What is CO2e? Global warming potential explained, and which GWP values to use

CO2e means carbon dioxide equivalent: a single unit that expresses every greenhouse gas in terms of how much carbon dioxide would cause the same warming. It exists because companies emit seven regulated gas families, not one, and adding kilograms of methane to kilograms of refrigerant would be meaningless. The conversion factor is called global warming potential, or GWP, and it tells you how much heat one kilogram of a gas traps over a chosen time horizon compared with one kilogram of CO2. Multiply the mass of each gas by its GWP, add the results, and you have tonnes of CO2e.

That much is on every glossary page. What almost nobody checks is which set of GWP values their spreadsheet is using, and that quiet detail changes reported emissions by several percent on the lines it touches. The values are revised every time the IPCC publishes a new assessment report, and plenty of working files still carry numbers from 2007.

How global warming potential actually works

GWP compares two things about a gas: how strongly it absorbs infrared radiation, and how long it stays in the atmosphere before breaking down. Carbon dioxide is the reference point and is defined as 1. Methane absorbs heat far more effectively but only survives about a decade, so its GWP is high over a short horizon and falls as the horizon lengthens. Nitrous oxide is both potent and long-lived. Industrial gases such as sulfur hexafluoride sit in a different league entirely, thousands of times stronger than CO2 and lasting for millennia.

The time horizon is a choice, not a fact. The default in corporate reporting is 100 years, written GWP-100, and that is what the GHG Protocol, the EPA and every disclosure regime expect unless they say otherwise. A 20-year horizon, GWP-20, makes short-lived gases such as methane look far more damaging, which is why you sometimes see methane quoted at 80-odd rather than around 30. Both numbers are correct. They answer different questions. For a corporate inventory, use GWP-100 and say so.

Which GWP values to use in 2026

Use the IPCC Sixth Assessment Report, AR6, unless a program you report into specifically requires an older set. AR6 is the current science, the GHG Protocol publishes a consolidated GWP values table covering AR4, AR5 and AR6 so you can align with whichever a scheme mandates, and most voluntary and regulatory frameworks have moved or are moving to AR6. Here are the values that matter to almost every corporate inventory.

Two things in that table are worth pausing on. First, AR6 split methane into fossil and biogenic for the first time, because methane from fossil sources eventually oxidizes into fossil CO2 and carries a small extra penalty. Natural gas leakage takes 29.8; methane from landfill, wastewater or livestock takes 27.0. Older reports had one value for both. Second, the numbers do not simply increase over time. Nitrous oxide fell between AR4 and AR5 and rose again in AR6. You cannot guess the direction of a restatement, which is exactly why the version has to be documented rather than assumed.

The clearest real-world case of two frameworks wanting different values for the same physical gas is a US facility that files with EPA and also answers CDP. EPA's Part 98 Table A-1 has used AR5 since January 1, 2025, so methane is 28 there, while CDP expects AR6 and its 29.8. Neither is wrong; they adopted different assessment reports. What is wrong is moving a figure from one to the other without restating it. The federal side of that is on EPA GHG reporting.

The methane trap that quietly understates emissions

The most common GWP error in practice is inherited rather than chosen. An inventory spreadsheet gets built one year, methane is entered at 28 from AR5, and the file is copied forward every year after that. Nobody revisits the constant because it is not the kind of thing that looks stale. Reporting fossil methane at 28 instead of 29.8 understates every methane-bearing Scope 1 line by roughly 6%, consistently, in the same direction, year after year.

On its own that rarely changes a headline total, because methane is a small share of most corporate footprints. It matters for a different reason: an assurance provider testing your calculation will find it immediately, and a systematic, undisclosed factor error is precisely the kind of finding that turns a routine review into an extended one. The fix is trivial if you know where your constants live, which is why serious inventories keep a documented chain from each reported figure back to the factor and version that produced it, the discipline data teams call lineage for reference data.

