Brazil’s IMO proposal for ISWG-GHG 22: how compliance, surplus trading, and the facility could work

The negotiations around the IMO Net-Zero Framework have moved into an interesting phase. Following the decision not to adopt the draft amendments to MARPOL Annex VI in October 2025, IMO Member States have spent much of 2026 considering how the framework could be adjusted without reopening the entire package negotiated at MEPC 83. Brazil's contribution ISWG-GHG 22/2/12, submitted ahead of the September 2026 intersessional working group meeting (ISWG-GHG 22), is particularly relevant in this context because it attempts to address several of the concerns that have prevented agreement while preserving most of the underlying compliance architecture.

The IMO ISWG-GHG 22 Brazil proposal's starting point is relatively conservative. Its stated preference remains adoption of the IMO Net-Zero Framework as negotiated, with changes only where they are considered necessary to achieve agreement. The alternatives contained in its submission are therefore presented as compromise options. The global GHG Fuel Intensity standard would remain, as would ship-level compliance balances, Surplus Units, Remedial Units and financial support for zero- and near-zero fuels. Changes concern the initial stringency of the targets, the presentation of the economic element, and the institutional structure through which compliance contributions would be collected and redistributed.

Similar to the initial proposal, Brazil wants the framework to operate explicitly as a flexible compliance system. A ship that reaches the required GHG intensity through its own fuels or technologies would comply directly and would not make a compliance payment. A ship that performs even better could create transferable surplus. Only where physical performance is insufficient would the company need to balance the remaining deficit through compliance units.

At the same time, the first year of FuelEU Maritime gives the industry a useful real-world example of how such flexibility can develop once a compliance market becomes operational. FuelEU has shown that surplus markets can facilitate compliance very effectively, but it has also exposed a more complicated question: the cheapest way of producing compliance is not necessarily the fuel pathway that policymakers ultimately may want to see scale in the long term. The current European experience therefore provides a useful lens through which to assess both the strengths and the remaining weaknesses of the Brazilian proposal.

The framework would remain built around GHG Intensity

The basic technical architecture remains unchanged. Ships would be measured against an annual GHG Intensity calculated on a well-to-wake basis and expressed in gCO₂eq/MJ. The draft framework uses a reference value of 93.3 gCO₂eq/MJ, representing the average GHG intensity of international shipping in 2008, and annual targets are expressed as percentage reductions against this reference value (just as in the original IMO NZF proposal).

An important feature of the IMO framework is that it contains two separate GFI thresholds. The Base Target establishes the less stringent minimum trajectory, while the Direct Compliance Target requires a significantly deeper reduction. The position of a vessel relative to these two thresholds determines whether it generates surplus, incurs a Tier 1 deficit or incurs both Tier 1 and Tier 2 deficits. This two-level design is central to the economic architecture because it prevents the market for transferable surplus from becoming the only mechanism through which the regulatory signal is transmitted.

Brazil proposes retaining this structure but making the choices available to shipping companies more explicit in the legal text. Its proposed wording for Regulation 32 would state that a ship can comply either directly through fuels, technologies or energy sources, or by balancing its compliance deficit through the cancellation of banked or traded surplus units and Remedial Units (RU) obtained through contributions to the IMO mechanism. Draft Regulation 36 would similarly clarify that each ship may voluntarily choose between direct compliance and balancing its deficit.

While nuanced, this is politically important because Brazil also proposes removing references to “GHG emissions pricing”. Its argument is that the system should not be understood as imposing a general levy on every tonne of shipping emissions. Instead, the economic element provides an alternative compliance route where direct physical compliance is unavailable or commercially unattractive. A company would therefore compare the cost of changing its fuel or technology with the cost of acquiring the units necessary to balance the resulting deficit.

A softer introduction, followed by a more demanding trajectory

Brazil also proposes a slower initial reduction trajectory. The submission shifts the implementation dates by one year following the delay in adoption and defers several of the originally negotiated reduction levels. Brazil explains that the adjustment is intended to respond to concerns about the availability of compliant fuels, the maturity of supply chains and the economic impact of introducing the mechanism too quickly.

