies, processes, resources and documentation are appropriate to achieve its objectives. This will have to be certified and subsequently subject to annual audit. AIBs will have to share with Commission/EMSA the results of the annual audits carried out by the accredited body such that the AIB can retain its QMS certification.

-Horizontal analysis required 90 to be carried out by EMSA (giving an indication of how the legislation is functioning and identifying gaps and what can be done to address them) and reported to the Commission and Member States (discussed in workshops).

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

The lead DG is DG MOVE, Unit D2: Maritime Safety

DECIDE reference number: PLAN/2019/5433

Item 35 in Annex II to Commission Work Programme 2021

2.Organisation and timing

The impact assessment follows the ex-post evaluation of the maritime accident investigation Directive performed as part of the overall maritime transport policy fitness check in 2018.

The impact assessment started in 2020, with inception impact assessment published on 9 October 2020 91 . 

The impact assessment on a possible review of the Accident Investigation Directive was coordinated by an Inter-Service Steering Group (ISSG). The Commission Services participating in the ISG were: Secretariat-General, DG Maritime Affairs and Fisheries, DG Climate Action, DG Migration and Home Affairs, DG for Employment, Social Affairs and Inclusion, DG Industry, Entrepreneurship and SMEs, DG Environment, DG Health and Food Safety and the European Maritime Safety Agency (EMSA).

The Inter-Service Steering Group met 4 times: on 22 January 2021, 3 March 2021, 15 October 2021 and 24 March 2022. It was consulted throughout the different steps of the impact assessment process: notably on all stakeholder consultation materials and deliverables from the external contractor and on the draft Staff Working Document.

3.Consultation of the RSB

The draft report was discussed by the Regulatory Scrutiny Board on 27 April 2022, which issued a positive opinion. The Board also made several main recommendations for improvement which were addressed in the final impact assessment report as follows:

The draft report was submitted to the RSB on 30 March 2022 and was discussed by the Board on 27 April 2022.

4.Evidence, sources and quality

The impact assessment is based on several sources, using both quantitative and qualitative data. This includes:

•    Ex-post evaluation of the maritime accident investigation Directive

•    Maritime Fitness Check 2018 

•    Stakeholder consultation activities (see Annex 2)

•    External support study carried out by an independent consortium (lead by Ecorys)

•    Commission experience in monitoring and implementing the Directive

•    Various databases managed by EMSA.

Annex 2: Stakeholder consultation (Synopsis report)

This annex provides a summary of the outcomes of the consultation activities which have been carried out for the review of the Accident Investigation Directive, including in the context of the external support study. It notes the range of stakeholders consulted, describes the main consultation activities and provides a succinct analysis of their views and the main issues they raised.

The objective of the consultation activities were to collect information and opinions of stakeholders on the key problems and associated drivers, definition of relevant policy objectives linked to those problem areas and the identification, definition and screening of policy measures that could eventually be incorporated into policy options for this Impact Assessment, as well as gather information and opinions on their likely impacts.

1.    Overview of consultation activities

A consultation strategy, covering the stakeholder consultation activities carried out, including as part of the support study, has been developed from the start and further fine-tuned throughout the different phases of the impact assessment. The consultation activities were aimed at a range of relevant stakeholders representing accident investigation bodies in EU and industry representatives (including relevant associations of shipowners, fishermen, port operators, recognised organisations seafarers and social partners).

Consultation activities have taken place since the publication of the Inception Impact Assessment in October 2020 and continued until early 2022, with the bulk of consultations taking place in 2021.

As part of the initial feedback mechanism, stakeholders had the opportunity to provide feedback on the Inception Impact Assessment 92 via the relevant website. The Commission received 5 responses, during 09 October to 20 November 2020. There was one response provided by one EU Member State, one from an EEA country and 3 provided by business associations, actors from shipping industry, including a trade association of seafarers and an NGO. A further response was received from a Member State which did not use the relevant website.

Afterwards, the following main consultation activities were carried out:

-Considering the highly technical nature of the file, it was agreed with the Secretariat General that there is no added value in running an Open Public Consultation. However, the general public was offered the opportunity to provide any views on this initiative, via an announcement on DG MOVE web page 93  and a dedicated functional mailbox. One contribution was received from a shipowner association relating to accident investigation which welcomed the administrative burden reduction and possible support to AIBs from EMSA.