Where GWP actually changes your numbers

GWP only bites where you are converting raw quantities of non-CO2 gases yourself. That is mostly Scope 1. Fuel combustion produces small quantities of methane and nitrous oxide alongside CO2. Refrigerant leakage from HVAC, chillers and cold storage is pure high-GWP gas and is often the largest GWP-sensitive line a company has. Fugitive methane matters in oil and gas, waste and agriculture. Process gases such as SF6 dominate the inventories of electrical utilities and semiconductor manufacturers.

In Scope 2 and most of Scope 3, the conversion has already happened for you. Grid electricity factors and spend-based input-output factors are published in CO2e, with the GWP applied by whoever built the factor. That is convenient and slightly dangerous: you inherit their GWP version without seeing it. If a factor library states the assessment report it uses, record it. If it does not, note that too, because an inventory that mixes AR5-based purchased factors with AR6-based direct calculations is normal, but it should be disclosed rather than discovered. Our guide to which EPA emission factors to use covers the source libraries for each scope.

What is the GWP of CO2?

The global warming potential of carbon dioxide is 1, in every IPCC assessment report and every time horizon. That is not a measurement, it is the definition: CO2 is the reference gas the whole scale is built against, so one tonne of CO2 equals one tonne of CO2e by construction. Every other gas is expressed as a multiple of it, which is why fossil methane at 29.8 means one tonne of it warms roughly as much as 29.8 tonnes of carbon dioxide over 100 years.

The practical consequence is that a footprint made almost entirely of CO2, which is the case for a company whose emissions are mostly purchased electricity and fuel combustion, barely moves when the GWP set is updated. A footprint with meaningful methane or refrigerant content moves a lot. That is the real test of whether the GWP version you picked matters to your number.

What is the unit of global warming potential?

Global warming potential has no unit. It is a dimensionless ratio: the heat trapped by one kilogram of a gas over a chosen period, divided by the heat trapped by one kilogram of carbon dioxide over the same period. What carries a unit is the result of applying it. Mass in kilograms or tonnes multiplied by a dimensionless GWP gives kilograms or tonnes of CO2e.

This trips people up in spreadsheets more often than it should, because a GWP column with "kg" in the header invites someone to convert it. The time horizon is the part that does need labeling. GWP values are almost always the 100 year figures, written GWP100, and the 20 year figures are far higher for short-lived gases like methane. Say which horizon you used, because the two are not comparable.

Is CO2e the same as CO2?

No. CO2 is one specific gas; CO2e is a converted total covering all greenhouse gases, expressed in the equivalent mass of CO2. A company reporting 10,000 tonnes CO2e might have emitted 9,400 tonnes of actual carbon dioxide plus quantities of methane, nitrous oxide and refrigerants that convert to the remaining 600. The distinction matters when someone asks you to break the number down by gas, which most disclosure formats now require, and it matters when comparing against a target expressed in one unit or the other. Report CO2e as your headline and keep the gas-by-gas breakdown behind it.

How to document GWP so it survives a review

Three lines in your inventory management plan close this issue permanently. State the assessment report version you use, AR6 for a new inventory. State the time horizon, GWP-100. State whether your values include climate-carbon feedbacks, since AR6 publishes both variants and the feedback-inclusive numbers are higher. Then apply the same choice across every year you report, including your baseline.

That last point catches people out. Changing GWP version is a methodology change, not new information, so it does not mean your emissions went up. If you move from AR5 to AR6, recalculate your baseline year on the same basis and disclose the restatement, otherwise your trend line shows a jump that is pure arithmetic. The same rule applies to any factor library update large enough to cross your significance threshold.

None of this is difficult once the plumbing is right. The problem is that GWP constants tend to live inside spreadsheet cells nobody owns. Our carbon accounting software keeps the factor and its version attached to every calculated line, so switching assessment report is a setting rather than an archaeology project, and every figure can be traced back to the invoice and the factor that produced it. For the full first-inventory sequence, see the GHG inventory checklist, and for the rulebook underneath all of it, the GHG Protocol Corporate Standard. Disclosing which GWP set you used is a basic expectation of carbon reporting software, not an optional footnote. If your reporting trigger is California, SB 253 reporting software covers what that regime expects.

CO2e and GWP FAQ