Policy approaches to mitigate in-use methane emissions from natural gas use as a marine fuel

Unregulated in-use methane emissions (or methane slip) can reduce or even eliminate the overall climate benefits of using liquefied natural gas as a marine fuel. We conduct critical review and expert interviews to identify methane slip mitigation measures, and then identify and evaluate potential policy instruments that could incentivize their uptake while considering the shipping sector’s climate targets. We find that regulatory instruments are expected to perform the best across a range of criteria when they are at the global level, include methane on a CO2-equivalent and lifecycle basis, promote polycentric approaches to climate governance, and allow flexibility in how the industry incorporates decarbonization measures. Market-based approaches and informational governance policies complement regulatory instruments by improving cost-effectiveness and increasing the availability of relevant information on emissions mitigation. The urgency and scale of shipping climate targets underscore the need for policy approaches that support planning for long-term decarbonization pathways and that can avoid locking into fossil-carbon intensive systems.


Introduction
Maritime transportation is the backbone of global trade and economy, with more than 80% (by volume) of the world's trade transported by ships [1]. However, marine shipping is also a major source of air pollutants and greenhouse gases (GHGs), owing to its heavy fuel oil (HFO) use during operations and maintenance. Its contribution to annual global carbon dioxide (CO 2 ), sulfur oxides (SO x ), and nitrogen oxides (NO x ) emissions are 3%, 13%, and 15%, respectively [2,3]. To manage these emissions, the global shipping regulator, the International Maritime Organization (IMO), currently regulates NO x emissions using engine standards, SO x emissions through a sulfur-in-fuel limit, and GHG emissions via a CO 2 intensity (CO 2 emissions per transport work) index [3].
Liquefied natural gas (LNG) is an alternative fuel that provides significant air quality benefits compared to HFO: NO x emissions are reduced by 80%-94% (and up to 98% if also using aftertreatment systems such as selective catalytic reduction), and SO x emissions are almost eliminated [4][5][6][7][8]. It has lower CO 2 emissions per unit of energy than HFO, thereby meeting all the IMO emissions regulations [9,10]. The use of LNG as a marine fuel has grown substantially in the shipping industry, by 28%-30% from 2012 to 2018 [3], as its use is linked to a burgeoning and profitable LNG trade, while also supporting compliance with regulatory targets [11]. Factors contributing to this uptake include: a favorable regulatory environment at the IMO level; the US shale gas boom and subsequent shift in natural gas geopolitics; and technological advancements in LNG infrastructure, liquefaction, and dual-fuel gas engine concept [7,10,[12][13][14]. Compared to the air quality and economic benefits, the benefits of fossil-based LNG towards decarbonization are more limited and less clear [5,9]. Fugitive methane (CH 4 ) emissions throughout the LNG value chain (natural gas extraction, processing, liquefaction, storage, transportation, regasification, and distribution), venting emissions (intentional methane emissions during operations and maintenance when using LNG as a marine fuel), and in-use engine methane emissions threaten GHG reductions required to meet the IMO 2050 climate target-at least 50% reduction in GHG emissions from international shipping compared to 2008 levels [9,15,16]. In-use engine methane emissions or methane slip happens when unburnt methane emits from marine gas engines into the atmosphere [6]. In this study, we focus on fossil-based LNG; and unless otherwise stated explicitly, LNG here means fossil-based LNG.
The IMO estimates that methane emissions from shipping have increased by 151%-155% (59-148 thousand tonnes (vessel-based); 55-140 thousand tonnes (voyage-based)) during the period from 2012 to 2018 [3]. These changes reflect the increased uptake of LNG as a marine fuel, where the low-pressure dual-fuel (LPDF) and lean-burn spark-ignited engines, with relatively high methane slip levels (1.5%-16% of throughput), dominate the market share (47% of all LNG-fueled engines in 2018) [3,16,17]. Despite the significant increase, current regulatory frameworks are not equipped to address methane emissions from LNG as a marine fuel, particularly during the use phase. Much research has focused on upstream components of this value chain, however, engine methane slip during fuel combustion onboard ships are a growing area of concern. Methane slip may constitute up to 99.8% of the total methane emissions of an LNG-fueled voyage [6,16]. The downstream combustion component, including methane slip has been estimated to constitute nearly 70%-91% of lifecycle GHG emissions in the LNG value chain under certain conditions [4,9,10,[18][19][20]. Here, we conduct a policy analysis to address this regulatory gap, identifying and evaluating policy approaches to mitigate methane slip from LNG-fueled ships and contribute to absolute GHG emissions reductions in the marine transportation sector. We consider a range of policy evaluative criteria, including emissions reduction effectiveness, stakeholder acceptability, cost-effectiveness, amongst others. For potential future (and on order) shipping assets, including LNG-fueled shipping, we intend for our study to better inform policy decision making. We draw on a critical review of academic and gray literature, and semi-structured interviews with multi-sectoral experts on LNG in shipping.
Policies are formed and influenced by a diverse set of stakeholders at all stages of the policy cycle (from goal setting to policy formulation, adoption, implementation, and evaluation), and at multiple governance levels (local, regional, national, and global) [21]. The shipping industry reflects this heterogeneous reality, underscoring the need to understand stakeholder perspectives and opinions to evaluate policy interventions. In this policy analysis study, we add to the maritime transportation decarbonization literature by emphasizing this multistakeholder perspective. Our study contributes to the larger body of work on energy transition policy, marine policy, and shipping decarbonization. We build on the previous policy studies applicable to shipping decarbonization [22][23][24], and environmental and economic assessment studies of LNG-as-a-marine fuel in the following ways: (a) We first identify and characterize downstream methane slip mitigation measures (technological, operational, and hybrid) applicable to new-builds and retrofittable for existing LNG-fueled ships. (b) We then identify and evaluate potential policy instruments that can incentivize uptake of these measures to mitigate methane slip from LNG-fueled shipping while considering the broader IMO 2030 and 2050 climate targets. (c) Subsequently, we discuss key attributes of the policy approaches identified as best suited for GHG emissions mitigation from LNG-fueled shipping while considering the upstream LNG value chain, the IMO and the Intergovernmental Panel on Climate Change (IPCC) climate targets.
The remainder of this article is structured as follows: section 2 provides a review of types of policy alternatives that can be used to mitigate emissions, and discusses the factors that led to the rise of LNG in the shipping industry and the frameworks used to assess lifecycle GHG emissions of LNG as a marine fuel. Section 3 provides details on the methods used to collect and analyze data. In section 4, we present the results from our analysis of methane slip mitigation measures and potential policy instruments, supported by further discussions on our findings. Section 5 provides the main conclusions and policy implications of this work.