Year	Base Target reduction	Direct Compliance Target reduction
2029	3.0%	3.0%
2030	4.0%	4.0%
2031	6.0%	19.0%
2032	8.0%	21.0%
2033	12.4%	25.4%
2034	16.8%	29.8%
2035	21.2%	34.2%
2036	25.6%	38.6%

The unusual feature is the treatment of the first two years. In 2029 and 2030, the Base Target and Direct Compliance Target would be identical. Brazil explicitly proposes this transitional arrangement to reduce the volume of revenue collected during the early implementation period and to give IMO additional time to establish the institutional systems required for larger-scale collection and redistribution. From 2031 onward, however, the two targets separate sharply and the full two-tier architecture becomes relevant.

Brazil does not propose reducing the long-term ambition indefinitely. At the other end of the trajectory, the submission would replace the previously envisaged 65% Base Target reduction in 2040 with a 70% reduction in 2041.

Surplus units would turn over-compliance into a tradable asset

A ship whose GFI is lower than the Direct Compliance Target would generate a positive compliance balance and become eligible for surplus units. The units correspond to the positive balance in tonnes of CO₂-equivalent. They can subsequently be transferred to another ship, banked for future use or voluntarily cancelled as a mitigation contribution.

This creates a genuine global compliance market somewhat similar to the already known FuelEU Maritime. Vessels with access to low-GHG fuels or technologies could generate compliance beyond their own requirements and sell that value to vessels facing higher physical decarbonisation costs. The economic effect is similar in principle to FuelEU Maritime: reductions take place where they are cheapest, while ships with fewer available options can remain compliant through the market rather than being forced into an immediate fuel switch.

There is nevertheless an important distinction between simply meeting the Base Target and becoming a supplier of surplus. Under the IMO architecture, the positive balance is calculated against the more stringent Direct Compliance Target. A vessel can therefore satisfy the Base Target and still generate no surplus units at all. Once the two targets separate from 2031, the threshold for becoming a surplus generator becomes materially more demanding than the minimum GFI trajectory.

That structure means that not every marginal improvement in vessel GFI automatically creates tradable supply. If the Direct Compliance Target requires a 19% reduction while the Base Target requires only 6%, as proposed for 2031, a vessel operating somewhere between those levels remains compliant with the Base Target but has not reached direct compliance. The resulting difference is dealt with through the Tier 1 mechanism rather than being ignored.

Tier 1 and Tier 2 are the key to understanding the economic mechanism

The distinction between Tier 1 and Tier 2 is one of the most important elements of the draft framework as well as the Brazilian proposal and can easily be overlooked. A vessel operating above the Direct Compliance Target but still below the Base Target has a Tier 1 compliance deficit. If its GFI also exceeds the Base Target, the additional emissions above that second threshold constitute a Tier 2 deficit.

Under the current draft text, these two deficits cannot be balanced in the same way. Tier 1 must be covered by Tier 1 RUs, while transferred or banked surplus units can be used to address Tier 2 deficits alongside Tier 2 RUs. Surplus trading can therefore reduce the cost of non-compliance with the Base Target, but it cannot substitute for the entire economic signal between the Direct Compliance and Base thresholds.

This gives the IMO model a different character from a single-threshold system. Once the two targets diverge, a large supply of cheap surplus cannot simply satisfy the full compliance requirement of every higher-emitting vessel. The Tier 1 band remains linked to the remedial mechanism and consequently provides a source of demand for RUs even where the transferable surplus market is well supplied.

Note that Brazil’s proposed introductory period temporarily weakens that distinction because the two targets are identical in 2029 and 2030. There is effectively no Tier 1 band during those years. That makes the initial market particularly relevant when comparing the IMO design with the experience under FuelEU Maritime, because the first two years could allow a larger amount of relatively inexpensive surplus to enter the system before the more restrictive two-tier structure takes effect.

