VOTE: What is the right global warming potential for Methane versus CO2

[sichen1234]

Currently in our discussions of the Methane Reductions project, we've used the EPA's multipler for methane, which is 25 times the global warming potential of CO2. This is based on the global warming potential of methane over a 100-year period.

The more recent IPCC 6th assessment raised it to about 30x over a 100-year period and more importantly to 82.5x over a 20-year period.

More recently two papers from Stanford has pointed that if we want to keep global warming to 1.5 degrees C, we need to consider the 20-year time horizon, which is the much higher value.

The goal of this vote is to determine what Glboal Warming Potential for methane we should be using in our calculations.

How It Works

This is an experiment for setting up a DAO. Right now we're not using tokens or a DAO yet but testing the processes.

Everybody is welcome to comment.

Based on the quality of your comments, you may be given points (what will one day be reputation tokens in a DAO) to vote on the proposal. So please give thoughtful comments.

Remember the goal is not winning but The Truth.

References:

• EPA: Importance of Methane
• ERCEvolution.com: IPCC Sixth Assessment Report Global Warming Potentials
• IPCC: IPCC AR6
• Stanford Woods Institute: Rethinking how to measure methane’s climate impact
• Environmental Research Letters: Global temperature goals should determine the time horizons for greenhouse gas emission metrics

[hai-yr]

"Applying the methodology in this paper to calculate the implied intertemporal values of a 20-year GWP, a timescale that has received
some recent attention, results in an implicit discount rate of 12.6 % (interquartile range of 11.1 % to 14.6 %). These results provide support for the contention that 100 years is a reasonable timescale choice for the GWP given the assumption that the relative climate damage of pulses of different greenhouse gases is an appropriate means of valuation and that the 3 % discount rate is a reasonable measure of the value of the future" https://esd.copernicus.org/articles/9/1013/2018/esd-9-1013-2018.pdf —this is from one of the papers that the Stanford paper cites. The 20 year (although Stanford recommends a 24 year timescale to be precise, it’s 20 in this paper) timescale places a very high value on changes made in the very near future, and that makes sense for meeting the 1.5 degree target, very drastic changes need to be made immediately. Are we operating under the assumption that the 1.5 degree target is realistic/possible? My other question would be, by placing greater weight on the near-term relative importance of methane, is that sacrificing long-term improvements that are needed in co2 emissions and how might we quantify that?

[hai-yr]

Another take from the CEO of Crusoe Energy:
"I know this is a debate, but it really shouldn't be. CH4 oxidizes into H20 and CO2 with a half life of about 9 years. So what that means is that by year 20, over 75% of the CH4 emitted at year 0 has become CO2 and the majority of methane’s warming impact has already happened. Cully has a 6 month old daughter and he made the analogy that it'd be like picking between two different statistics for his daughter’s average daily diaper usage. You could say that his daughter uses 5 diapers a day over the first 2 years of her life OR you could also say she only uses 0.5 diapers a day for the first 20 years of her life. Which one gives you more meaningful information? Additionally most of the major corporate climate goals are in a time frame of the next 10-30 years, so if we are going to keep warming below 2 degrees, the next 20 years are the most important 20 years."

[brioux]

That's an interesting analogy, and a good pitch for the 20 year GWP impact of "young" emissions. However if we are building a system that tracks emissions over time in a shared ledger, attributing a single GWP to CH4 would both loose accuracy depending on the window of time of interest.
The time factor is important in any applied measure, and will help validate a company's offset activities by ranking (or pricing) offsets and emissions activities based on expected lifetimes.
Take the following example of buying offset credits. If we want to offset the GWP of current methane emissions, buying an offset credit with a 20 year lifetime should match the 20 yr-GWP (not the 100). However, if you buy a more expensive offset credit with an expected 100 year lifetime or more (assuming it could be guaranteed, e.g. a carbon sink at the bottom of the ocean) there is no reason not to use a 100 yr-GWP. You can match the >100 year offset to the 20 yr-GWP of methane, as long as this is clearly reported...
Time (i.e., GWP discounting, lifetime) will always be an important factor, we just need to attribute it accordingly so companies can make informed decisions...