Environmental governance of maritime transportation
The international shipping industry is a complex and heterogenous entity, regulated by the IMO [25]. IMO regulations are first ratified by two-thirds of the member states present during an IMO voting session [26]. The member states of the IMO then enforce these regulations individually [26]. Additional details on international shipping emissions regulations are provided in the supplemental information (SI).
Several types of policy alternatives are used in global and heterogenous transportation sectors such as marine and aviation to mitigate GHG emissions: regulatory instruments, market-based mechanisms, and informational governance policies. Regulatory instruments play a critical role in these sectors because they impose certain direct regulations to reduce emissions, demonstrating dependability and uniformity [27]. The governments at all levels (local, regional, national) may intend to go above and beyond the requirements of a global regulator such as the IMO, and introduce additional or complementary policies within their jurisdictional purview [25]. Such multipronged approaches are termed as polycentric governance [25]. Even with such an approach, regulatory instruments can be inflexible, low in cost-effectiveness, and distort the level playing field in a sector such as international shipping with complex operations [25,27]. Shipping entities such as port authorities and shipping companies may also establish voluntary agreements to reduce emissions and the impact of shipping on local air quality in coastal areas [28].
In contrast, market-based mechanisms provide flexibility and promote cost-effectiveness [27,29]. In market-based systems, the negative externalities or emissions are priced, or financial incentives are provided to influence desirable behavior from the relevant stakeholders in a certain market [29]. However, ex-post analyses have shown that to date, the impact of carbon pricing on GHG emissions reductions has been limited [30]. Informational governance strategies can be used to provide support to both regulatory and market-based policy instruments [31]. They are flexible and in most cases, voluntary and non-coercive in nature [27]. Informational governance policies promote transparency and disclosure of environmental data to encourage good environmental behavior [31]. However, they are not as effective when implemented in isolation [27].

Initial use of natural gas in the maritime transportation sector
Liquefied natural gas carriers (LNGCs) were the first users of natural gas as a marine fuel. These are liquid tankers used for long-distance transportation of LNG. Inadequate insulation of the tankers contributes to natural evaporation of LNG, known as boil-off gas (BOG). To maintain tank pressure, BOG is removed from the tankers and used as a marine fuel for LNGCs propulsion. LNGCs mainly use steam boiler propulsion systems that simultaneously combust both BOG and HFO, the dominant marine fuel in the shipping industry in recent decades. Initially, BOG was so significant due to poor insulation that it contributed 100% of LNGC fuel consumption. However, as insulation technology developed over the years, natural evaporation of LNG reduced considerably, which demanded the use of either forced BOG or HFO for fuel consumption [13,14].
Considering the relatively low efficiency of steam boiler propulsion and the requirement of forced BOG, other fuel-efficient propulsion systems such as dual-fuel engines and slow-speed diesel propulsions have now gained traction due to their economic advantages and represent around 61% of the current LNGC fleet [13,32]. The dual-fuel engine concept allows a ship to operate interchangeably on natural gas (with diesel pilot fuel) or 100% diesel fuel [6].

Growth of natural gas as a fuel in the maritime transportation sector
The US shale revolution has driven a significant shift in the geopolitics of natural gas as an energy source and global supply and demand. The shale gas revolution and development of LNG import/export terminals and bunkering infrastructure around the globe have made LNG a genuinely global and tradable commodity, driving down LNG prices and creating an LNG supply glut. The political and economic relations among the importing, exporting, and transit nations govern the new geopolitics of global gas and LNG markets. Landlocked Europe, which is traditionally reliant on the gas-producing nation of Russia for its energy needs through long-term pipeline contracts, now has the opportunity to diversify its gas imports through US-based LNG. Such diversification decreases Europe's reliance on Russia, strengthens its energy security, and avoids potential commercial or political concessions. Changes in gas geopolitics also impact transit nations (e.g. Ukraine) critical for gas pipeline infrastructure, where favorable transit fees and gas prices might be negotiated with importing and exporting nations [12,33].
LNG geopolitics sustained by the shale boom, liquefaction technology and LNG infrastructure advancements, and LNG demand, are primarily facilitated by LNGCs [7,12,13]. Shipping companies and shipowners are strategically choosing to diversify into the LNG business to grasp the profitable trend of the shale boom and global LNG trade, which is expected to increase by 21% by 2025 compared to 2019 [11,34]. This diversification is achieved in two ways: by increasing the LNGC fleet (including small-scale LNG bunkering and delivery vessels) and by introducing LNG-fueled ship(s) into the existing fleet [11]. Currently, there are 572 active LNGCs (130 on order) and 266 active LNG-fueled ships (442 on order) [32,35]. Many of these orders, especially for LNG-fueled ships, were initially driven by the Norwegian NO x fund that incentivizes low-NO x solutions, including LNG-based propulsion [36]. However, in 2018, the Norwegian government removed the carbon tax exemption on LNG as a marine fuel for Norwegian-registered vessels, making the fuel there 25% more expensive than conventional marine diesel fuels [37]. Globally, while making LNG-related investments, shipping companies consider low-risk profitability in the short-term and long-term environmental regulatory targets [11]. Since LNG as a marine fuel meets all the current and upcoming NO x and SO x emissions standards, and CO 2 -intensity-based energy efficiency design index (EEDI) regulations mandated by the IMO, it makes for a solid business proposition [10,11].