Remedial Units provide certainty when physical compliance or surplus is unavailable

Unlike a surplus unit, which originates from another vessel’s physical over-performance, an RU is obtained through a financial contribution to the IMO mechanism. Brazil argues that RUs should guarantee an adequate supply of compliance units and make the cost of meeting the regulation sufficiently predictable for shipping companies.

The original draft framework circulated in 2025 set initial prices of USD 100/tCO₂eq for Tier 1 RUs and USD 380/tCO₂eq for Tier 2 RUs for reporting periods 2028 to 2030, with IMO subsequently required to establish the mechanism for prices from 2031 onward. Brazil’s contribution does not itself replace those prices, although its proposal to shift the implementation timetable means that the corresponding years would need to be adjusted in the final legal text.

For the transferable market, the Tier 2 RU price is to be particularly important because it provides an alternative to purchasing another ship’s surplus units. Provided RUs remain readily available, a buyer has little reason to pay materially more for a Surplus Unit than the equivalent remedial compliance would cost. The RU price can therefore provide an effective upper reference point for surplus.

The reverse does not apply. The availability of a USD 380 remedial option does not mean that surplus units will trade anywhere close to USD 380. If surplus is abundant, its market price can fall towards the marginal cost of creating the next unit. The development of FuelEU over the past year provides a useful demonstration of exactly this effect.

FuelEU already shows how quickly a compliance market can find the cheapest solution

The first FuelEU Maritime compliance cycle provides an unusually relevant comparison because it combines a GHG-intensity target with banking, pooling and a financial fallback. Data from a forthcoming European Commission report shows that more than 90% of ships used pooling for the 2025 compliance year, while fewer than 2% used borrowing. Around 7% made no flexibility request, approximately half of that group complied directly and the remainder paid the FuelEU penalty.

From the perspective of regulatory compliance, this is an important result. A very large number of companies were able to match positive and negative compliance balances rather than relying on the statutory penalty, indicating that a market for over-compliance can become functional quickly. It also illustrates why flexibility is commercially attractive: a company does not necessarily need to decarbonise every vessel independently when lower-cost reductions can be sourced elsewhere in the market.

The fuel outcome is more nuanced. The same Commission data indicates that biofuel represented more than two-thirds of the renewable and low-carbon energy used for FuelEU compliance, while its modelling suggests that existing LNG use may have provided roughly one-third of the overall emissions reduction required in 2025. More capital-intensive options such as e-fuels, onshore power, and wind-assisted propulsion remained a small part of the compliance mix.

That demonstrates something fundamental about market-based fuel standards: when different pathways create the same fungible compliance value, the market tends to favour the cheapest available source first.

Current FuelEU pricing makes this effect particularly visible

The development of the FuelEU surplus market during 2026 provides an even clearer illustration. BetterSea’s transaction data shows that weighted end-of-month surplus prices reached €83.13/tCO₂e at the end of August 2026, with the weighted monthly average for August at €108.89/tCO₂e. This is materially below the values around €190-200/tCO₂e observed across parts of the first compliance cycle.

The fall in surplus prices has occurred alongside a fundamental change in the cost of generating new compliance. The Strait of Hormuz shock sharply increased conventional bunker prices while the increase in some biofuel prices was less pronounced, compressing the normal premium between fossil and renewable fuels. Several industry sources frequently reported during 2026 that FuelEU abatement costs considering biofuel and MGO fell below zero, meaning that under the relevant assumptions the biofuel switch is already economically advantageous before any positive value is assigned to the resulting FuelEU surplus.

This changes the composition of the surplus market. A company that previously purchased surplus because using alternative fuel was more expensive may now find that replacing MGO with an eligible biofuel is attractive on the underlying fuel and EU ETS economics alone. When more companies reach that conclusion, new supply enters the surplus market at a very low marginal generation cost and places downward pressure on the price.

The consequence is particularly challenging for fuels with higher production costs. A company using a more advanced alternative fuel may require a relatively high compliance value to compensate for the physical fuel premium, but it sells the same unit of FuelEU over-compliance as a company whose biofuel switch happened to generate surplus at almost no additional cost. The surplus market does not distinguish between those generation costs. Unless there is an additional fuel-specific support mechanism, the cheapest marginal supplier increasingly determines the market value received by every generator.

