2.1. Distinguishing CDR from Related Technologies

Within the realm of ‘greenhouse gas removal’ or ‘GHG removal’, which refers to the ‘withdrawal of a GHG and/or a precursor from the atmosphere by a sink’,Footnote ³¹ CDR consists of those measures that target the removal and storage of CO₂.Footnote ³² GHG removal encompasses CDR measures but is broader in scope, referring to the set of technologies that remove and store any GHGs from the atmosphere.Footnote ³³ This article will generally refer to CDR as distinct from GHG removal, in line with the practice of EU institutions.Footnote ³⁴

Carbon capture and storage (CCS) and carbon capture and utilization (CCU) can form an integral part of CDR technologies when they are applied to CO₂ that has been captured from the atmosphere.Footnote ³⁵ The capture of CO₂ from the atmosphere can be achieved either indirectly through biomass or directly from the air.Footnote ³⁶ CCS and CCU do not qualify as CDR technologies if they are applied to CO₂ from fossil fuel use, as no CO₂ is removed from the atmosphere.Footnote ³⁷ CCS can be defined as a ‘process in which a relatively pure stream of CO₂ from industrial and energy-related sources is separated (captured), conditioned, compressed and transported to a storage location for long-term isolation from the atmosphere’.Footnote ³⁸ CCU is referred to as a ‘process in which CO₂ is captured and then used to produce a new product’.Footnote ³⁹ For example, CCS is an essential part of CDR when used in conjunction with bioenergy (BECCS). In contrast, CCS combined with a coal-fired power plant does not remove any CO₂ from the atmosphere and therefore cannot be considered as CDR. Only when CCS and CCU are components of CDR technologies, are they relevant to the scope of this article.

2.2. Required Levels of Carbon Dioxide Removal

Although GHG emissions reductions are the most important measure for climate change mitigation, GHG removal, of which CDR technologies are the most relevant, will be necessary to achieve the global temperature increase targets of the Paris Agreement:Footnote ⁴⁰ specifically, if we intend to hold ‘the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels’.Footnote ⁴¹ To reach this objective, the Paris Agreement stipulates that parties must aim to ‘achieve a balance between anthropogenic emissions by sources and removals by sinks of greenhouse gases in the second half of this century’.Footnote ⁴² Emissions reductions alone are not sufficient to avoid catastrophic climate change, as scientific research indicates.Footnote ⁴³ Of the scenarios in integrated assessment models considered at the time of the Paris Agreement, 87% of those that aim to achieve 2°C, and all those that expect 1.5°C, relied partly on GHG removal to reach the temperature targets of the Paris Agreement.Footnote ⁴⁴ Without GHG removal, very drastic emissions reductions would be necessary to reach 2°C, while 1.5°C would remain out of reach.Footnote ⁴⁵ According to a report from the National Academies of Sciences, CDR efforts must annually remove 10 gigatons (Gt) CO₂ by the late 2050s and 20 Gt CO₂ by the late 2090s globally to limit global warming to below 2°C.Footnote ⁴⁶ For comparison, global net anthropogenic GHG emissions amounted to 59 Gt CO₂-eq in 2019.Footnote ⁴⁷ The need to scale up CDR was also emphasized in the Paris Agreement's first global stocktake concluded in the United Arab Emirates in 2023. The document ‘calls on Parties’ to contribute to the global efforts to accelerate removal technologies such as CCS and CCU, particularly in hard-to-abate sectors.Footnote ⁴⁸

CDR technologies are not only essential in helping to reach net zero emissions but also for maintaining a net zero status by removing residual hard-to-abate emissions, including those from sectors such as agriculture and aviation. After complete decarbonization, CDR could deliver genuine removal by drawing down ‘legacy carbon’ remaining in the atmosphere from past emissions.Footnote ⁴⁹

The Report on the State of Carbon Dioxide Removal by Smith and co-authors finds a ‘yawning gap’ between the extent to which countries are planning CDR deployment and what will be required to achieve the global temperature increase targets of the Paris Agreement.Footnote ⁵⁰ This also holds for the EU, which is currently not on track to deliver the required level of CDR deployment.Footnote ⁵¹

2.3. The European Union Emissions Trading System

The EU ETS was set up in 2005 as the world's first international emissions trading system. It operates in all EU Member States, as well as Liechtenstein, Iceland, and Norway. The EU ETS currently covers installations in the power sector and manufacturing industry, and airlines operating between the participating countries, which make up for approximately 40% of the EU's GHG emissions.Footnote ⁵² It is one of the largest emissions trading systems worldwide.Footnote ⁵³ The EU agreed in 2023 to expand the EU ETS to other sectors, such as international aviation and maritime transport, and to create a separate emissions trading system for fuel combustion in buildings, road transport, and additional sectors (so-called ‘EU ETS II’), which will become operational from 2027 onwards.Footnote ⁵⁴ The European Commission referred to the EU ETS as a ‘cornerstone’ of the EU's strategy to reduce GHG emissions.Footnote ⁵⁵ The EU ETS regulatory framework is set out in the EU ETS Directive, which has been updated over time.