[sichen1234]

Papers like A quantitative approach to evaluating the GWP timescale
through implicit discount rates have a lot of assumptions: what the Damages are, how much the economy will grow, and the relationship between those. Small changes in the assumptions mean huge differences when compounded over this much time. For example, for every $100 of GDP today, 2% growth for 100 years takes it to $724 but 1.5% for 100 years takes it to $443.

But basically what they're saying is that the longer we wait, the more resources we'll have to deal with it later. Is this really true?
Partha Dasgupta wrote Discounting Climate Change about some of the problems with these discounting rates. I remember he made a point like this: We believe discount rates will be positive, but what if climate change and loss of biodiversity reduces available future resources versus today? Then discounting rates should be negative.

Meanwhile the diapers example is a great way to think about it. Yes, lifetime diaper needs are not great, and if you wait 50 years, the baby would (hopefully) become a successful adult who would have more than enough resources to deal with the diapers they used when they were a baby. But could you really wait for 50 years before dealing a baby's diapers?

[hai-yr]

I agree with the note on discount rate uncertainty -- given a relatively weak global economy due to the pandemic (and pandemics are viewed as likely to increase in frequency with climate change), energy/commodity crises, and high debt levels we may already be seeing a reduction of potential (from an economic standpoint at least) to deal with climate change and associated potential shocks.

[brioux]

@sichen1234 thanks for kick starting this.
First, some clarifying points as I formulate my vote.

• The goal is to determine a GWP timescale for methane emissions (CH4).
• We vote is on selecting a 20 versus 100 year timescale, or more precisely ~100x vs. ~30x factor on the ton CO2e per ton CH4 basis.
• An important factor in selecting a GWP is the implicit discount rates used to set climate policy by different stakeholders, @hai-yr thanks for the link.

This vote raises an important issue regarding how tokens are issued in NET. In my view the outcome of this vote should address more than the choice of 100x vs 30x. It is about selecting a reference value that will always be applied when issuing kgCO2e tokens for CH4 emissions.

As long as there is a debate/discussion on GWP (and interest rates for climate impacts), 1 kgCO2e of CH4 will never be truely fungible to 1 kgCO2 as different stakeholders may set different values. If we think about this in terms of exchange rates, if a KgCO2e CH4 token is issued with a GWP of 30x, but a party uses a rate of 100, they will inflate the token amount (3.3x). The important outcome of this vote is making sure our users, that may conform to different GWP factors, are not limited in how they use our token system.

NET is designed to issue multiple token IDs and types. It would be possible to assign an GHG type - e.g., CH4 vs CO2 (or any other), to each token. By doing so, whatever reference GWP value is applied, a user can adjust (inflate/deflate) the kgCO2e emission balances when using a NET interface or contract API. They can select the GWP they choose.

For my vote, a reasonable reference value should be the current standard or global consensus, i.e., the 100 year GWP. This is not an endorsement of the lower factor, i see value in using 20 years from a discount perspective and accelerating investments. It is a question of GHG identification and standardization.

However, I have an alternate proposal to avoid making a potentially biased decision on conforming to a reference GWP. This could be setup as a separate vote: Assign all tokens in kg of gas emitted (or removed) applying no GWP assumption. A GHG type should be assigned to each token ID (CO2, CH4, N2O,... ) with user defining their own GWP factors.

[arezd]

Thank you very much for opening this discussion,

@Bertrand: I think we should not emit only pure-based methane tokens as this would mean we are leaving the issue to someone else not necessarily more competent than us to solve the problem.

If we would propose something: let me propose something at the intersection of truth and consensus.

Governments are alerting on the so-called net zero trajectory by 2050 in the EU (at least). This is why companies have the goal to have net zero emissions.

This means that when someone is reporting its emissions, governments or companies are using these reports in order to CHECK if a country will be aligned with this net zero trajectory.

This means we should not consider a GWP of 20 years or 100 years because of these underestimations all the more true from emissions from 2030 onwards.

The more we will move overtime, the more the GWP of 100 years (and also the GWP of 20 years) will make no sense and we will probably think short-term in the future.

GWP of 100 years has probably made sense today, but no sense in 2040. It completely underestimates emissions of methane from 2040, 2045 or 2050.