LNG and banking add a second source of surplus pressure

FuelEU also shows how existing fleet characteristics can contribute surplus without a new renewable-fuel investment. Fossil LNG can remain below the FuelEU GHG-intensity target depending on engine technology and methane slip.

The commercial treatment of this surplus differs between companies. A charterer controlling an LNG-fuelled vessel for a limited period may have little interest in preserving the compliance position of that ship ten years later and can instead monetise the current surplus through pooling. An owner and operator may reach the opposite conclusion and bank the positive balance to hedge against the vessel’s future FuelEU exposure. One can assume that a meaningful amount of surplus was retained rather than immediately monetised, although the publicly available first-year figures reviewed here do not provide an aggregate figure for the amount of surplus that was banked.

This matters because FuelEU banking has no expiry date. Positive compliance balances can be carried forward successively and banked surplus does not expire. A ship can therefore accumulate positive balances during relatively easy early compliance years and use that stock against considerably more stringent requirements later.

From a market perspective, this creates an additional layer of uncertainty. The surplus currently offered for pooling represents only part of the potential supply. Another part may sit in company accounts as banked compliance and return to the market or be consumed internally many years later. As the market develops, the size of this accumulated stock becomes increasingly relevant to any assessment of future supply, pricing and the marginal incentive for new alternative-fuel consumption.

The IMO architecture limits indefinite surplus accumulation

This is one area where the IMO draft takes a materially different (better) approach. A surplus unit would be valid for only two calendar years after the calendar year in which it was issued. If the unit remains unused when that period expires, it is cancelled as a mitigation contribution.

The rule does not prevent short-term oversupply. If market conditions make physical over-compliance unusually cheap, the fleet could still generate a large amount of surplus and depress prices for the following reporting periods. What the expiry mechanism prevents is the creation of an indefinitely growing stock of legacy compliance capable of competing with future physical decarbonisation many years later.

Combined with the rising Direct Compliance Target, that should make the market more responsive to current fuel economics. Surplus generated cheaply under an early and relatively weak target will eventually disappear, while new units will have to be produced against increasingly demanding GFI thresholds. This does not guarantee a high surplus unit price, but it reduces the risk that accumulated early over-compliance suppresses the market indefinitely.

The Direct Compliance Target provides a second safeguard. Once the Base and Direct Compliance targets diverge, a ship must achieve the deeper threshold before it generates transferable surplus. A vessel that simply meets the Base Target may avoid Tier 2 exposure, but it still has to deal with the Tier 1 deficit. This limits the extent to which inexpensive surplus from another ship can substitute for progressively improving the fleet’s GHG intensity.

The introductory period could nevertheless resemble FuelEU

Brazil’s proposed treatment of 2029 and 2030 makes the first two years more exposed to the same market dynamics now visible under FuelEU. With the Base and Direct Compliance targets both set at 3% in 2029 and 4% in 2030, there would effectively be no Tier 1 band. A vessel beating that single threshold could generate surplus, while a vessel above it would face the corresponding deficit without the stronger separation introduced from 2031 onward.

Whether this produces large quantities of low-cost surplus will depend heavily on the final well-to-wake GHG factors. Some LNG configurations could potentially perform favourably against an early target, while unexpected market movements could again (or still) make biofuel-based compliance unusually inexpensive. In that situation, a significant amount of surplus could be generated early on.

The crucial difference from FuelEU is that the effect would be temporary. Units cannot remain in circulation indefinitely. Still one may argue that an already delayed IMO NZF should not further soften the market.

ZNZ rewards address a problem that surplus trading cannot solve on its own

The FuelEU experience also explains why the separate zero- and near-zero, or ZNZ, reward mechanism is important. If all over-compliance is fungible, the ordinary surplus price will tend towards the cost of the cheapest available generator. That price may be sufficient to support a biodiesel blend or an existing LNG vessel but likely entirely insufficient to bridge the cost gap for e-methanol, e-ammonia, renewable hydrogen or other more capital-intensive pathways.