Inconsistent frameworks to assess lifecycle GHG emissions from using natural gas as a marine fuel
The primary constituent of LNG is methane. LNG may also contain propane, other highly flammable saturated hydrocarbons, and traces of noble gases; unlike coal and oil, which are composed of more complex compounds with higher carbon/hydrogen ratio and nitrogen and sulfur contents [38]. Combustion of LNG for energy consumption purposes, therefore, provides clear air quality benefits, significantly reducing NO x emissions and almost eliminating SO x and particulate matter emissions [9,38]. Theoretically, LNG could provide climate benefits since it has lower CO 2 emissions per unit of energy compared to HFO; however, in practice, its climate benefits are less clear because of fugitive and in-use methane emissions across the entire LNG value chain [9,10,15]. This ambiguity is further exacerbated by the different frameworks applied to assess the lifecycle GHG emissions of LNG as a marine fuel. Reports on methane assessments from upstream oil and gas operations through atmospheric measurements suggest that methane emissions are 60% greater than the regulatory emissions inventory estimates in the US and Canada [15,39]. These discrepancies highlight methane leakages from fugitive and unplanned processes that are underreported or underestimated, and minimum thresholds below which reporting is not required [15,39].
There is no consistent framework applied to judge GHG benefits from using LNG as a marine fuel. Depending on the level and type of parameters used in the framework of the lifecycle GHG assessments, studies may recommend LNG as an alternative marine fuel for shipping decarbonization because they conclude that LNG generates considerable GHG benefits and can serve as a transition fuel [19,[40][41][42], or they may recommend augmenting LNG-fueled ships with other shipping decarbonization measures to generate GHG benefits and meet long-term decarbonization targets [4,14,17,43], while others caution against any LNG uptake in the shipping industry due to concerns related to natural gas-based carbon lock-in and GHG disbenefits from using LNG [9,10]. The parameters in the framework vary based on the length of the upstream LNG value chain comprising of natural gas extraction, production, processing, and liquefaction, LNG transportation and distribution (through several bunkering options such as tank to vessel, truck to vessel, and vessel to vessel) processes, and regasification; the extent of fugitive methane emissions during these upstream operations; the magnitude of methane slip from marine gas engines onboard ships, which depend on the engine type and operational practices; the type of engine emissions testing methodologies applied such as the standard IMO E2/E3 emissions test cycle that may or may not represent real-world engine load conditions; and the type of climate impacts metrics used to assess lifecycle GHG emissions such as global warming potential (GWP) on a 100-year timeline, GWP100, a 20-year timeline, GWP20, technology warming potential (TWP), and global temperature change potential (GTP).
Methane is a potent GHG with a GWP of 29.8 for a 100-year time horizon and 82.5 for a 20-year time horizon [44]. A recent study reported that LNG as a marine fuel may contribute to decarbonization only when considering the best-case LNG supply chain, reducing GHG emissions by up to 28% compared to HFO on a GWP100 metric [17]. On a GWP20 basis, some reports have estimated that LNG provides no climate benefits and instead increases CO 2 e emissions compared to HFO, while others estimate GHG reductions of up to 9% [4,5,9]. The TWP metric, which considers cumulative radiative forcing of a technology over time, suggests that in the best-case scenarios, LNG as a marine fuel in high-pressure dual-fuel (HPDF) engines can produce GHG benefits within 30 years after transitioning from conventional marine diesel fuels [45]. For LPDF engines, it may take up to 190 years to generate any GHG benefits [45]. This period required for reaching climate parity is consistent with other studies comparing transportation systems fueled by natural gas with gasoline or diesel, and underscores the importance of multiple climate tipping points (MCTPs) [45][46][47]. MCTPs assessments emphasize the need to consider short-term climate impact metrics such as GWP20 or dynamic metrics [48], considering the urgency of the climate crisis and the dangers of passing tipping points triggering large-scale and abrupt irreversible changes in the climate system [46]. The parties to the Paris Climate Agreement are required to use the 100-year GWP to report their net GHG emissions, and determine their nationally determined contributions; lifecycle assessments of value-chain impacts using this metric therefore allow for consistency with national commitments. However, the IPCC's Sixth Assessment Report suggests including additional metrics and reporting to better represent the implications of global warming from current and future GHG emissions, particularly when considering the impacts of short-lived climate pollutants (SLCPs) such as methane [44]. These include separately reporting the multiple GHGs; using a cumulative CO 2 equivalent metric, GWP * (an alternative to GWP that makes use of the changes in the emissions rate of the SLCPs to determine the warming equivalent); and a combined GTP [49,50].
While a few studies have assessed methane slip during individual engine loads and actual vessel operation, most studies use weighted average methane slip value for their lifecycle GHG emissions assessment of LNG as a marine fuel [4,5,9,10,18,19,51,52]. Depending on the type and model of the engine, the weighted average methane slip value varies from 6.9 gCH 4 kWh −1 to 0.2 gCH 4 kWh −1 [6,9]. Weighted average methane slip is calculated using the standard IMO E2/E3 emissions test cycle methodology, which gives more weightings (70%-84%) to engine load of 75% and above, where methane slip is much lower (<5 gCH 4 kWh −1 ), and fewer weightings (16%-30%) to engine load of 50% and below, where methane slip is significantly higher (12-178 gCH 4 kWh −1 ) [5,41]. Real-world data from 2012 to 2018 show that majority of the ship feet (containers, oil tankers, bulk carriers) were, on average, operating at between 30% and 60% engine load [3], indicating potential underestimation of methane slip in several lifecycle GHG emissions studies of LNG as a marine fuel. In fact, using real-world vessel operation data during harbor service, lifecycle emissions analysis of LNG shows an increase in GHG emissions by 38% compared to marine diesel fuels, even on a GWP100 basis, and by 109% on a GWP20 basis [5].
The contributing factors (shale gas boom, LNG geopolitics, demand and infrastructure development, favorable regulatory landscape) that led to the rise of LNG in the shipping industry, the inconsistent framework applied to judge its climate benefits as a marine fuel, and the subsequent and ongoing increase in unregulated methane emissions from the industry underscore the need for policy development that addresses the use of LNG as a marine fuel.

Methodology
In this study, we use a combination of critical literature review, and semi-structured interviews with key stakeholders with subject matter expertise in LNG as a marine fuel. The interview data complement secondary data from the literature review in identifying methane slip mitigation measures, potential policy instruments for mitigating methane slip, policy evaluation criteria and their assessment scales, and finally, evaluation of the policy instruments. We triangulate across multiple data collection methods and sources (expert interviews, critical literature review) to support and validate our research findings, address biases, and provide additional perspectives that cannot be obtained if we relied solely on a single methodology [53]. Further description on the methods is given below.

Critical literature review
We conduct a literature review to identify and analyze methane slip mitigation measures applicable for gas engines onboard marine vessels. We also carry out a critical review of current and previous policy alternatives from several jurisdictions [54][55][56][57], and the road [58], aviation [59], and maritime [60][61][62][63] transportation sectors, on emissions mitigation to determine potential policy instruments for mitigating methane slip from LNG-fueled shipping. A description of the policy instruments under evaluation is summarized in table 1. We group them into three categories: market-based measures (MBMs), regulatory instruments, and informational governance policies.
The evaluative criteria we use in this policy study are those conventionally applied to assess energy and environmental policies [64,65]. For the assessment scales of the evaluative criteria (effectiveness, cost-effectiveness), we use natural attributes where the data available is directly used to measure the objective of the policy alternatives [66]. When no natural attributes exist for the assessing criteria (administrative feasibility, technical feasibility, stakeholder acceptability, policy robustness), we construct comprehensive, operational, and lucid scales that measure the relevant objectives [66]. Subsequently, we assign a low/medium/high indicator to each level of the scale to facilitate the later analysis of the policy instruments in this study. The evaluative criteria selected here are defined below: Effectiveness refers to the ability of a policy instrument to achieve the intended objectives, which are in accordance and relative to the IMO 2050 GHG target of at least 50% reduction in GHG emissions (in terms of carbon dioxide equivalent (CO 2 e)) compared to 2008 levels per the IMO initial GHG strategy [3]. Cost-effectiveness is defined here as the cost incurred per ton of CO 2 e reduced. The cost-effectiveness scale is based on the level of marginal abatement costs of shipping decarbonization measures and proposed maritime carbon levies [22,67]. Administrative feasibility refers to the operational burden on relevant administrators [65], including the IMO, governments at various levels, port authorities, and the private sector, due to bureaucracy, excessive information and monitoring requirements. Technical feasibility here refers to the range of monitoring technology required to enable a given policy instrument, including verified data collection of methane and other GHG emissions measurement, monitoring, and vessel operational parameters such as engine load, fuel consumption, and cargo carried. Stakeholder acceptability is defined as the level of support from the relevant stakeholders in the marine shipping industry. Policy robustness refers to the ability to consistently achieve GHG emissions reductions, under a range of political, economic, and technological future conditions [65], such as the 2019 coronavirus disease (COVID-19) pandemic, which may impact the total emissions reduction potential for the shipping industry [68]. Table 2 describes the assessment scales of the evaluative criteria created to measure the performance of each policy alternative.