The EU ETS is a ‘cap and trade’ system, which sets a cap on the absolute quantity of certain GHG emissions from entities covered by the system.Footnote ⁵⁶ Entities covered by the system must annually surrender sufficient emissions allowances to cover their emissions. Failure to do so may result in heavy penalties. The number of allowances available corresponds with the cap.Footnote ⁵⁷ The cap, and consequently, the total number of allowances, is reduced over time to decrease overall emissions. As a result of supply and demand, the price of emissions allowances is likely to rise, providing an incentive to reduce emissions and invest in low-carbon technologies.Footnote ⁵⁸ Entities covered by the system buy allowances at auctions or, in certain cases, receive them for free.Footnote ⁵⁹ Moreover, allowances can be traded among participants. The market approach is intended to ensure that emissions are reduced in the most cost-effective way.Footnote ⁶⁰ Participants in the EU ETS were able to use international credits from the Kyoto Protocol's CDM to fulfil part of their obligations under the EU ETS until 2020.Footnote ⁶¹ This was achieved by linking the EU ETS and the Kyoto project-based mechanism.Footnote ⁶²

Eventually, the cap will reach zero, putting an end to the supply of emissions allowances.Footnote ⁶³ Entities covered by the system could no longer purchase allowances to cover their GHG emissions. This could become problematic as some GHG emissions will be difficult or even impossible to mitigate.Footnote ⁶⁴ Integrating CRCs into the EU ETS emerges as a solution to cover GHG emissions from these so-called hard-to-abate sectors.Footnote ⁶⁵

2.4. The Current Regulatory Framework of CDR in the EU

At present, the EU ETS does not include a mechanism for the creation of additional allowances or credits through CO₂ removal.Footnote ⁶⁶ Article 2(1) of the EU ETS Directive limits the scope of the directive to positive emissions, excluding negative emissions technologies.

The European Commission published a legislative proposal for a regulation on a carbon removals certification framework in November 2022.Footnote ⁶⁷ The proposed regulation aims to establish a voluntary Union certification framework for carbon removals, with the objective of incentivizing the uptake of high-quality carbon removals.Footnote ⁶⁸ It lays down quality criteria, rules for verification and certification, and rules for the functioning and recognition of certification schemes.Footnote ⁶⁹ This framework intends to form a foundation that policymakers can draw upon to incentivize and govern CDR technologies. It does not currently facilitate the integration of CRCs into the EU ETS. Nevertheless, it could potentially serve as a regulatory foundation for another legislative proposal aimed at integrating CRCs into the EU ETS that were certified under an EU carbon removals certification framework. Participants in the EU ETS would be able to purchase CRCs and use them towards fulfilling their obligations under the EU ETS, or at least parts of it. This option is already foreseen in the EU ETS Directive.Footnote ⁷⁰ Article 30(5)(a) of the EU ETS Directive mandates the EU Commission to report by 31 July 2026 ‘how negative emissions resulting from greenhouse gases that are removed from the atmosphere and safely and permanently stored could be accounted for and how those negative emissions could be covered by emissions trading’.Footnote ⁷¹

In contrast, CCS is already subject to a detailed regulatory framework. The EU ETS Directive provides incentives for the deployment of CCS because the obligation to surrender allowances does not arise for emissions that are verified as captured and transported to an authorized facility for permanent storage.Footnote ⁷² The CCS Directive stipulates, inter alia, provisions for permits and obligations for the operation, closure and post-closure of sites.Footnote ⁷³ In the case of carbon leakage, the operator must notify the competent authority and must take corrective measures.Footnote ⁷⁴ Moreover, the operator must surrender EU ETS allowances as CO₂ capture, transport, and storage facilities are covered by the EU ETS.Footnote ⁷⁵ Further liability rules for operators are stipulated in the Environmental Liability Directive and at the Member State level.Footnote ⁷⁶ Most liabilities arising in relation to the storage site are transferred to the competent authority after the storage site has been closed and certain conditions apply so-called risk transfer.Footnote ⁷⁷ These conditions are (i) that all available evidence indicates that the stored CO₂ will be completely and permanently contained, and (ii) that a minimum period has passed (not less than 20 years), which is to be determined by the competent authority.Footnote ⁷⁸ The operator can be held liable, even though the risk transfer has taken place, if the clawback provision pursuant to Article 18(7) applies. This provision allows for the post-transfer recovery of costs if the operator has been at fault, which includes cases of deficient data, wilful deceit, or a failure to exercise due diligence.Footnote ⁷⁹ Before the risk transfer has taken place, the CCS operator is exclusively liable for any CO₂ leakage. For CCS projects under the CDM, a transfer of liability from the project operator to the host state would take place under similar conditions.Footnote ⁸⁰