The draft IMO framework therefore separates ordinary compliance value from additional support for qualifying ZNZ fuels and technologies. Under the existing draft, the initial ZNZ threshold is set at no more than 19 gCO₂eq/MJ and becomes more stringent later in the trajectory. Brazil does not propose removing this mechanism. On the contrary, the operational model illustrated in its submission retains a dedicated ZNZ Rewards Window, through which payments could be made following evidence that qualifying zero- or near-zero fuels, technologies or energy sources had actually been used.

Surplus units would reward performance relative to the GFI compliance standard and allow efficient trading between ships. ZNZ rewards would have a different function: providing an additional economic signal to the technologies capable of delivering the much deeper reductions required later in the transition.

The effectiveness of this approach will depend on the detailed reward methodology and the available funding. A dedicated window does not in itself guarantee that an expensive fuel will become competitive, and Brazil recognises that substantial guideline work remains. It specifically identifies the calculation of GFI and ZNZ rewards, Facility governance and auditing, and sustainable-fuel certification as priority workstreams requiring greater clarity.

A lighter Facility rather than a large IMO Fund

The institutional design of the economic mechanism represents the second major area of compromise. Some IMO Member States have objected to establishing a substantial fund administered directly through IMO, while others have argued that revenue collection and redistribution are necessary if the framework is to support ZNZ deployment and an equitable transition. Brazil’s proposal attempts to preserve those functions without requiring IMO to perform all of the financial-management activities itself.

Instead of an IMO Net-Zero Fund, Brazil proposes exploring an IMO Facility. IMO would establish the governing rules, eligibility criteria and policy direction, but accredited international, regional or national financial institutions could undertake much of the practical implementation. Brazil describes this as a more streamlined mechanism focused on collecting and deploying revenue through established implementing institutions, which would report back to IMO and operate according to its guidelines. The Facility could also be temporary and include a sunset clause. Financial contributions associated with deficit balancing, together with possible voluntary donations, would flow into the Facility. One channel would finance ZNZ rewards, while a separate Just & Equitable Transition Window would finance eligible projects through international, regional or national financial institutions.

The proposal seeks to reduce the need for IMO itself to become a large-scale asset manager, project evaluator and financing institution, addressing one of the major counter arguments posed against the previous draft.

A cheap surplus market could also affect the amount of money reaching the Facility

There is, however, an interaction between surplus prices and Facility funding that deserves attention. Where a Tier 2 deficit can be balanced either with a traded surplus unit or a Tier 2 RU, companies will naturally choose the less expensive route. If physical over-compliance becomes cheap and the market is well supplied with surplus units, fewer Tier 2 RUs may be purchased and less revenue may flow into the Facility. A similar behavorial pattern was observed in FuelEU Maritime when comparing the uptake of pooling vs paying the penalty, as discussed above.

The two-tier structure partly limits that effect. Once the Base and Direct Compliance Targets separate, Tier 1 deficits continue to require RUs and therefore create a more stable source of financial contributions irrespective of Tier 2 surplus availability. Brazil also proposes permitting voluntary donations.

Nevertheless, the relationship between RU demand and the size of the ZNZ reward pool will need careful consideration when the guidelines are developed. If the framework relies on compliance contributions to support expensive zero- and near-zero fuels, it must remain capable of funding those rewards even when ordinary Tier 2 compliance becomes inexpensive. FuelEU demonstrates that market conditions can change the marginal cost of compliance much faster than long-term fuel-production projects can adjust.

Better transparency could become another advantage of the IMO model

A further lesson from FuelEU concerns market visibility. FuelEU allows unlimited banking, but the publicly available first-year compliance data provides much better information on the use of pooling and borrowing than on the aggregate stock of positive balances retained for the future. For market participants trying to assess supply and demand several years ahead, the amount of surplus already accumulated is an important variable.

The IMO GFI Registry should provide a stronger basis for transparency and is capable of the same as surplus unit issuance, transfer, banking, cancellation and expiry would all take place within a central system. The existing draft also foresees reporting to MEPC on the operation of the compliance mechanism.

