<aside> 👀 If you haven't seen carbon.fyi yet, check it out! It's an emissions calculator for the Ethereum blockchain.

</aside>

This page summarizes Offsetra's rationale and methodology for the calculation of CO2e emissions from the Ethereum network. This methodology is behind the Carbon.FYI calculator.

We initially published this work on September 28th 2020, it has since been cited by others in the community. We ask that if you do use our assumptions and approach, or iterate on this work, that you please cite us and reference the data source.

In the case that you can improve our assumptions with more up-to-date data, we welcome the feedback! By working on these issues collaboratively, and helping to inform the community on the impacts and some of the solutions for crypto CO2-equivalent (CO2e) emissions, we can make a pretty big difference.

You can view the tabulated data and calculations here.

Update log

• 24th Jan 2021: Updated total power utilised on the Ethereum network for 2020, yielding an increase of ~900MWh
• 8th March 2021: Removed the 20% buffer that was previously considered potential losses through transmission and distribution (within the power system) and inefficiencies at the point of use (i.e. from heat). Also updated grid emissions intensity for China (see sources spreadsheet). Note, that both of these changes altered our emissions per unit of gas.
• 20th March 2021: Updated the total gas usage and transaction count data to be congruent with the 2020 energy data utilized. Emissions per unit of gas were modelled across three energy usage scenarios.

The Formula

Check out the spreadsheet itself for sources and data points, but below is presented the brief methodology that we followed to find the emissions per transaction on the Ethereum network.

i = country (see country emissions factor tab on the spreadsheet for full list of included countries)

t = total power consumption of Ethereum network (TWh)

pᵢ = proportion of total hash

cᵢ = national power consumption for Eth mining (TWh)

$$ t*p_i = c_i $$

bᵢ = national grid emission intensity (CO2e/TWh)

e = total emissions from eth network (CO2e)

$$ \sum_{i=1}^{n}{b_i \times c_i} = e $$

x = total transactions on Ethereum network in a given year

CARBON.FYI = emissions per transaction on the Ethereum network

$$ CARBON.FYI = x/e $$

The Problem

The Blockchain is a distributed ledger, which contains a record of all transactions, arranged in sequential blocks. The distributed nature means that users cannot spend any holdings twice or manipulate the Blockchain for gain.

The way that manipulation is detected and rejected is through the 'Proof of Work' consensus mechanism that requires network participants (miners) to undertake complex search problems. To add a valid block to the Blockchain, miners use specialized software to solve the problems and are then issued a certain number of Ethereum in return.

Over time, the complexity of these problems has increased in response to more miners, with more advanced hardware. As the frequency of mining has increased, the electricity expended to mine cryptocurrencies has too.

Given the electrical load required to undertake Proof of Work, and the frequency of cryptocurrency transactions, the energy demand from this activity is significant. This mining is therefore responsible for the greenhouse gas emissions proportional to the power used, considering the respective national emissions intensity of the power system.