We may use in the future GWP of 2050 - current year of emissions. Meaning for 2022, it would make sense to consider 28 years. In 2045, we may use GWP of (2050 - 2045) = 5 years.

Should we do something like this we will overestimate things constantly which at one point makes the quality of carbon credits premium.

[brioux]

@arezd thanks for stirring the debate. I agree with you that the issue should be handled by a competent authority! My point was that the decision should be made not on the open ledger where tokens are issued, but by the experts operating interfaces that use ledger to facilitate trading. You raise an important point that I agree with strongly, what matters here is attributing the time dimensions (GWP discounting, lifetime) accurately. My point is that we need to make all this information very clear to the USERS of our system so they can make informed decision: volume/mass, age, lifetime, etc...

See me response to @hai-yr second comment on matching offset credits to emissions.
Both emission and offsets should have lifetimes (known or expected) that impact how net-zero goals are achieved. E.g. an offset with a 20 year lifetime will not provide the same emission reduction as an offset with a 100> year lifetime. The former would be better matched to the 20 yr-GWP of CH4, the latter the 100 yr-GWP.

[sichen1234]

This is raising a lot of great questions.

@arezd You bring up a good point. I think 100-year time horizon was appropriate in the 1990's. Then it took about a decade to get a treaty (Kyoto), 20 years of ..., and 30 years later it's no longer 100 years. Furthermore, today climate goals such as SBTi are set with trajectories in mind, ie not just by 2050 but with specific goals in between, such as 50% reduction by 2030. With those shorter term targets I think we do need to focus on shorter term impacts.

@hai-yr's example of the diapers is a great way to think about it. Once the baby grows up, they'll be more than able to clean up the diapers from when they were a baby. But could you wait that long with their diapers around?

@brioux You're right that CH4 and CO2 are not truly fungible, and a lot of assumptions go into the time horizon. The difference between a blockchain/web3/DLT and SAAS or enterprise software is that the former is a marketplace where a lot of parties transact, whereas the latter is for a single company to use.

So while we should annotate the tokens as much as possible with details, for example break down the assumptions and the emissions factors in the meta data, we still would need to have a decide on a CH4:CO2 conversion factor for the market as a whole. Think of the choice of a GWP is an "exchange rate" that the market needs to follow, and it needs to be followed on both sides. A too low CH4:CO2 exchange rate would understate the impact of methane emissions from agriculture and oil & gas production, while a too high CH4:CO2 exchange rate would overstate the value of methane reductions.

This is exactly why we need a neutral DAO to set these. Otherwise couldn't you see two industries using two different conversion rates?

[vipinsun]

Instead of this multiplier being a constant, this multiplier can follow a term structure. A term structure is a rate (this could be the CH4:CO2 Conversion Factor CF) versus the term (1 yr, 3 yr, 5 year, 10 year, 30 yr, 100 yr, 200yr). This curve will be downward sloping polynomial (according to what I hear from previous comments). Such a term structure will take care of many of the issues pointed out in previous comments.

A well known term structure model is the yield curve (that is interest rate view as of today).
However several other concepts are contained in this idea.

1. The CF curve of today is based on todays view.
2. The CF curve of tomorrow will reflect the knowledge as of tomorrow (will not be the same as today). Including a truncation of the term.
3. The CF curve of any day can be used to do stress tests or derive certain sensitivities (by shocking the curve).
4. There are more advanced concepts such as the Forward Curve which may not (yet) exist for the CF curve

Interest rates have the luxury of discovery through liquid market instruments and a secondary market. They also have rich sources (a deep and liquid market), however we do not have that yet, so the choice of oracles will be very important for this.
Will be happy to discuss further.

[sichen1234]

I think it is more like a forward commodity or foreign exchange curve, where there is a supply (methane, CO2) and demand (amount of money spent to remove or mitigate against them) over time into the future. That level of detail would make the market much more sophisticated, but we're so far from that in climate. For example, in the methane reductions project, we're working on knowing how much methane is being emitted right now and then who is responsible for the current methane emissions. Perhaps one day, every emitting entity would need to sell their emissions into a forward market, and then pledge to pay for them.

[vipinsun]

The term structure does not have to be a forward curve. Meaning without a forwards market, even with a basic term structure for CF we can do some interesting things. Like I said we could talk more about this basic idea.