-Three rounds of interviews with EU level representatives of key stakeholders organised by the consultant in charge of the external support study, running intermittently from February 2021 to November 2021, to fill specific information requests, particularly to support and refine the overall problem definition and possible policy options.

-Targeted stakeholder surveys organised by the consultant in charge of the external support study, running from June to November 2021, to gather specific information, particularly to support the refinement of the problem definition, the baseline and where possible the assessment of impacts of possible policy measures.

-Additional consultation activities organised by DG MOVE and the consultant in charge of the external support study in order to consult the Member States and key stakeholders by providing their views on the different policy measures but also to validate the emerging and final results of the support study to the Impact Assessment in terms of the quantification of the impacts. These activities took place in the context of  meetings of the EU Sectoral Social Dialogue Committee on maritime transport (16 April 2021, 23 September 2021 and 16 December 2021),  the EU Sectoral Social Dialogue Committee on Sea Fisheries (29 January 2021, 8 March 2021 and 16 November 2021), the EU Sector Social Dialogue Committee on ports (19 November 2021), meetings of the Permanent Cooperation Framework of AIBs (10 March 2021, 18 June 2021, 29 August 2021, 22 September 2021 and 7 October 2021), an informal meeting of the EU/EEA Maritime Transport Directors (30 November 2021), meetings of the of the EU Committee on Safe Seas and the Prevention of Pollution from Ships (17 May 2021 and 11 November 2021). A final validation workshop to validate the conclusions of the support study attended by Member State and industry representatives was also organised (20 January 2022).

The information collected from stakeholders was key in allowing the Commission to refine the design of the policy options as well as to assess their economic, social and environmental impacts, compare them and determine which policy option is likely to maximize the benefits/costs ratio for the society and fully contribute to achieving a more effective and efficient mechanism to investigate the maritime accidents across the EU. Findings from those processes complemented the desk research carried out in the context of the external support study.

Methods have been adapted to take account of the development of the COVID-19 pandemic. For this reason, interviews and meetings have been held by videoconference.

Table 15:  Overview of responses to different stakeholder consultation activities

2.    Limitations of the Stakeholder consultation

Not all stakeholders were very responsive to the various targeted consultation activities, in particular fishermen representatives (slow responses or incomplete answers). However, since all relevant stakeholder groups have provided their views and positions to the various targeted consultations, a meaningful comparison and analysis of opinions gathered from all consultation activities was nevertheless possible.

It was particularly difficult to gather input from stakeholders on possible expected costs of implementing proposed measures, as well as estimations on the number of accidents because of the scarcity and incompleteness of existing data. In certain instances, the responses of the accident investigation bodies were not very consistent, which may be explained by their differing histories and stage of development (pre-2009 or post), structures (maritime only or multi-modal), size and resources, differing appreciations of issues such as definitions and whether there are problems with resources, concerns over additional, increased workload in the future, while ensuring that their full independence and responsibility in deciding the accidents they think are worth being investigating and the conduct of the investigations themselves.

3.    Analysis of the key results of the stakeholder consultation 

The remainder of the annex presents key findings from the analysis of stakeholder contributions to the consultation process. They are structured around the main elements of the intervention logic, including the problem areas and their drivers, the policy objectives as well as the key aspects of the design of possible policy measures. The technical support study for this Impact Assessment contains the detailed presentation of findings from the targeted consultation activities.

While AI Bodies are overall of the opinion that the fact that the fishing vessels below 15 meters are not covered by the Directive is not an issue per se, some rather pointing to difficulties in interpreting a number of the definitions provided in the Directive, they nevertheless acknowledge the high risks of accidents in this particular sector:

Figure 7: Are there certain types of fishing vessels that are more accident prone? (n = 12)

Figure 8: Do you agree that unclear definitions in the Directive are an issue? (n = 18)

Figure 9: Do you agree that a lack of resources available to AIBs is an issue? (n = 17)

Figure 10: AIB survey - In your expert opinion, what grounds should justify the inclusion of smaller fishing vessels (under 15m) in the scope of the Directive? (multiple answers possible) (N=16)

Figure 12: How will this solution affect the workload of your AIB? (n = 20)

Annex 3: Who is affected and how?

1.Practical implications of the initiative

Summary of the preferred policy option implementation

The revision of the Maritime Accident Investigation Directive aims at maintaining and improving the level of maritime safety of EU registered ships operating worldwide and of ships operating in EU waters. The impacts of the preferred policy option are expected to fall on different stakeholder groups: national accident investigation bodies, EMSA, industry (i.e. ship owners/operators), crews of the vessels, port workers and passengers of maritime vessels.

