The hledger wiki suggests accounting for carbon footprint by creating a 'C' unit at a fixed rate to every transaction; the idea is that all money you spend probably leads to carbon emission, and that spending is a good proxy for your carbon footprint.
While this is probably true, it's not really accounting. My energy company sends me a specified monthly bill by email. Using a little Go program, it's easily parsed and turned into these hledger transactions:
2023-02-01 Electricity Costs
expenses:home:energy:electricity
liabilities:emissions:scope2 -99 kWh
liabilities:energycompany -39,60 EUR
2023-02-01 Variable Gas Costs
expenses:home:energy:gas
liabilities:emissions:scope1 -153 cm
liabilities:energycompany -221,85 EUR
2023-02-06 Advance payment;
assets:bank:rabobank:betaalrekening -174,00 EUR
liabilities:energycompany
First the financial aspect: Dutch energy companies collect a fixed advance payment every month. At the end of the year, this is compared against your actual usage, and the balance is something you either get paid back, or have to pay.
That's why the bill is split up; my actual monthly costs increase my (financial) liability. My advance payment decreases it. It's perfectly normal, and even expected, for these to be unbalanced throughout the year.
Then there's the emissions part. Unfortunately, ledger doesn't allow you to use numbers in units. This decreases the possibility of undetected typos, but means I have to write cm instead of m3, and later, more awkwardly, kgcotwo rather than kgco2.
Scope 1 and 2 emissions are borrowed from corporate carbon accounting. In short, scope 1 emissions are emissions you cause directly. Your gas heating, your car, your barbecue. Scope 2 emissions are emissions produced by others, but proportionally linked to your consumption; electricity is the obvious example.
Scope 3 emissions are the most important ones. They make up the majority of our emissions impact, but they're almost impossible to account for. If I buy a rose, it may be locally grown (they grow like weeds in some parts of the Netherlands), it may be grown in a greenhouse heated with natural gas, or it may have been flown in from Africa. Each has a different environmental impact, but there's no way to know, and I don't know of a way to account for them.
commodity 1.000,00 kWh ; kilowatt-hours of electricity
commodity 1.000,00 cm ; cubic meters of gas
commodity 1.000,00 kgcotwo ; kilograms co2
P 2023-01-01 cm 1,78 kgcotwo ; this is correct for Dutch low-caloric gas.
; If you're elsewhere in the world, this number
; is different!
P 2023-01-01 kWh 0,43 kgcotwo ; this number is correct for the Dutch national
; energy grid in the past few years. If you're
; elsewhere in the world, this number is different!
The above commodity definitions tie it all together. Gas converts to CO2 at a fixed rate. The carbon impact of electricity production is actually highly variable, due to the changing nature of our energy mix (wind and sun are variable, and in some cases nuclear power plants are taken offline), but in most countries you should be able to obtain a yearly average.
The above means I can now do:
hledger -X kgcotwo bs
...
liabilities:emissions:scope1 || 679,96 kgcotwo
liabilities:emissions:scope2 || 98,04 kgcotwo
...
How useful is this? Not in an obvious way. Like I mentioned, my scope 3 impact is likely higher than scopes 1 and 2 combined. This makes the method precise without being accurate; the linked method of equating the amount of dollars spent to carbon impact is probably far more accurate.
It does offer some insights; the carbon impact of simply heating my home vastly outstrips the energy usage of any gadgets and household appliances. If I want to reduce my impact, I probably shouldn't try to replace my fridge; I should fix my home! And I think that this kind of decision making is what accounting is supposed to help with.