3.1. Why Permanence Matters

Public policy needs to define parameters to determine equivalence between CRCs stemming from different CDR technologies and emissions allowances.Footnote ⁸¹ Broad fungibility is important for promoting a liquid market that provides sufficient investments to scale up CDR at a high rate.Footnote ⁸² The only common parameter that seems to apply to these units is one metric tonne of CO₂ equivalent.Footnote ⁸³ However, it is difficult to define scientifically the equivalence between CO₂ captured and stored through different CDR technologies and avoided CO₂ emissions.Footnote ⁸⁴ This is because the various CDR technologies available differ significantly regarding their levels of permanence.Footnote ⁸⁵ For instance, carbon theoretically can be stored indefinitely in terrestrial ecosystems, such as forests, peatlands, and soils.Footnote ⁸⁶ However, the risk of reversal as a result of human action – such as deforestation, or natural disturbances, including fire and drought – is relatively high.Footnote ⁸⁷ Carbon stored in biochar in soil, on the other hand, will reverse after a certain period.Footnote ⁸⁸ Some works suggest it can persist on a centennial scale.Footnote ⁸⁹ CO₂ stored in well-selected and well-managed geological formations through BECCS or DACCS can be stored for a thousand years or longer.Footnote ⁹⁰ Under these conditions, reversal is much less likely to occur, although it remains a concern.Footnote ⁹¹ The quantity of CO₂ escaping from appropriately selected and managed geological formations will ‘very likely’ remain below 1% over the first 100 years.Footnote ⁹²

Generally, more permanent CDR technologies are more significant from a climate repair perspective than less permanent technologies.Footnote ⁹³ This is because, in the event of a reversal, the original benefits of the CDR project are reversed to some extent, depending on when the reversal occurs.Footnote ⁹⁴ However, the value of temporary removal of CO₂ is not zero.Footnote ⁹⁵ A CDR project does not have to remove a tonne of CO₂ indefinitely to effectively offset a tonne of CO₂ emitted. This is because the bulk of the original emission does not stay in the atmosphere eternally, as atmospheric CO₂ is being sequestered through natural processes over time. Only a fraction of the CO₂ emitted remains in the atmosphere essentially forever.Footnote ⁹⁶ Even if the removed CO₂ reverses before the original emission has fully decayed, a certain percentage of CO₂ has already been taken from the atmosphere, thus reducing the global net atmospheric CO₂ concentration. For instance, a CDR project that reverses after 100 years has the net effect of a 39% reduction relative to no CDR. This increases to 66% after 1,000 years.Footnote ⁹⁷ Moreover, temporary removal of CO₂ can buy humanity time to develop methods to permanently remove CO₂.Footnote ⁹⁸ Therefore, CDR technologies with low levels of permanence entail benefits.Footnote ⁹⁹

In other words, CDR technologies with high levels of permanence are more valuable than those with low levels of permanence, but CDR technologies with low levels of permanence still provide benefits.

3.2. Risks of Low Levels of Permanence for Integration

Integrating CRCs issued for CDR technologies with low levels of permanence into the EU ETS could pose two main challenges.

Firstly, by compensating for emissions with CDR that have low levels of permanence, the environmental integrity of the EU ETS could be jeopardized. For instance, if CRCs can be used for compliance under the EU ETS, one tonne of CO₂ can be emitted by submitting a CRC. This would be acceptable from a climate perspective because the tonne of CO₂ emitted is offset by the removal. However, this calculation holds only if the removal is sufficiently permanent. If the captured and stored tonne of CO₂ reverses after a certain period, a certain net amount of CO₂ is added to the atmosphere.Footnote ¹⁰⁰ The transfers of CRCs would then lead to an increase in aggregated global emissions and undermine the environmental integrity of the system. This result could be addressed through regulations on liability. However, liability presents its own set of challenges and should not be the primary mechanism to manage permanence issues, as further discussed in the next section.

Secondly, an EU legislative act that does not consider the different levels of permanence would create economic incentives to deploy currently cheaper and less permanent CDR technologies.Footnote ¹⁰¹ Therefore, permanent CDR technologies would not be scaled up and the goal of reaching carbon neutrality in the long term would be threatened. Less permanent CDR technologies, such as afforestation, are generally cheaper, whereas most engineering-based approaches that show high levels of permanence, such as BECCS and DACCS, are currently relatively expensive.Footnote ¹⁰² This is problematic, as sufficient investments in engineering-based CDR are critical for reaching net zero emissions and maintaining it.Footnote ¹⁰³

Therefore, an EU legislative act must consider the different levels of permanence of the CDR technologies when integrating CRCs into the EU ETS.