Semi-structured expert interviews
We conduct semi-structured expert interviews that complement the literature review of methane slip mitigation measures and the evaluation of policy instruments. An interview guide that lists the topics of interest and questions covered during the interviews is provided in the SI. The interview participants contributed to identifying methane slip mitigation measures, the potential policy instruments for methane slip mitigation from LNG-fueled shipping, and determining the most suitable evaluative criteria and their respective assessment scales. We contacted 34 experts in the marine shipping industry, out of which 20 agreed to be interviewed (table 3). The final number of interview participants was determined by thematic saturation-which is based on grounded theory and is the point at which additional interview data no longer contributes new insights due to repetitive themes [69]-and the breadth of the entire LNG value chain, to reflect the heterogeneity of the shipping industry. The experts were selected based on their specialized knowledge of LNG as a marine fuel, and in-depth understanding of the maritime transportation industry, shipping decarbonization, and related policies. This expertise was evidenced by their published works on LNG as a marine fuel, lifecycle GHG emissions studies, and shipping decarbonization in peer-reviewed journals and reports such as the IMO GHG studies; participation in IMO meetings such as the IMO Marine Environment Protection Committee meetings, and the intersessional meetings of the Working Group on Reduction of GHG Emissions from Ships (ISWG-GHG); affiliations with organizations that have consultative status with IMO; and experiences working in the LNG value chain in leadership and  management roles. Considering the focused criteria for expert recruitment, there are a relatively small number of people in this space. Therefore, in addition to our judgment, we consulted our research collaborators and conducted snowball sampling to recruit experts for this study. The semi-structured interviews were conducted online over nine months, from August 2020 to May 2021. The interview participants were ensured confidentiality. They represent the entire breadth of the LNG value chain and the diverse perspectives and opinions of key stakeholders (natural gas producers, LNG bunkering suppliers, port authorities, ship owners and operators, charterers, marine gas engine manufacturers, non-governmental organizations (NGOs), regulators, marine shipping and gas engine researchers, trade associations, marine engineers, and naval architects) with interests in LNG as a marine fuel and with a global representation in the marine shipping industry. The interviews lasted 60-90 min and elicited key themes regarding LNG as a viable alternative marine fuel, its role in shipping decarbonization and meeting IMO climate targets, methane slip mitigation strategies and any applicable regulations, potential policy alternatives for methane slip mitigation from LNG-fueled shipping, criteria for judging such policies, and measurement scales to support the criteria. Interviewed experts also provided several relevant information through email. We recorded, transcribed and manually coded the interviews before using the interview data for our analysis.

Measures to reduce methane slip onboard ships
Methane slip mitigation measures are a niche category in the shipping sector, and relevant information is sparse. Here, we identify and characterize methane slip mitigation measures into three broad themes by triangulating expert interviews with a critical literature review. They are technological, operational, and hybrid measures. These measures are summarized in table 4, and briefly reviewed in the text, with additional technical details provided in the SI.

Technological measures
We define technological measures as the measures that require addition or upgrade of methane-slip-reducing technologies in the propulsion system of a marine vessel. These include: engine exhaust emissions aftertreatment; exhaust gas recirculation (EGR) systems; HPDF engines; and optimization of combustion chamber design, engine valve overlap, the combustion process within the engine combustion chamber, and the air-fuel ratio and mixture quality. Studies have shown that aftertreatment systems, such as palladium-based methane oxidation catalysts, can achieve high levels of methane slip reduction, i.e. up to 80% [70,77]. However, interviewed experts and past studies suggest that maintaining the ideal performance conditions for these catalysts can be challenging during ship operations [70]. EGR systems and negative engine valve overlap mechanisms work by reusing the unburnt methane for the next engine combustion cycle, which would have otherwise been emitted through the exhaust valve [71,72]. The reuse of such exhaust gas streams reduces both NO x emissions and methane slip from marine dual-fuel engines [71]. An alternative approach to the aftertreatment and EGR systems is optimization of the air-fuel ratio and mixture quality. For instance, to comply with the IMO NO x regulations, marine dual-fuel engines (particularly lean-burn spark-ignited and low-pressure types) in gas mode are typically operated under lean conditions with high air-fuel ratios, leading to lower peak combustion temperatures and corresponding lower NO x emissions [6]. However, this also contributes to unacceptable methane slip levels, particularly at low engine loads [6]. As explained by one interviewed shipowner-operator, 'at low loads, they are basically methane pumps'. Several measures have been identified that can improve combustion efficiency and reduce methane slip; from fuel dilution with ethane gas, combustion chamber with reduced crevice volume, to engine compression ratio reduction [72][73][74][75][76]. More details on these measures are provided in the SI.
Dual-fuel engine technologies based on the diesel cycle also contribute to reduced methane slip [6]. However, because of their economic advantages and ease of handling, the LPDF and lean-burn spark-ignited engines, which are based on the Otto cycle, have had increased uptake in LNG-fueled shipping, comprising 47% of all LNG-fueled engines in 2018, rising from 22% in 2012 (number of LNG-fueled engines in 2012: 423, in 2018: 678); despite their high levels of methane slip [3,5,9,13]. HPDF engines run on the diesel cycle concept, where only air is compressed [6]. When the pilot diesel fuel (for ignition) and natural gas are injected into the compressed air, instant combustion occurs as the high compression ratio in a diesel cycle allows the compressed air to reach the ignition temperature of the fuel [6]. This type of combustion process leads to negligible methane slip [6]. HPDF engines comprised 8% of all LNG-fueled marine engines in 2018 [3]. The main caveats and safety concerns with such a system are that it requires a high-pressure natural gas supply of up to 350 bar [6], larger floor height inside a ship compared to other engine types, and the addition of NO x abatement technologies such as exhaust recirculation and selective catalytic reduction to comply with the NO x Tier III regulations [7].

Operational measures
We define operational measures as those that involve modifications in the existing operations and the standard operating procedures of a marine propulsion system to reduce methane slip. These include engine cylinder deactivation and avoidance of running engines at low loads. During cylinder deactivation, particularly at low loads, some of the cylinders in a multi-cylinder engine are not fueled, while the remaining cylinders are operated at higher loads to generate the required power at the given low load. This deactivation process leads to higher low-load exhaust temperature and combustion efficiency, reducing methane slip [5,51]. An interviewed shipowner-operator recommended cylinder deactivation as a methane slip mitigation measure but also cautioned that operational patterns of a ship, safety, and engine life would also need to be considered. Ideally, it is safer to avoid running marine gas engines at low loads, if possible, where methane slip is the highest, instead of deactivating engine cylinders. However, such operational profiles might not be possible for short-sea shipping such as short-sea and ferry operations and coastal shipping, which typically operate at low engine loads.