Ensuring a high level of safety is important for the users of transport transporting goods as well as for passengers. It is also important for vessel crews as these persons make up the largest number of persons killed and/or injured in maritime accidents. It is important for consumer protection as well as for the integrity of the internal market that occurrences are systematically investigated in a coherent and harmonised manner across the European Union. There can be no gaps in the maritime safety net. Any and all lessons that can be learned should be taken advantage of to avoid future reoccurrence.

It is also important for the environment that accidents are avoided and when these do unfortunately occur that their effects are mitigated to the extent possible, this is why it is essential that accidents are investigated and that the results of the investigations are disseminated to the widest extent possible throughout the Union in a systemic manner.

The preferred policy option identified in the context of this Impact Assessment, policy option C, provides for a minimum requirement that any very serious marine casualty (in essence a fatality or the loss of a vessel) involving a fishing vessel of below 15 metres in length overall is subject to a preliminary assessment to determine if a full safety investigation should be carried out.

The preferred policy option also clarifies on a number of problematic or ambiguous definitions by amending the Directive. It aligns the Directive with the most up to date IMO provisions and also requires that AIBs have a certified quality management system in place to assure their operations.

Finally the preferred policy option envisages assistance to those AIBs that request it in terms of training as well as operational support whereby EMSA can assist AIBs to better discharge their obligations under the Directive.

Implications on consumers, market actors and public authorities

The following key target groups of this initiative have been identified:

·Accident investigation bodies in EU Member States

·European Maritime Safety Agency (EMSA)

·Owners/operators of fishing vessels of below 15 metres length overall

·Owners/operators of fishing vessels of above 15 metres length overall

·Owners/operators of maritime transport vessels

·Crews of the above-mentioned categories of vessel

·Port workers

·Passengers of maritime vessels

AIBs will be affected in three ways: firstly the partial extension of scope to smaller fishing vessels means that as a minimum they will have to carry out a preliminary assessment of very serious marine casualties involving these vessels. This is expected to increase the amount of work they have to do. AIBs will also face a slightly increased number of investigations because of changes and increased clarity in definitions. The increase is somewhat mitigated by the fact that the AI body can decide not to carry out a full investigation if it feels that there are no lessons to be learned. These investigations are expected to result in new findings as fishing vessels below 15 meters are currently not in the scope of the AI Directive.

AIBs will benefit from the technical and operational support that they will receive from EMSA which should allow them to better discharge their obligations in an efficient and timely fashion particularly when confronted with multiple investigations which have to be carried out at the same time.

The third aspect is that AIBs will have to put in place a quality management system (QMS). The AIBs internal systems will have to be certified and then the AIB will have to be audited each year to retain its certification. This implies one-off costs for putting in place the system and ongoing costs for the yearly audit.

EMSA will be impacted in that it will have to provide different forms of operational support as well as training to AIBs. This will take the form of support to the AIBs to carry out investigations either in terms of soft skills in the form of expertise or hardware such as drones or laboratory assistance to deal with technical aspects.

In fisheries, the owners/operators and crews of fishing vessels of below 15 metres in length overall are the stakeholder category mostly impacted by the proposed intervention. The extension of the scope of the Directive to this category of vessel should lead to an improvement in the safety profile of this vessel segment and at the very least to a better understanding of the safety problems faced by vessels within this category. Additional costs for the sector of participating in accident investigations (i.e. enforcement costs) are however limited.

Maritime transport operators and the owners /operators of fishing vessels of above 15 metres (both of which are currently within the scope of the Directive) will be impacted in that a limited number of additional investigations will be carried out as a result of increased clarity as regards definitions of when an investigation has to be carried out. The additional costs for the sector of participating in accident investigations (i.e. enforcement costs) are however limited.

On the other hand, benefits are expected in terms of avoided vessels lost, due to improved safety. These benefits overcompensate the costs for the industry (i.e. shipowners/operators).

Given that crews are systematically the most impacted category of people as regards injuries and death in maritime transport any improvement to safety will impact on them positively. Positive impacts in terms of safety improvements are also expected for the port workers and the passengers of maritime vessels.