3.3. Liability as a Risk Management System

A common approach under the CCS Directive,Footnote ¹⁰⁴ and under many voluntary carbon schemes,Footnote ¹⁰⁵ is to hold the project operator liable for any re-emissions. There are several mechanisms available to hedge reversal risks. In many voluntary carbon schemes, project operators have to retain a certain portion of the received credits in a risk buffer account, often ranging from 10 to 40%.Footnote ¹⁰⁶ Under some schemes, the portion relies on the project's ex ante risk rating.Footnote ¹⁰⁷ This portion cannot be sold for a predetermined period and should be used to compensate for reversals.Footnote ¹⁰⁸ Risk pooling is a variant of risk buffer accounts, where multiple CDR projects keep a common buffer account. An advantage of this pooled approach is that individual contributions may be less than those for individual buffer accounts.Footnote ¹⁰⁹ Moreover, there are non-permanence insurances that cover a portfolio of different CDR projects.Footnote ¹¹⁰ A combination of these approaches is also possible.

Situations where liability becomes necessary should be avoided in the first place as liability itself entails many risks. For instance, it can be very challenging to identify the relevant emitter and enforce the liability against them, depending on with whom the liability lies and the time frame.Footnote ¹¹¹ Companies may be insolvent, untraceable, or no longer exist.Footnote ¹¹²

Moreover, if CDR were to reverse simultaneously on a large scale, exacerbated by extreme weather events caused by climate change in the future, the liability system could be at risk of collapsing.Footnote ¹¹³ For instance, in the event of catastrophic forest fires, many CDR projects for which CRCs have been issued could be reversed. The liable entities, most likely the project operators, would have to compensate for the climate damage by surrendering CRCs or emissions allowances. If such reversal events were sufficiently large, the demand for CRCs and emissions allowances would soar, most likely driving up prices significantly and potentially pushing many project operators into bankruptcy. This will become even more problematic as the number of emissions allowances in the market will continue to decline. Once there are no more emissions allowances left in the EU ETS, CRCs issued for failed CDR can be replaced only with other CRCs. Depending on the availability of CRCs and demand at the time, this could be a costly undertaking. Risk buffer accounts and non-permanent insurance can increase the resilience of the system, but they too have limitations. The question then becomes, who is going to ensure that the reversed CO₂ is compensated for if the project operator fails? Most likely, this will be the responsibility of the Member States. In the event that the reversals are not fully offset, the environmental integrity of the EU ETS would be undermined, as more CO₂ would have been emitted than in a situation where there was no trading of CRCs. This is because integrating CRCs would expand the total gross GHG emissions permissible under the EU ETS. This expansion would be acceptable as the additional GHG emissions would be offset by CDR, leaving the net GHG emissions trajectory unaffected. Nonetheless, this reasoning remains valid only if the CO₂ is stored permanently. Should the stored CO₂ reverse prematurely, the original emission is no longer (fully) offset. In this scenario, more GHG emissions would have occurred than in a scenario where no CRCs were traded under the EU ETS and the environmental integrity of the system would be disrupted.

For these reasons, liability should be seen as a legal safeguard for exceptional circumstances and not as a general solution for the permanence problem. Situations that give rise to liability need to be prevented from the outset to avoid the above issues in advance.Footnote ¹¹⁴

4.1. Clean Development Mechanism

Under the former Clean Development Mechanism of the Kyoto Protocol, removal projects in developing countries (non-Annex I parties) could generate temporary carbon credits.Footnote ¹¹⁵ Each temporary credit would represent a tonne of CO₂-equivalent removed. Developed state parties (Annex I parties) could buy and use these temporary credits to meet part of their emissions reduction commitments under the Kyoto Protocol. These temporary credits would expire after a certain time because of concerns regarding the limited permanence of removal projects. Eventually, the temporary credits had to be replaced by the purchaser with permanent credits at a one-to-one ratio.Footnote ¹¹⁶

Only a limited range of CDR technologies, primarily afforestation and reforestation, were eligible under the CDM.Footnote ¹¹⁷ An important reason to make credits temporary was concern about the potential reversibility of the stored carbon. Afforestation and reforestation were considered temporary solutions to gain time in which to develop the technologies to effectively mitigate climate change.Footnote ¹¹⁸ They were not regarded as long-term solutions that could generate credits equivalent to credits from emissions reduction projects. This practice ensured the environmental integrity of the CDM regarding removal projects, as no emissions could be offset by CDR in the long term. This does not pertain to problems regarding the accuracy of monitoring and additionality of removal projects as they are outside the scope of this study.