Hybrid measures
We define hybrid measures as the measures that combine various alternative fuels and technologies, methane-slip-reducing technologies, shipping decarbonization measures, and operational changes in a ship to maximize methane slip mitigation at the lowest possible voyage costs, such as a hybrid battery-LNG propulsion system. In this system, during low engine loads, initial engine ramp rates, or berthing, power for ship propulsion or auxiliary needs is driven by battery technology, thus avoiding the region of high methane slip. Dual-fuel engines in gas or diesel mode are used for a ship's power requirements at medium to high loads. In addition to batteries as an energy storage medium, cold-ironing or shore-to-ship power is another option suggested for use during ship berthing. However, this may not always be feasible for LNG ships due to BOG management, either through reliquefaction (where cold-ironing may be utilized) or consumption [13]. Previous assessments of hybrid systems including battery-diesel-electric, dual-fuel engines in gas or diesel mode with energy from waste heat recovery systems, found such systems to reduce NO x , SO x , CO 2 emissions from ship exhaust and meet the EEDI limits [14,43,78,79]

. Several interviewed stakeholders prefer hybrid measures and consider that 'hybrid [battery-] LNG should be the de facto installation [for LNG-fueled ships]' .
Since it provides operational flexibility and technological neutrality in reducing methane slip, both for ship retrofits and new-builds. Several studies proposed to combine LNG with more efficient engines and shipping technologies, and decarbonization measures to reduce GHG emissions and meet the IMO climate targets [10,17,79,80]. Experts from the environmental NGO and regulatory sectors even suggest skipping LNG as a marine fuel due to its methane slip issues, and using a combination of very-low-sulfur fuel oils, other alternative marine fuels (including but not limited to renewable natural gas (RNG), biofuels, hydrogen, ammonia) and technologies (including but not limited to fuel cell, battery, wind-assisted propulsion), and shipping decarbonization measures (including but not limited to slow steaming, voyage optimization, hull design optimization) for overall GHG emissions reductions, while complying with the existing air quality regulations.
A description of all the methane slip mitigation measures identified from the literature review and elicited from the expert interviews is summarized in table 4 with their references.
The aforementioned methane slip mitigation measures apply to new vessels, and also existing vessels through retrofitting. Either option may be preferred depending on voyage planning [81], the ship type and size, the ship's lifetime, marginal abatement cost of the mitigation measure, subsequent fuel (natural gas) savings, payback period, and regulatory compliance timeline. Since methane emissions from the shipping industry are currently unregulated, uptake of methane slip mitigation measures is crucial, especially for existing and new-build LNG-fueled vessels, and considering the low-cost and abundant nature of natural gas [82]. Holistic policy approaches are required to incentivize the uptake of mitigation measures. We discuss the projected outcomes of such potential policy instruments in the next section.

Projected outcomes of the policy instruments
In this section, we discuss the projected outcomes of potential policy instruments for mitigating methane slip from LNG-fueled shipping, while considering the long-term decarbonization targets. We engaged with key stakeholders in this policy analysis study who may have an active or passive influence on the policy-making processes. We triangulate their diverse perspectives and opinions with empirical evidence from the literature to assess the range of outcomes of the evaluated policy instruments. A summary of the projected outcomes against the evaluative criteria is organized in table 5.