2.Summary of costs and benefits

3.Relevant sustainable development goals

Annex 4: Analytical methods

1.Description of the analytical methods used 

The main model used for developing the baseline scenario for this initiative is the PRIMES-Maritime transport model by E3Modelling, a specific sub-module of the PRIMES and PRIMES-TREMOVE models. The model has a successful record of use in the Commission's energy, transport and climate policy assessments. In particular, it has been used for the impact assessments underpinning the “Fit for 55” package 95 , the impact assessments accompanying the 2030 Climate Target Plan 96 and the Staff Working Document accompanying the Sustainable and Smart Mobility Strategy 97 , the Commission’s proposal for a Long Term Strategy 98  as well as for the 2020 and 2030 EU’s climate and energy policy framework.

For the assessment of the impacts of the policy options an excel-based tool has been developed by Ecorys in the context of the impact assessment support study 99 . The tool draws on the Standard Cost Model for the assessment of the administrative costs and also includes an assessment of the impacts on the maritime safety. The excel-based tool builds extensively on data from EMCIP, provided by EMSA, and the analysis of stakeholders' feedback. The proposed measures which involve the amendment of the Directive are assumed to be implemented from 2025 onwards, so that the assessment has been undertaken for the 2025-2050 period and refers to EU27. Costs and benefits are expressed as present value over the 2022-2050 period, using a 3% discount rate.

PRIMES-Maritime model

The PRIMES-Maritime transport model is a specific sub-module of the PRIMES and PRIMES-TREMOVE models and aims to enhance the representation of the maritime sector within the energy-economy-environment modelling nexus. The model, which can run in stand-alone and/ or linked mode with PRIMES and PRIMES-TREMOVE, produces long-term transport activity, energy and emission projections, until 2070, separately for each EU Member State. The coverage of the model includes the European intra-EU maritime sector as well as the extra-EU maritime shipping. The model covers both freight and passenger international maritime. PRIMES-Maritime focuses only on the EU Member State, therefore trade activity between non-EU countries is outside the scope of the model. The model considers the transactions (bilateral trade by product type) of the EU-Member States with non-EU countries and aggregates these countries in regions. Several types and sizes of vessels are considered.

PRIMES-Maritime features a modular approach based on the demand and the supply modules. The demand module projects maritime activity for each EU Member State by type of cargo and by corresponding partner. Econometric functions correlate demand for maritime transport services with economic indicators considered as demand drivers, including GDP, trade of energy commodities (oil, coal, LNG), trade of non-energy commodities, international fuel prices, etc. The supply module simulates a representative operator controlling the EU fleet, who offers the requested maritime transport services. The operator of the fleet decides the allocation of the vessels activity to the various markets (representing the different EU MS) where different regulatory regimes may apply (e.g. environmental zones). The fleet of vessels is disaggregated into several categories. PRIMES-Maritime utilises a stock-flow relationship to simulate the evolution of the fleet of vessels throughout the projection period and the purchasing of new vessels.

PRIMES-Maritime solves a virtual market equilibrium problem, where demand and supply interact dynamically in each consecutive time period, influenced by a variety of exogenous policy variables, notably fuel standards, pricing signals (e.g. Emission Trading Scheme), environmental and efficiency/operational regulations and others. The PRIMES-Maritime model projects energy consumption by fuel type and purpose as well as CO2, methane and N2O and other pollutant emissions. The model includes projections of costs, such as capital, fuel, operation costs, projections of investment expenditures in new vessels and negative externalities from air pollution.

The model serves to quantify policy scenarios supporting the transition towards carbon neutrality. It considers the handling of a variety of fuels such as fossil fuels, biofuels (bioheavy 100 , biodiesel, bio-LNG), synthetic fuels (synthetic diesel, fuel oil and gas, e-ammonia and e-methanol) produced from renewable electricity, hydrogen produced from renewable electricity (for direct use and for use in fuel cell vessels) and electricity for electric vessels. Well-To-Wake emissions are calculated thanks to the linkage with the PRIMES energy systems model which derives ways of producing such fuels. The model also allows to explore synergies with Onshore Power Supply systems. Environmental regulation, fuel blending mandates, greenhouse gas emissions reduction targets, pricing signals and policies increasing the availability of fuel supply and supporting the alternative fuel infrastructure are identified as drivers, along fuel costs, for the penetration of new fuels. As the model is dynamic and handles vessel vintages, capital turnover is explicit in the model, influencing the pace of fuel and vessel substitution.