The usefulness of that information will ultimately depend on the disclosure rules adopted in the supporting guidelines. If the market can see aggregate surplus issuance, transfers, vintage distribution and upcoming expiries, companies will be better able to assess the balance between existing supply and future demand. A compliance market in which a substantial quantity of banked surplus is invisible to participants would be considerably harder to predict efficiently (as currently seen in FuelEU Maritime), despite neccessary to drive long-term investments.

What FuelEU tells us about the Brazilian proposal

The first FuelEU cycle offers a useful way of assessing the design. It demonstrates that flexible compliance can work, that companies can create an active market for positive compliance balances and that only a comparatively small proportion of vessels need to rely on the expensive statutory fallback when sufficient surplus is available. At the same time, it shows that this success does not automatically translate into broad deployment of the most advanced fuels.

The development of 2026 prices makes the structural issue clearer. Once market disruption reduced the cost of switching from MGO to biofuel, companies that would previously have purchased compliance could become low-cost surplus generators themselves. Additional supply then placed downward pressure on surplus prices, weakening the revenue available to more expensive generators. BetterSea’s August weighted end-of-month price of €83.13/tCO₂e illustrates how far the market has moved from levels around €190-200/tCO₂e seen during the earlier compliance cycle.

The Brazilian version of the IMO framework (just as the previous draft) contains several features that could reduce the persistence of the same problem. Surplus expires after two years rather than remaining banked indefinitely. From 2031, surplus is generated against a substantially more stringent Direct Compliance Target rather than the Base Target. Tier 1 compliance cannot be purchased from the ordinary surplus unit market, preserving demand for the remedial mechanism. Finally, ZNZ fuels and technologies receive a separate reward channel rather than being expected to compete solely on the price of generic compliance.

None of these mechanisms eliminates the underlying market effect. If one pathway can generate an IMO Surplus Unit at very low cost, it will put downward pressure on the value of every other transferable surplus unit. The 2029 and 2030 introductory period may be particularly exposed because the two GFI thresholds are temporarily identical. The difference is that the IMO architecture does not rely on surplus trading alone to carry the entire decarbonisation incentive. It separates flexible compliance, remedial compliance and support for deep decarbonisation into different parts of the system.

That distinction may ultimately be one of the most important lessons from FuelEU. A compliance market is naturally designed to find the cheapest way to meet a regulatory target; it is not automatically designed to create a business case for every fuel pathway that will eventually be required. A regulatory framework that wants to achieve both objectives needs to distinguish between the value of ordinary compliance and the additional support required to move technologies down the cost curve.

Next steps at IMO

ISWG-GHG 22 did not agree on Brazil’s proposal or select a single revised architecture. The official outcome records the Chair’s observation that delegations showed willingness to continue working towards text that could be presented to MEPC 85, while interested Member States were invited to consult further and submit proposals reflecting greater convergence. Work on several of the implementation guidelines was also deferred because of time constraints.

The next formal negotiating stage is ISWG-GHG 23 from 23 to 27 November 2026, immediately ahead of MEPC 85 from 30 November to 3 December. The adjourned second extraordinary session of MEPC is then scheduled to resume on 4 December 2026, subject to confirmation by MEPC 85.

Brazil argues that the mid-term measures should be adopted no later than MEPC 85 and notes that the possible amendments have already been circulated through Circular Letter No. 5214 to satisfy the relevant MARPOL circulation requirements. It also proposes that the wider review of the 2023 IMO GHG Strategy should begin only after sufficient implementation experience exists, suggesting a review 16 months after entry into force of the mid-term measures.

Brazil’s contribution provides one possible way of bringing all the NZF elements together without abandoning the work already completed at IMO. FuelEU’s first compliance year, however, shows why the final design needs to be assessed as a market rather than only as a set of reduction percentages. The important question will not only be whether ships can comply, but whether the interaction between physical fuel choices, surplus units, Remedial Units, banking and ZNZ rewards creates the investment signal needed as the targets become progressively more demanding.