Market-based measures
MBMs to reduce shipping GHG emissions have been considered at the IMO since the early 2000s [23]. However, no such approach currently exists at the global level [23]. A maritime carbon tax, as a levy on carbon emissions or bunker fuel, is under ongoing discussion at the IMO, with proposed carbon prices ranging from 2 USD per tonne of bunker fuel (equivalent to ∼0.65 USD per tonne of CO 2 e) to 300 USD per tonne of CO 2 e emissions [22]. Currently, Norway has a carbon (CO 2 ) tax on LNG as a marine fuel to incentivize the production of other alternative energy sources such as hydrogen, battery, and battery-hybrid systems [37]. At the regional level, the European Union (EU) in 2021 proposed to include shipping into the European Union Emissions Trading System (EU ETS) from 2024 [83]. The updated EU ETS includes CO 2 emissions from ships (above 5000 gross tonnage and independent of the flag) calling at an EU port for voyages within the EU, during berthing in an EU port, and 50% of the emissions from voyages that start or end outside the EU [84]. As LNG uptake as a marine fuel increases, such MBM-related discussions must consider and include methane emissions. Interviewed experts view the inclusion of methane as CO 2 e in the MBMs as one of the key factors in mitigating methane slip. In late 2022, the Council of the EU and the European Parliament agreed to include marine shipping emissions within the scope of the EU ETS, with provisions to include non-CO 2 emissions, such as CH 4 and nitrous oxide (N 2 O), from 2026 [85]. LNG use might be further incentivized if methane emissions are not considered in MBMs since it produces less CO 2 emissions than conventional marine diesel fuels.
Stakeholders expect MBMs such as ETS and carbon tax to have low-to-moderate effectiveness, depending on several factors such as the price of carbon, the scope and coverage of the MBMs, and the economic conditions that may considerably impact the policy robustness of the MBMs, contributing to uncertainties. A myriad of uncertainties exists in maritime transportation, which could be addressed explicitly in future policy studies [86,87]. The effectiveness of maritime MBMs also depends on whether shipping methane emissions are included in the MBMs, and the revenues generated from the MBMs are recycled into the shipping industry. The stakeholder acceptability of the MBMs is moderate. Several interviewed experts doubt whether MBMs could promote environmentally sustainable behavior in the industry due to the split incentive problem, environmental governance issues, carbon leakage and potential low carbon pricing, which may insulate long-term financial investments such as ships from making any decarbonization-related decisions. The split incentive problem occurs when shipowners who generally bear the capital costs of decarbonization or methane slip mitigation technologies do not benefit from such measures [88]. Instead, the charterers who pay for the fuel costs under certain charter types reap the fuel savings achieved through such technologies [88]. Shipowners may recoup their investments through higher charter rates, but it is not always feasible, limiting the incentive to adopt methane slip mitigation technologies. In the case of long-term contracts, though, charterers might be willing to bear the capital costs of such upgrades with the expectation that the investment will be paid back within the contract period [89].
Stakeholders suggest that a higher carbon price is needed for any considerable GHG emissions reductions, while maintaining that a higher price will receive strong pushback from the industry at the IMO level. Interviews with representatives from the shipbroking and industrial NGO sectors reveal that higher carbon prices may be possible where part or entire portion of the levy is passed on to the end consumers, making the taxing fairer for the shipping industry-considering that the global supply chain drives the industry. Nonetheless, it is essential that the 'polluter pays' principle is at the core of potential MBMs such as maritime carbon tax and ETS. MBMs are considered cost-effective, where the lowest marginal abatement costs for the industry vary significantly depending on the ship type and size, whether a vessel is new-build or requires retrofitting, and voyage planning. The monitoring, reporting and verification (MRV) requirements of emissions trading systems, especially if methane is included, make them more complicated to implement than carbon taxes, considering methane emissions vary significantly with changes in operational practices of LNG-fueled ships. In the case of a maritime carbon tax, fuel consumption data may be used as a proxy for CO 2 e emissions. Stakeholders concur that such requirements and lack of trustworthiness within the shipping industry towards potential bunker levy administrators such as bunker suppliers contribute to increased operational burden on the IMO, leading to low-to-moderate administrative feasibility. Consortium benchmarking, a voluntary version of a cap-and-trade system, has similar requirements as an ETS. However, interviewed experts do not expect this policy instrument to contribute to any considerable methane emissions reductions due to its voluntary nature. Such voluntary programs, which non-state actors often conduct, are also known as private standards [90]. They result in rewarding only the good performers and low industrial legitimacy, leading to one stakeholder suggesting to include the Poseidon Principles framework for responsible ship finance and the ship rating scheme Clean Cargo Working Group (CCWG) with such standards to influence shipowners to participate. Several port authorities manage emissions-and environmental-specific programs that give a discount in port dues to shipowners or charterers to encourage good environmental behavior and incentivize air pollutant emissions reductions [91,92], especially near port communities that are disproportionately affected by air pollution from shipping activities [93]. Port authorities conduct such discount schemes using private standards and ship environmental programs such as CCWG, Environmental Ship Index, RightShip GHG rating, environmental class notations from ship classification societies, regulatory standards such as the EEDI, or through direct reporting by ships about their use of alternative fuels such as LNG and methane slip mitigation technologies. Interviewed experts do not expect such port schemes to be effective, mainly because the incentive is not large enough compared to the total operational costs during a voyage to change any behavior within the shipping industry. Despite the low effectiveness, the industry still prefers discounted port dues schemes, especially the consumer-facing sectors such as cruise ships that intend to have a green image, and container shipping lines with clients, i.e. cargo owners who want to green their supply chains. A coalition of leading ports around the globe providing such incentives would help increase their effectiveness, where a vessel accrues discounted port fees at multiple stops in its voyage. Discounted port dues programs are typically revenue-neutral, and therefore, high emitters pay higher dues. Stakeholders caution that discounted port dues programs should not discourage early adopters of LNG as a marine fuel, considering the air quality benefits of the fuel, subsequent compliance with SO x and NO x regulations, but potentially high methane slip near the port areas where LNG-fueled ships are either berthing or maneuvering at low engine loads. Instead, GHG emissions from the entire voyage should be considered in determining the discount level, which can be facilitated by private environmental ship programs, regulatory standards, or class notations updated with CO 2 e basis that includes methane; engine certifications that indicate methane slip values, or proof of onboard application of methane slip mitigation technologies such as diesel cycle gas engines. Though discounted port fees programs are no panacea to reducing methane slip, interviewed experts from the natural gas industry highlight the critical and proactive role of the port authorities in promoting the use of LNG in the shipping industry, through forming coalitions with governments, LNG producers, shipowners, bunkering operators, and building joint ventures to invest in LNG infrastructure [91,94].
Potential government subsidies evaluated here are to incentivize the uptake of methane slip mitigation measures for locally-registered LNG-fueled ships. Most stakeholders prefer subsidies to be provided in the ship design process. Such financial incentives tend to have low effectiveness overall, considering the high capital costs required for LNG-fueled ships and the limited GHG emissions reduction potential of LNG compared to conventional marine diesel fuels. Due to the split incentive problem, subsidies may be given to shipowners as they are liable to their investments or charterers depending on ships' operational profile and charter arrangement. Subsidies may be used to cover capital costs for methane slip mitigation technologies or incentivize seafarers to operate ships using best practices in reducing in-use methane emissions when possible. Experts from the natural gas industry suggest that subsidies should not be provided for shipping LNG infrastructure in the long term as it may distort the market. While those representing environmental NGOs, trade associations, and regulators entirely oppose subsidies for LNG ships-considering it is based on the fossil fuel system-and instead support financial incentives for zero-emissions shipping and related research. Experts from industrial NGOs and engine manufacturing sectors prefer subsidies for further research and development in marine gas engines to reduce methane slip. Subsidies may also take the form of reimbursements based on assessments in achieving specific methane slip-related targets through technological, operational, or hybrid measures.

