Should carbon offsets only include removing CO2 from the atmosphere?

April 17, 2020, by Derik Broekhoff

One idea gaining currency lately is that, if organizations offset their greenhouse gas (GHG) emissions at all, they should do so only with activities that remove CO2 from the atmosphere – such as forestry projects or geologic carbon capture and storage. The argument appears to stem from a rather formalistic interpretation of a “science based” approach to reducing GHG emissions. In consultation documents put out by the Science Based Targets initiative (SBTi), for example, the underlying rationale seems to be:

  1. Globally, countries must reduce GHG emissions to “net zero” by the middle of the century, with any residual CO2 emissions balanced by CO2 removals.
  2. Organizations setting “science-based” targets should follow a path for reducing their own emissions that mirrors what science says must occur globally.
  3. This means organizations should aggressively reduce their own carbon footprints towards zero emissions, and – if they have residual emissions that cannot be eliminated – they should balance them out with investments in activities that remove CO2 from the atmosphere.

The first premise is indisputable. The second is prudent as a general rule, and – putting aside questions of feasibility and equity – it is a core premise of any responsible approach to carbon neutrality.

The third premise, however, is debatable and does not logically follow from the previous two. By the middle of the century, if the world responds aggressively to climate change, removals should be the only available option for balancing out residual emissions (all opportunities for reducing or avoiding emissions should be exhausted). But it is not yet the middle of the century, and we have a lot of mitigation to accomplish to reach that milestone. An exclusive focus on removal-based carbon offsets today is problematic for three reasons.

  1. Nature-based carbon removals are less reliable than other offsetting options

The most prominent strategies for achieving removals today involve so-called “nature-based climate solutions” (NCS), which sequester carbon in forests, soils, and other terrestrial reservoirs. While these are popular options in today’s voluntary carbon offset market, they present numerous challenges where “environmental integrity” is concerned, including challenges related to additionality, quantification (including emissions “leakage”), local impacts, and permanence.

For these reasons, NCS removals should be treated cautiously – even from the perspective of a strictly formal “science-based” approach. The fundamental issue is that NCS removals cannot reliably balance out fossil carbon emissions over the long run and at large scales. Substituting NCS mitigation for fossil fuel reductions means, in essence, shifting carbon from highly stable geologic reservoirs to more precarious terrestrial ones, which may release carbon to the atmosphere due to natural disturbances – including disturbances induced by climate change itself. This basic fact has led some scientists to object to any use of NCS removals to offset fossil carbon emissions (see here and here, for example).

These scientists may be overstating the case. From a practical standpoint, we need to invest in both fossil- and nature-based mitigation; leveraging voluntary carbon markets for both purposes is a reasonable approach in the near term. From a long-term global perspective, however – the perspective that supposedly informs “science-based targets” – NCS offsets are not a prudent or reliable way to achieve “net zero” emissions.

  1. Geologic removal methods are unproven

The primary means for achieving more “permanent” removals – geologic sequestration through carbon capture and storage (CCS) – is still a largely unproven approach. In particular, geologic reservoirs for CCS are finite, and the total viable long-term storage potential is unclear. This has led a number of commentators to warn against making premature assumptions about the reliability of “negative emission technologies”, and to argue for developing emission reduction targets separate from considerations about sequestration or negative emissions. A prudent strategy for setting a “science-based’ target, then, would be to act as if CCS and other negative emission technologies will not be widely available, even over the long term.

  1. Today, avoiding emissions is just as important – if not more so – than removing them

Regardless of the ultimate role of CO2 removal technologies in global climate change mitigation efforts, we currently still have a long way to go to simply reduce fossil carbon emissions and emissions of other greenhouse gases. By mid-century, “net zero” emissions may indeed require balancing emissions with removals. In 2020, an arbitrary focus on emission removals rather than reductions is missing the trees (what is needed now) for the forest (what – in an ideal world – will be needed 30 years from now). This kind of distraction can have unintended and adverse consequences.


The idea of “science-based” targets is useful as a thought exercise for individual companies and organizations. It can illuminate their role in a sustainable, decarbonized world, at a tangible, “nuts and bolts” level; this can lead to concrete steps and help drive business strategies. But when the idea is taken to logical extremes, the results can be counterproductive. A narrow focus on achieving “net zero” emissions using only removal-based carbon offsets would be just such an extreme.

Ideally, organizations (along with the rest of the world!) will reduce their own emissions in line with what is needed to limit global warming to 1.5°C or 2°C. If they use carbon offsets to achieve even more mitigation, they should strongly consider investments that avoid greenhouse gas emissions – and thus accelerate decarbonization. It does not make sense to limit investment to sequestration offsets based on an idealized notion of what “net zero” emissions should look like in 2050.

7 responses to “Should carbon offsets only include removing CO2 from the atmosphere?”

  1. David Jaber says:

    If you look more closely at SBTi, you’ll see it does not allow offsets at its core and on my read, it is in alignment with what you’re stating. Currently, companies that set targets have to do so with real reductions in operations and supply chain. Where removals may be relevant as “insets” in one’s supply chain.

    • Hi David,

      You are right that SBTi does not currently allow offsets to meet science-based emission targets. However, they are in the process of developing criteria for achieving “net zero” emissions, for which they are contemplating removal-based offsets as a way to compensate for “residual” emissions. See, for example, slide 24 of the consultation document linked to in the first paragraph, above.

  2. Klaus Radunsky says:

    I agree with most of the above. However, several aspects are missing:
    – the magnitude of the expected residual GHG emissions by 2050
    – the magnitude of the various approaches to achieve net negative emissions
    – the significant amount of time required to develop NET at the scale required
    – the need for a joint effort to live up to those challenges indicated above – a large corporate in IT is planning to invest 1 bio USD in the development of CDR, acknowledging that a much stronger effort is needed now! to meet those challenges!

  3. It is a good insight into the “science-based” carbon offset to reach a low carbon economy by the middle of the century. And also NBS is important to if they have residual emission can not be eliminated Them can be a balance with investing for carbon removal project activities. However, I think technology innovation and technology transforming are more important to do so with green financing.

  4. Alex says:

    Great article. Would you be able to double-check the links? Some of them generate a 404 page not found.

  5. Matthew Brander says:

    Thanks very much for this analysis – it’s extremely useful.

Leave a Reply

Your email address will not be published. Required fields are marked *