Temporary credits represented only a marginal piece of the CDM carbon market. They were excluded from most emissions trading schemes, such as the EU ETS, particularly because of concerns about permanence, the accuracy of monitoring, and liability.Footnote ¹¹⁹ For example, temporary credits would have created liability risks under the EU ETS because a company that wanted to cease operations could sell its permanent credits and replace them with cheaper temporary credits. If the company ceased to exist, it could no longer replace the temporary credits with permanent credits. Consequently, the Member State in which the company operated had to cover the expired credits.Footnote ¹²⁰

Investments in CDR projects under the CDM were very weak;Footnote ¹²¹ this was mainly as a result of the restrictive regulatory approach under the CDM.Footnote ¹²² Temporary credits were complex and not fungible with other carbon assets.Footnote ¹²³ They were not admitted under the EU ETS, which was the largest market for CDM credits.Footnote ¹²⁴ They were difficult to manage and transfer because of the obligation to replace them.Footnote ¹²⁵ Perhaps most importantly, temporary carbon sequestration was not valued as the replacement ratio with permanent credits was one-to-one.Footnote ¹²⁶

4.2. Voluntary Carbon Markets

Many companies purchase carbon credits on voluntary carbon markets to offset their emissions.Footnote ¹²⁷ Voluntary carbon credits are issued by carbon standardsFootnote ¹²⁸ for the reduction or removal of GHG emissions.Footnote ¹²⁹ Each carbon standard sets its own rules with which CDR projects must comply to be certified. The quality and price of the credits can consequently vary significantly, as each project uses a specific technology and shows disparate levels of permanence.Footnote ¹³⁰ In 2021, the value of voluntary carbon market transactions exceeded US$ 1 billion, more than doubling since 2020.Footnote ¹³¹ Currently, they mainly finance nature-based solutions such as afforestation or soil carbon sequestration in the removal sector.Footnote ¹³²

A typical approach taken by participants in voluntary carbon markets to ensure some degree of permanence is through long-term project contracts, typically ranging from 3 to 100 years.Footnote ¹³³ The length of these contract periods is based on contractual realities rather than any notional definition of permanence or atmospheric residence of CO₂.Footnote ¹³⁴ Contractual realities in this sense refer to the period deemed appropriate to require private parties to ensure the safe storage of carbon. The carbon credits are typically issued at the beginning of a project and the carbon must remain stored during the contract period.

If any part of the carbon reverses before the project period ends, liability and risk-buffer accounts are intended to ensure the environmental integrity of the system. Many standards distinguish between intentional and unintentional reversal.Footnote ¹³⁵ In the case of unintentional reversal, such as wildfires, the credits from the (pooled) risk buffer are generally used to replace the loss. If the reversal was caused intentionally, such as through the resumption of deforestation by the project operator, all previously issued credits must generally be replaced by the project operator. Thus, the project operator loses all the benefits it has received.Footnote ¹³⁶

The approach of voluntary carbon markets does not ensure that carbon is removed and stored for a significant period in atmospheric terms. Most project contracts stipulate project lengths of up to 100 years, yet many contract periods are significantly shorter.Footnote ¹³⁷ From an atmospheric perspective, 100 years is a relatively short period because CO₂ resides in the atmosphere for much longer.Footnote ¹³⁸ As mentioned above, the storage of CO₂ for 100 years has the effect of reducing atmospheric carbon stock by just 39% compared with no CDR.Footnote ¹³⁹ However, from a contractual perspective, 100 years is a long period of time as many project operators are not willing to or are not capable of committing further into the future.

The practice of many voluntary carbon standards to issue credits at the beginning of the project period could jeopardize the stability of the system. Large-scale simultaneous reversals could push the risk management system of liability and risk buffer accounts to its limits. The possibility of such a scenario, coupled with relatively short contract periods, forges a vulnerability that questions the environmental integrity of the system.

Furthermore, most voluntary carbon schemes do not reward temporary carbon storage if it is reversed intentionally before the end of the contract period. In this case, temporary carbon storage is worthless even though the project did create some climate benefits and investments were made. Consider, for example, a reforestation project with a contract period of 100 years. The project reverses fully after 90 years as a result of the resumption of deforestation by the project operator. The project operator must then replace all the credits it has received for this project and the investments that it has made are lost. One could argue that this result is justified to prevent project operators from intentionally undoing CO₂ storage. However, the project operator might have substantial impending reasons, including a financial crisis, and has already created some benefits by storing CO₂ for 90 years. This ‘all-or-nothing approach’ may deter the deployment of CDR projects at the forefront.