Regulatory instruments
The current regulatory landscape at the IMO level includes the EEDI, and the recently adopted energy efficiency existing ship index (EEXI) and annual operational carbon intensity indicator (CII) standards for GHG emissions mitigation. More details on these regulations are provided in the SI. The IMO standards, however, are based on CO 2 intensity and do not include methane [95,96]. They consider technological measures for energy efficiency while excluding ship operational profiles that significantly affect methane slip levels. The EU Stage V regulation for marine engines in EU inland waterways (regulation 2016/1628) includes a total-hydrocarbon limit, which considers non-methane hydrocarbons and methane emissions [97]. According to an interviewed expert from the environmental NGO sector, a methane slip standard explicitly for marine gas engines is under consideration for the 'Blue Angel' certification, an ecolabel managed by the federal government of Germany. The engine methane slip limit for this certification is expected to be less than 0.6 gCH 4 kWh −1 . Such a limit could mean that the current lean-burn spark-ignited and LPDF engines are ineligible for the ecolabel, considering their high methane slip levels unless they are augmented with methane slip mitigation measures.
Interviewed experts concur that regulatory standards are the most effective in shipping decarbonization and limiting methane slip from LNG-fueled shipping, primarily if the standards are implemented at the IMO level. Stakeholders strongly favor engine standards for methane slip, where some suggest that low-pressure engines should not be in production due to their unacceptable level of unburnt methane emissions. In contrast, those representing engine manufacturers and industrial NGOs caution that such a standard might be challenging to implement and that the evaluation of marine gas engines should be on a CO 2 e basis. Engine manufacturers expect that the next-generation low-pressure engines will have lower methane slip of 1-2 gCH 4 kWh −1 . Though, there are currently no known incentives to reduce methane slip at the required rate. In their interviews, most stakeholders from the shipping industry point out that the onus is on the engine manufacturers to reduce methane slip to an acceptable level. Similar methane slip concerns are observed in freight trucks running on natural gas [36,98,99]. Engine standards alone are expected to be of low effectiveness due to LNG's limitation in reducing overall shipping GHG emissions, which include not only methane but also CO 2 and N 2 O. However, if complemented and augmented with other shipping decarbonization measures, the effectiveness could be considerably higher. For instance, approaches that target shipping GHG emissions reductions more broadly would enable such augmentation.
Amendment of the CO 2 -based IMO indexes such as the EEDI, EEXI, CII and the Energy Efficiency Operational Indicator (EEOI), and Ship Energy Efficiency Management Plan (SEEMP) to include GHGs such as methane, in terms of CO 2 e, would allow the shipping industry flexibility to incorporate suitable decarbonization measures, including other alternative fuels and energy systems. Currently, only the CO 2 emissions reduction capabilities of the decarbonization measures can be incorporated into the EEDI formula and the other IMO-based intensity indexes [100]. In addition, there are no provisions in the formula that consider emissions offsetting through methods that do not involve marine shipping. Considering the sensitivity of operational parameters such as engine load to methane slip levels, several stakeholders recommend a mandated EEOI but acknowledge the complications in continuous monitoring of fugitive methane emissions during ship operations. Therefore, others suggest using a fuel proxy where engine testbed or model data under realistic engine load conditions are considered to calculate methane slip values and applied as CO 2 e in the overall EEDI formula. Current EEDI test measures are conducted at 75% engine load where methane slip is much lower, using ship design values instead of operational performance data [101].
Phasing out fossil-based LNG as a marine fuel could be high in effectiveness depending on the timing of the ban, total uptake of the gaseous fuel by that time, and level of absorption of other decarbonization measures and alternative fuels in the shipping industry. Interviewed experts have contrasting viewpoints regarding the lifespan of LNG as a marine fuel-considering that it is a fossil fuel. In general, they predict that LNG will peak by 2030 or 2040 and then phase out or remain in decreasing uptake trend by 2050. The predictions are in line with the McKinsey 2050 global gas outlook study that forecasted LNG demand to peak by 2037 [102]. The 2021 World Bank report, in its 'temporary role for LNG' scenario, suggests LNG will peak by 2030 and then decline rapidly in a timeframe that would contribute to stranded assets, which could be around 850 billion USD in 2030 [103,104]. The study expects LNG uptake as a marine fuel to be limited; however, it also suggests that the fuel might play an important role in shipping decarbonization as a feedstock to low/net-zero carbon marine fuels such as blue hydrogen and ammonia [103]. Taking a more progressive approach to climate actions, stakeholders representing environmental NGOs and regulators suggest that LNG needs to be phased out as a marine fuel within the next 15-20 years, considering the IMO 2050 GHG goals and the risks of a potential natural gas-based carbon lock-in. In contrast, those from the natural gas and shipbroking industry, industrial NGOs, and port authorities suggest that LNG uptake might last for several decades, barring some technological breakthroughs, with a 50-100 year horizon and peaking by 2050. They consider LNG to be a vital fuel for the IMO 2050 GHG strategy, and that it is a 'transition' fuel to low/zero/net-zero carbon energy systems for ships, mainly green ammonia, methanol, hydrogen, and RNG. The renewable version of LNG, RNG or upgraded biogas, and synthetic renewable LNG (e-LNG) have greater potential to reduce GHGs and generate climate benefits than fossil LNG or when blended with fossil LNG; however, fugitive methane emissions across the natural gas chain, fuel feedstock availability, and costs are of major concern, when considering decarbonizing deep sea shipping [52,[105][106][107].

Informational governance policies
Informational governance policies use disclosure of environmental information as a tool to encourage environmentally sustainable behavior of stakeholders in the shipping industry, such as savings on fuel consumption and reducing emissions [31]. The EU MRV and the IMO data collection system (DCS) are existing mandated policy instruments that rely on information disclosure to foster CO 2 emissions reductions. The EU MRV requires direct reporting of CO 2 emissions from ships above 5000 gross tonnage within, arriving at, or departing from ports under the jurisdiction of EU member states [108]. In late 2022, the EU Council and Parliament agreed to include non-CO 2 emissions in the EU MRV regulation from 2024 [85]. While the IMO DCS provides global coverage and requires reporting of fuel consumption, a proxy for CO 2 emissions, for ships above 5000 gross tonnage [31]. Data compiled from the EU MRV serve as the basis for including maritime transport in the EU ETS [108]. The EU adopted the MRV system for shipping in 2015, a year before the IMO DCS, given the insufficient progress in GHG emissions mitigation at the IMO level [31]. Neither of these systems currently includes methane. Considering the rise of LNG as a marine fuel in the shipping industry, interviewed experts suggest including methane in the emissions reporting schemes either through direct continuous emissions monitoring (such as EU MRV) or through fuel proxy (such as IMO DCS) with third-party certified ship operational data such as engine load.
According to several stakeholders, continuous monitoring of in-use methane emissions is unrealistic and leads to more complexity with little benefits and increased costs. In contrast, a shipowner-operator suggests that though continuous monitoring of methane has complexities, they are essential for increased transparency, and the monitoring costs are likely to be minute compared to the total capital expenditure of a vessel. Such differences in opinions about costs allude to the importance of managing profit margins in the shipping industry, with one natural gas bunker supplier arguing, 'as a shipowner, I am incentivized to spend the least amount on a ship and charter it for the highest amount into the market' . Representatives from environmental NGOs, port authorities, and the shipbroking industry concur with the importance of continuous emissions monitoring in transparency and also in terms of regulatory enforcement and evasion avoidance. They anticipate the shipping industry to be moving towards such a monitoring system, similar to the current onboard monitoring requirements for scrubber systems in HFO-fueled ships. Interviews with engine-emissions researchers reveal that electrochemical and optical sensors are potential options for hydrocarbon measurement in ships, including methane. Electrochemical sensors provide a low-cost alternative to shipowners [109], and researchers expect this version to scale up in the future. Satellite measurements to detect methane [110], and atmospheric measurements of methane emissions from a remote station near shipping lanes, using cavity ring-down spectrometer or unmanned aerial vehicles such as drones are other possibilities [111].
Stakeholders strongly favor increasing awareness of best practices in managing methane slip among shipowners, operators, and seafarers. Best practices include adopting relevant mitigation measures, managing operational practices (such as slow steaming, intentional gas discharge during maintenance) and parameters (such as engine load) that significantly impact methane slip levels, and avoiding high methane slip conditions when possible to do so safely and within operational constraints. Improved understanding of methane slip leads to industrial collaboration among stakeholders and innovation, such as hybrid battery-LNG systems, and implicitly encourages stakeholders to save fuel. Shipowners and shipping companies, especially consumer-facing sectors and those that operate primarily near the emissions control areas such as roll-on/roll-off (RoRo) vessels and handysize tankers, demonstrate and promote their environmental sustainability performance through third-party ship environmental certifications and class notations to conform to client demands, increase awareness and mold public perceptions, generate interest from potential clients, or complement the Poseidon Principles to secure future funding. Though there are currently no private certifications and class notations that explicitly consider methane, there is potential for such voluntary measures to contribute to methane slip mitigation as part of a universal shipping GHG emissions index that recognizes both upstream and downstream emissions instead of competing certifications. Interviewed experts caution that most environmental ship certifications and class notations are a marketing tool and do not necessarily contribute to absolute emissions mitigation, and instead focus on energy efficiency to improve carbon intensity as demand for shipping increases. Such certification programs lack transparency, ambition to go beyond regulatory requirements, and have data reliability issues [90].
Stakeholders do not expect mere disclosures of emissions and environmental performance to be effective but recognize their flexibility in terms of technical and administrative feasibility. While split incentives, environmental governance issues and challenges are problematic, informational governance policies promote capacity building and increase the availability of relevant information regarding GHG and methane slip mitigation in an industry characterized by asymmetric and imperfect information [88,89], thereby alleviating some of the barriers to MBMs.