Data inputs

The main data sources for inputs to the PRIMES-Maritime model, such as for activity and energy consumption, comes from EUROSTAT database and from the Statistical Pocketbook "EU transport in figures” 101 . Other data comes from different sources such as research projects (e.g. TRACCS project) and reports. PRIMES-Maritime being part of the overall PRIMES and PRIMES-TREMOVE transport model is calibrated to the EUROSTAT energy balances and transport activity; hence the associated CO2 emissions are assumed to derive from the combustion of these fuel quantities. The model has been adapted to reflect allocation of CO2 emissions into intra-EU, extra-EU and berth, in line with data from the MRV database 102 . For air pollutants, the model draws on the EEA database. In the context of this exercise, the PRIMES-Maritime model is calibrated to 2005, 2010 and 2015 historical data.

2.Baseline scenario

In order to reflect the fundamental socio-economic, technological and policy developments, the Commission prepares periodically an EU Reference Scenario on energy, transport and GHG emissions. The socio-economic and technological developments used for developing the baseline scenario for this impact assessment build on the latest “EU Reference 2020 scenario” (REF2020) 103 . The same assumptions have been used in the policy scenarios underpinning the impact assessments accompanying the “Fit for 55” package 104 .

Main assumptions of the Baseline scenario

The main assumptions related to economic development, international energy prices and technologies are described below.

Economic assumptions

The modelling work is based on socio-economic assumptions describing the expected evolution of the European society. Long-term projections on population dynamics and economic activity form part of the input to the model and are used to estimate transport activity, particularly relevant for this impact assessment.

Population projections from Eurostat 105 are used to estimate the evolution of the European population, which is expected to change little in total number in the coming decades. The GDP growth projections are from the Ageing Report 2021 106 by the Directorate General for Economic and Financial Affairs, which are based on the same population growth assumptions.

Table 16: Projected population and GDP growth per Member State

Beyond the update of the population and growth assumptions, an update of the projections on the sectoral composition of GDP was also carried out using the GEM-E3 computable general equilibrium model. These projections take into account the potential medium- to long-term impacts of the COVID-19 crisis on the structure of the economy, even though there are inherent uncertainties related to its eventual impacts. Overall, conservative assumptions were made regarding the medium-term impacts of the pandemic on the re-localisation of global value chains, teleworking and teleconferencing and global tourism.

International energy prices assumptions

Alongside socio-economic projections, transport modelling requires projections of international fuel prices. The projections of the POLES-JRC model – elaborated by the Joint Research Centre and derived from the Global Energy and Climate Outlook (GECO 107 ) – are used to obtain long-term estimates of the international fuel prices. The table below shows the oil prices assumptions of the baseline and policy options of this impact assessment.

Table 17: Oil prices assumptions

Source: Derived from JRC, POLES-JRC model, Global Energy and Climate Outlook (GECO)

Technology assumptions

Modelling scenarios is highly dependent on the assumptions on the development of technologies - both in terms of performance and costs. For the purpose of the impact assessments related to the “Climate Target Plan” and the “Fit for 55” policy package, these assumptions have been updated based on a rigorous literature review carried out by external consultants in collaboration with the JRC 108 . Continuing the approach adopted in the long-term strategy in 2018, the Commission consulted on the technology assumption with stakeholders in 2019. In particular, the technology database of the PRIMES and PRIMES-TREMOVE models (together with GAINS, GLOBIOM, and CAPRI) benefited from a dedicated consultation workshop held on 11th November 2019. EU Member States representatives also had the opportunity to comment on the costs elements during a workshop held on 25th November 2019. The updated technology assumptions are published together with the EU Reference Scenario 2020 109 . The same assumptions have been used in the context of this impact assessment.

Policies in the Baseline scenario

Building on the EU Reference scenario 2020, the baseline scenario for this impact assessment has been designed to include the initiatives of the ‘Fit for 55’ package 110 .

The baseline scenario assumes no further EU level intervention beyond the current AI Directive and the continuation of the application of the IMO Casualty Code as it is. Smaller fishing vessels would continue to be outside the scope of the Directive and therefore Member States would adopt different approaches to similar accidents. Member States with inadequate financial, technical or operational resources to be able to report on and investigate accidents in a timely and efficient manner would have to make choices as to what accident to investigate and how.