4.3. Tonne-Year Accounting

Another approach, referred to in the literature as ‘tonne-year accounting’, involves issuing credits incrementally over the project period. Permanent credits are issued after a certain number of carbon tonne-years of storage have been achieved.Footnote ¹⁴⁰ This way, storage periods with varying lengths can be put in equivalent and fungible units. Murray and Kasibhatla describe this approach as ‘less conservative than temporary crediting’, which was practised under the CDM, and ‘more conservative than prepaid credits and buffer provisions’, which are typically used in voluntary carbon markets.Footnote ¹⁴¹

Murray and Kasibhatla suggest that policymakers should stipulate a time period, such as 100 years, for carbon storage after which full permanence is earned. This period must not represent a permanent emissions reduction in atmospheric terms but should serve as a ‘policy-relevant target for de facto permanence’.Footnote ¹⁴² Once a certain number of carbon tonne-years have been accumulated, a certain degree of permanence has been achieved and cannot be reversed. The more carbon that is stored for longer periods, the more permanence is gained. Hence, the level of permanence and therefore the value of the project accumulate exponentially over time.

This approach tries to pursue a middle ground between ‘prepaid credits’ that are given out at the beginning of a project, such as in voluntary carbon markets, and the option to receive the credits at the end of the project period. In this case, operators earn credits incrementally over the course of the project. Murray and Kasibhatla claim that this approach ‘seems to reduce vulnerability of the entire system to reversal risk’.Footnote ¹⁴³ It builds on the rationale that short-term removal of CO₂ still entails some benefits and should be rewarded, which is reflected in an incrementally increasing value of the CO₂ storage. It seeks to ensure the environmental integrity of the system by allowing no more emissions than have been ‘earned’ through the accumulation of carbon tonne-years. Contrary to voluntary carbon schemes, this approach does not rely on liability. The credits are irrevocable, even though a CDR project might reverse in the future.Footnote ¹⁴⁴ Therefore, the system is resilient and challenges relating to liability are being avoided.

It is important to carefully define the period that is considered permanent, after which the full credits are received. If it were too short, the environmental integrity of the system would be low, and it would incentivize the deployment of currently cheaper low permanent CDR technologies. For instance, if the period were to be 100 years, this threshold could be reached through afforestation. CO₂ storage beyond this point would not be rewarded, as full credits are earned after 100 years and no liability arises in the event of reversal after that time.

This approach has been criticized on the basis that it is questionable that the delayed revenue stream would provide sufficient financial incentives to scale up CDR rapidly and significantly.Footnote ¹⁴⁵ For many CDR technologies, costs are incurred at the beginning of the project, while payment would be made after permanence has been earned. This might deter investors. An EU report on carbon farming flagged a practice in voluntary carbon schemes as problematic from a cash-flow perspective, which issued credits after the carbon has been sequestered.Footnote ¹⁴⁶ The report states that the ‘associated lagging stream of revenues is the main structural barrier that has led to the limited uptake of reforestation activities’.Footnote ¹⁴⁷ This criticism applies a fortiori to tonne-year accounting as carbon sequestration occurs before a certain degree of permanence has been earned. Moreover, the final report of the Taskforce on Scaling Voluntary Carbon Markets identified the ‘long lag times between investment and return’ and the accompanying ‘lack of financial attractiveness’ as a significant mobilization challenge.Footnote ¹⁴⁸ This issue of scattered financial flows could be addressed by reducing the level of liability in proportion to the reduction in the global warming potential of the original emission, as opposed to earning credits once carbon tonne-years have been accumulated. In other words, instead of issuing credits incrementally, the level of liability is decreasing according to the accumulation of carbon tonne-years. In this way, the full number of credits can be issued at the beginning of the project period. CDR deployment, therefore, is made more financially attractive. In the case of a reversal, only the remaining global warming potential of the original emissions at the time of the reversal must be compensated, rather than demanding the submission of all the credits that were originally issued for the project.

However, there are more fundamental criticisms of tonne-year accounting in the scientific community. It has been criticized for the use of subjective economic discount rates and arbitrary time horizons to assess the costs of emissions and the benefits of temporary storage.Footnote ¹⁴⁹ Others have argued that equivalence claims between temporary carbon storage and avoided emissions or permanent storage are flawed.Footnote ¹⁵⁰ The Article 6.4 Mechanism Supervisory Body stated that it will ‘focus on measures that address reversals on a tonne-for-tonne basis, and not on a tonne-year basis’ because of concerns ‘within the scientific community, regarding its underpinning methods and assumptions, and ecological implications, and insufficient confidence in its suitability for international applications and effectiveness at addressing reversals’.Footnote ¹⁵¹

4.4. Lessons Learned

The restrictive approach to CDR taken under the CDM did ensure that no emissions were permanently offset by non-permanent carbon removal, thereby ensuring the environmental integrity of the system in this respect. However, the CDM failed to encourage extensive investments in CDR. CDR were not sufficiently valued, as temporary credits had eventually to be replaced with permanent credits on a one-to-one ratio. Such an approach might have been satisfactory at this time because afforestation and reforestation were perceived as temporary solutions. However, it will not be sufficient to achieve the current goal of scaling up CDR rapidly and significantly.