Linkages to upstream emissions and carbon lock-in
Shipping companies, including LNGC businesses, are driven by global consumption and supply chain. Depending on the business model, cargo owners (e.g. oil and gas majors in tanker shipping) may greatly influence ships' operational profiles, safety standards, and fuel consumption, thereby impacting the shipping sector's environmental performance [112,113]. The myriad actors influencing the already complex marine transportation sector, in addition to the 'flags of convenience' model prevailing in the industry [113,114], highlight the importance of a polycentric approach to environmental and climate governance in shipping that considers the entire LNG value chain including upstream emissions; where local (e.g. Port of Vancouver EcoAction Program, Norwegian NO x Fund, Maritime Singapore Green Initiative) and regional (e.g. EU ETS and EU MRV) policies complement global policies (e.g. IMO DCS, EEDI, EEXI, CII, GreenVoyage2050), forging ways to optimize policy design, implementation, and enforcement. Interviewed experts suggest that upstream low carbon fuel standards (LCFS) complement the policy instruments evaluated in this study. LCFS mandates reduction in the carbon intensity of transportation fuels sold in a jurisdiction [115]. The carbon intensity (in gCO 2 e/MJ) in LCFS considers the full lifecycle of the fuel, and credits are earned based on the current limits, which can either be traded or banked for future years [116]. Other complementary policy instruments for the upstream LNG value chain include renewable fuel standards [117] and methane fees for natural gas producers [118].
Most of the projected LNG demand stems from coal-to-gas switching in China and other fast-growing Asia Pacific regions, where natural gas is considered a suitable alternative to coal, owing to its air quality benefits and lower CO 2 emissions per unit of energy [119,120]. However, as fossil LNG infrastructure replaces coal, it can also replace low carbon/net-zero/renewable fuels and energy systems that would otherwise provide long-term climate benefits. One interviewed expert argued that heavy investments in LNG infrastructure are for the long-term (more than 30 years), creating a new carbon-based energy system supported by natural gas and locking-in, not just the shipping industry but also society, considering the fuel's use as an energy source for the buildings, electricity, and other transportation sectors. These concerns about lock-in understand technological (shale boom, liquefaction technology, LNG infrastructure, dual-fuel engine technology, LNG tanker shipping) and institutional (LNG geopolitics, environmental governance in shipping) coevolution, driven by path-dependent increasing returns to scale, as potentially contributing to a potential LNG-based carbon lock-in in the shipping industry. If climate targets identified in the Paris Agreement and by the IMO are to be achieved [121,122], large-scale decarbonization of all systems will be required, including shipping. In light of these targets, opportunities to avoid reliance on a system locked into carbon-intensive fossil fuel-based energy systems are emphasized by many environmental NGOs, regulators, and researchers.

Limitations of the study
Our focused criteria for expert recruitment meant that we follow a small-n approach, and emphasize depth of engagement with expert informants over breadth. Despite considering thematic saturation and the diversity of perspectives across the LNG value chain, the small-n approach means that there may be limitations in terms of representativeness and generalizability of the results for each stakeholder group. Further, considering that the environmental governance landscape in maritime transportation is rapidly evolving, this study may have limitations in terms of evolving information availability and capturing experts' evolving opinions and perspectives. The perspectives presented here therefore represent a selection of perspectives from 2020 to 2021, that nevertheless illustrate the diversity of views and reasoning on how methane slip might be addressed through policy interventions, considering the broader shipping climate targets. Combining the interview data with critical review of gray (i.e. outside of commercial publishing, including from governments, think tanks, civil society, and industry) and academic literature allowed for more breadth in the policy analysis, drawing on both the strengths of in-depth stakeholder engagement and the wider scope of documentary evidence.

Conclusions and policy implications
Overall, based on our critical review and semi-structured expert interviews, regulatory instruments (amendment of IMO's CO 2 intensity indexes and energy efficiency strategies to include methane, engine methane slip standards) at the IMO level are recommended to mitigate methane emissions from LNG-fueled shipping and contribute to the climate targets of the IMO and IPCC, because of their potential to achieve high effectiveness, strong industrial legitimacy, and global purview. Potential regulatory measures need to avoid 'all-in-one' approaches and instead be flexible to recognize the heterogeneity of the shipping industry. A flexible approach to regulatory instruments would support and be neutral to the uptake of the different operational, technological, and hybrid methane slip mitigation and shipping decarbonization measures, including alternative energy systems, both for new-builds and existing vessels. A flexible approach could support planning for long-term decarbonization pathways. Both MBMs and informational governance policies complement regulatory instruments. MBMs improve cost-effectiveness further, and informational governance policies help increase the availability of relevant information on emissions mitigation and measures in the shipping industry, thus reducing some of the barriers to MBMs, such as asymmetric and imperfect information and split incentives problem. Given the complexity of environmental governance in maritime transport and the uncertainty in shipping decarbonization pathways, polycentric governance and adoption of policy mixes of regulatory instruments, MBMs, and informational governance policies, are highly preferred. Such policy mixes enable taking advantage of the strengths of the individual policy instruments while overcoming their deficiencies when used in isolation. The policy approaches here are better operationalized by real-world verified measurements of methane emissions, which are severely lacking for both upstream LNG value chain and marine gas engines, and require urgent attention.
As the use of LNG as a marine fuel increases in the shipping industry, it is critical that the evaluated policy instruments include methane, on a CO 2 e and lifecycle basis, and ratchet up over time to more stringent absolute GHG emissions reduction targets. The potential for climate tipping points and the limited time and carbon budget available for 1.5 • C warming, has increased attention to reducing emissions of short-lived GHGs such as methane [123]. The attributes of the policy approaches aforementioned can better inform policy decision making, as the IMO and regional regulators consider regulating methane slip in the industry. It is essential that the commercial drivers of the industry, including voyage planning, global consumption and supply chain, are also under consideration for potential policy interactions and opportunities for reducing methane emissions in the entire LNG value chain.

Data availability statement
The data cannot be made publicly available upon publication because they contain sensitive personal information. The data that support the findings of this study are available upon reasonable request from the authors.