As previously stated, the role of EMSA in the implementation of the Directive is central. EMSA operates the EMCIP database, provides training to investigators on the request of AIBs and assists in the management of the PCF. The Commission has launched an impact assessment on the possible review of EMSA founding Regulation. 111 However, the outcome of this impact assessment cannot be prejudged and thus the baseline scenario does not account for changes in the EMSA founding Regulation.

Baseline scenario results

The COVID-19 pandemic had a major impact on global shipping, affecting all its segments from passenger ships to container ships and oil tankers. In the baseline scenario, international maritime freight transport activity (intra and extra-EU) is projected to be 21% lower in 2020 relative to 2015. From 2021 onwards however it is projected to start recovering and grow strongly by 2025 and beyond (i.e. 19% growth for 2015-2030 and 48% for 2015-2050), due to the rising demand for primary resources and container shipping. Relative to 2019, this is equivalent to 8% increase in transport activity by 2030 and 33% growth by 2050.

The number of port calls for 2025-2050 is projected to grow at a lower rate than transport activity, following similar evolution over the historical period 112 . This reflects the fact that transport activity is also driven by other factors such as the increase in the size of vessels over time, and of the distance travelled. In the baseline scenario the number of port calls is projected to go up by 14% by 2030 relative to 2015 and by 36% by 2050 (equivalent to 6% growth by 2030 relative to 2019 and 26% increase by 2050), following the recovery from the COVID-19 pandemic.

Driven by the increase in the transport activity and the number of vessels, the number of marine casualties is projected to increase over time in the baseline scenario. Considering only the scope of the AI Directive (i.e. excluding fishing vessels below 15 meters) the number of casualties is projected to increase by 14% by 2030 relative to 2019 and by 45% by 2050 without further EU level action. At the same time, the degree of severity of marine casualties is projected to decrease, leading to a decrease in the number of vessels lost (by 5% for 2019-2030 and for 2019-2050) and a relative stabilisation of the number of fatalities by 2050 (11% decrease for 2019-2030 and 3% increase for 2019-2050).

Table 18: Projected number of marine casualties, vessels lost, fatalities and injuries in the baseline scenario in EU27

Source: Ecorys (2022), Impact assessment support study

Accounting also for the fishing vessels below 15 meters, which are currently outside the scope of AI Directive, the number of marine casualties is projected to increase by 9% by 2030 relative to 2019 and by 27% by 2050 in the baseline scenario.

The projected numbers of marine casualties, vessels lost, fatalities and injuries in the baseline scenario, by vessel type, are provided in Table 19. They are derived based on the projected growth in the number of vessels and the occurrence ratios. For all vessels types, except for fishing vessels, the occurrence ratios 113 are assumed to remain constant over time at their 2019 levels, drawing on information for the historical period from EMCIP. This is also the case for the ratios between vessels lost, fatalities, injuries and the vessel fleet.

For fishing vessels the occurrence ratio is projected to slightly increase over time (from 5.6% in 2019 to 6.2% in 2030 and 7.5% in 2050), drawing on historical developments but assuming a slower pace than in the past. At the same time, the ratios between vessels lost, fatalities, injuries and the vessel fleet are assumed to remain constant over time, at their 2019 levels. This is acknowledging the past trends observed, showing that while the number of marine casualties involving fishing vessels is increasing over time the degree of severity of the casualties has been slightly decreasing 114 . The number of fishing vessels is projected to reduce over time (by 6% between 2019 and 2030 and by 16% during 2019-2050), in line with historical developments 115 and also taking into account the moderate increase in the number of catches projected in the future (0.4% per year) 116 . This is the reason why the number of vessels lost, fatalities and injuries involving fishing vessels is projected to decrease over time.

Thus, the lower growth in the number of marine casualties for fishing vessels below 15 meters (6% for 2019-2030 and 15% for 2019-2050) relative to other types of vessels is explained by the projected reduction in the number of fishing vessels by 2050, while the marine casualties’ rate would still increase over time. The degree of severity of marine casualties including fishing vessels below 15 meters is projected to decrease over time but 210 fatalities are still projected in 2030 and 203 in 2050 in the baseline scenario, while 141 vessels are projected to be lost in 2030 and 129 in 2050. By 2050, the number of fatalities involving fishing vessels below 15 meters would be 135 out of total 203, representing 67% of the total number of fatalities while the vessels lost 85% of the total (110 lost fishing vessels below 15 meters out of 129 total).

Table 19: Projected numbers of marine casualties, vessels lost, fatalities and injuries in the baseline scenario by vessel type (EU27)