The approach of voluntary carbon schemes has the potential to drive investment, as evidenced by high demand and growth rates. However, the contract periods that guarantee safe storage of CO₂ are relatively short and the system relies heavily on liability and risk buffer accounts. This approach does not seem to offer the best solution to ensure environmental integrity.

The tonne-year approach does not rely on liability and risk buffer accounts, which avoids challenges regarding execution, and which bolsters the system's resilience. Short-term CDR technologies are valued based on the benefits they provide. However, concerns about the underlying methods and assumptions, as well as the environmental impacts, make this approach less favourable. In addition, the dispersed financial flow would most likely not encourage sufficient investment in CDR.

From the analysis of these approaches, it appears that a right balance must be struck, particularly between encouraging investments and ensuring environmental integrity. Too stringent rules on permanence and liability can present barriers to participation while loose rules can jeopardize environmental integrity.Footnote ¹⁵²

5.1. Permanence

The EU intends to create a certification framework for all types of carbon removal, not all of which will be eligible for integration into the EU ETS because of their varying levels of permanence. Therefore, the article argues that the EU should create a distinct, more stringent definition of permanence, specifically for the purpose of integrating CRCs into the EU ETS.

The Commission's proposal for a carbon removals certification framework defines permanent carbon storage as ‘a carbon removal activity that, under normal circumstances and using appropriate management practices, stores atmospheric or biogenic carbon for several centuries’.Footnote ¹⁵⁴ This definition must be further elaborated. The EU should explicitly define a time frame that it deems permanent. Currently, it is not specified what exactly is meant by ‘several centuries’. This could lead to uncertainties about whether specific CRCs are eligible for integration into the EU ETS, and therefore hinder investment and deployment.

This article proposes to define permanent CDR as technologies that have a relatively low risk of significant CO₂ reversals prior to the end of the period considered permanent. This definition uses three variables to determine the permanence of CDR technologies: (i) what storage period is deemed permanent; (ii) what constitutes a significant reversal of the stored CO₂; and (iii) what constitutes a low reversal risk. Depending on the exact definition of these variables, many nature-based solutions could be considered non-permanent, whereas engineered CDR technologies (such as BECCS and DACCS) could be regarded as permanent.

Firstly, the EU must determine the permanent storage period. Defining the permanent storage period is about finding the right balance between ensuring the environmental integrity of the EU ETS and scaling as many CDR activities as possible. If the permanent storage period is too long, many CDR technologies would be excluded from the market, which otherwise could provide valuable (temporary) storage. If it is too short, the level of environmental integrity of the EU ETS would diminish. There is no uniform definition of permanence.Footnote ¹⁵⁵ Murray and Kasibhatla define permanence ‘as the point in time at which stored carbon has essentially fulfilled its role as offsetting the global warming potential of the original emission’.Footnote ¹⁵⁶ This definition should guide the determination of the permanent storage period. The stored carbon must offset a significant percentage of the global warming potential of the original emission, rather than fully offset it. This approach does not guarantee complete environmental integrity as a certain percentage of the original emission will remain in the atmosphere. However, it is necessary to draw a line because a certain percentage of the original emission will inevitably remain in the atmosphere nearly forever, making it almost impossible for it to be fully offset. In addition, the temporary removal offers other benefits, such as buying humanity time in which to develop more permanent CDR technologies.

Secondly, the EU must stipulate what constitutes a significant reversal of the stored CO₂ in addition to what is considered a high versus low reversal risk. A possible problem in that respect is the lack of information regarding the reversal risks of certain CDR technologies.Footnote ¹⁵⁷ To address this knowledge gap, the precautionary principleFootnote ¹⁵⁸ should apply. When the reversal risk is not sufficiently clear, a CDR technology shall be deemed non-permanent in order to ensure the environmental integrity of the system. Further research is required to better understand the reversal risks of CDR technologies.

The Commission's proposal suggests that the ‘validity of the certified carbon removals should depend on the expected duration of the storage and the different risks of reversal associated with the given carbon removal activity’.Footnote ¹⁵⁹ CDR technologies that store carbon in geological formations should not be subject to an expiry date as they provide enough certainty on the long-term duration of the stored carbon. The validity of the certified carbon removals from carbon farming or carbon storage in products, conversely, should be subject to an expiry date as they are more exposed to the risk of re-emission. After the expiration date, the carbon should be considered as re-emitted unless the operator demonstrates the maintenance of carbon storage through continuous monitoring activities. This article argues that CRCs with an expiration date should not be integrated into the EU ETS, as the experiences with temporary credits under the Kyoto Protocol have demonstrated.Footnote ¹⁶⁰ The expiry of certain credits would make them not fungible with indefinite CRCs and emissions allowances, thus increasing complexity and administrative work. Therefore, CRCs with an expiration date are likely to be unattractive to market participants. Instead, this article proposes a clear cut between permanent and non-permanent CDR technologies. Only CRCs issued for permanent CDR technologies should be integrated into the EU ETS. Non-permanent CDR technologies, particularly those with strong co-benefits, should be promoted through other mechanisms such as the Innovation Fund,Footnote ¹⁶¹ the Common Agricultural Policy,Footnote ¹⁶² the Regional Development Fund,Footnote ¹⁶³ or by means of a distinct market for non-permanent CRCs. This is because the benefits of short-term removal and storage of CO₂ cannot outweigh the risks that their inclusion would pose for the environmental integrity of the EU ETS as a result of potential large-scale reversal events. It is crucial to remember that the EU ETS is a cornerstone, if not the leading mechanism, in the EU ambition to reduce its emissions. Therefore, the standard of environmental integrity must be high.

5.2. Liability

The Commission's proposal for a carbon removals certification framework does not specify what a liability system could look like. It only calls for ‘appropriate liability mechanisms’, which could include discounting of carbon removal units, collective buffers, or up-front insurance mechanisms.Footnote ¹⁶⁴ With regard to the geological storage of CO₂, the liability mechanisms and corrective measures of the EU ETS Directive and the CCS Directive should apply to avoid double regulation.Footnote ¹⁶⁵ The Commission's proposal makes it clear that the project operator or a group of operators should be liable for any re-emission of a CDR project.Footnote ¹⁶⁶ This is compelling as the project operator is the best place to address any reversal, both with regard to knowledge of the project and access to the site.Footnote ¹⁶⁷ Passing the liability risk to the purchaser of the removal unit would provide a ‘perverse incentive’ to the project operator not to bear responsibility for safe storage.Footnote ¹⁶⁸

One way of establishing a liability regime is to include CDR technologies in Annex I of the EU ETS Directive, which lists the activities covered by the EU ETS. In the event of a reversal, allowances or CRCs would have to be submitted as for any other source of GHG emissions. This approach has been taken for CCS activities.Footnote ¹⁶⁹ This liability regime would be robust if only CRCs issued for permanent CDR technologies were integrated into the EU ETS. Large-scale, simultaneous reversals that overwhelm the system would then be unlikely. Liability should become relevant only when a CDR project reverses before the permanent storage period ends. It should not apply after that period, as a significant percentage of the original emission will have decayed by that time.

This article distinguishes between two reversal scenarios: disastrous and non-disastrous. A ‘disastrous reversal’ should be defined as a kind of reversal caused by catastrophes beyond the control of the project operator, such as drought, wildfires, flooding, earthquakes, storms, tornados, or human-induced events. A ‘non-disastrous reversal’ could be defined as a reversal that does not fall within the category of disastrous reversals, particularly because the project operator had been at fault. This applies, for instance, in situations where the project operator acted intentionally or failed to exercise due diligence. The distinction between these scenarios is practised widely in voluntary carbon schemes.Footnote ¹⁷⁰ The clawback provision of Article 18(7) of the CCS Directive introduces a similar differentiation.

This article proposes that the project operator should be liable in the event of a non-disastrous reversal until the end of the permanent storage period. In the case of a disastrous reversal, the liability risk should be transferred from the project operator to the competent authority when certain conditions apply. Lessons can be drawn from the CCS liability regime because both CCS and permanent CDR demonstrate high levels of permanence. BECCS and DACCS, which most likely fall within the category of permanent CDR, rely on CCS for the storage of removed CO₂. Under the CCS Directive, a transfer of liability risk from the operator to the competent authority takes place.Footnote ¹⁷¹ For CDR, a risk transfer should apply under similar conditions because liability periods that are too long are unpredictable and uninsurable, and therefore are likely to discourage the deployment of and investment in CDR.Footnote ¹⁷² The conditions for a risk transfer should be, alongside other possible requirements, that (i) all information available must indicate that the bulk of the stored CO₂ will be safely contained until the end of the permanent storage period, and (ii) that a minimum period has passed. The allocation of part of the liability to the Member State would be justified by the public interest in the widespread application of CDR to tackle climate change.Footnote ¹⁷³ Moreover, large reversals that force the Member State to make payments under the liability scheme are unlikely to occur when only CRCs issued for CDR technologies with high levels of permanence are integrated.