When we started our NFT self-experiment, it did not take long before questions relating to energy consumption came up. According to our estimates, the two transactions (i) minting of the token and (ii) authorisation of an eventual sales processing, produced a CO2-footprint of over 100 kg of CO2EQ. If a tailor-made smart contract would have been employed, this number would have been ten times higher. In short, the minting, transfer and even the sheer existence of NFTs require a lot of energy. But what is the reason for all this energy consumption and are there regulatory instruments to curb it?
Reasons for the energy hunger of NFTs
Very simply put, the reason for the energy hunger of NFTs lies
in the underlying validation system: Today, most NFTs run on
the Ethereum blockchain. The Ethereum
blockchain is a way of conducting and recording transactions
through a decentralised network of computers. Because of the
absence of a centralised authority or ledger, a different mechanism
is required to gain trust: consensus. In the case of Ethereum, the
consensus is currently achieved by the
so-called "Proof-of-work"-method:
Pending transactions are processed into blocks on the blockchain
via intensive computational
work ("mining"). In order to
"mine" a new block, computers on the network (so-called
"miners") compete in solving a complex maths puzzle
through guessing (each guess being a "hash"). To shield
the network from attacks and regulate the speed at which new blocks
are formed, the difficulty of the maths puzzle depends on the total
amount of computational power (the "total hashrate of the
network"). For their computational work, the successful miner
(the miner who correctly guessed the maths puzzles solution)
receives a reward; on the Ethereum blockchain in the form of Ether,
the network's currency.
In short, the more computational work ("hashrate") is
used for the processing of transactions on a blockchain, the more
computational work is required.
Where is the energy used?
The primary factor for a blockchain's energy consumption is
unsurprisingly the energy required to perform the
computational work. In the early days of blockchain
technology, central processing units (CPUs) of PC systems were
enough to power the first blockchain networks. With increasing
interest in this technology (and the total hashrates of major
blockchains), miners started to use the more powerful graphics
processing units (GPUs). While other blockchain networks today are
powered by "application-specific integrated circuits"
(ASICs) with enormous hashrates, the Ethereum blockchain runs an
"ASIC-resistant" proof-of-work algorithm. Thus, Ethereum
mining is still done with GPUs, which are, of course, less
effective than ASICs.
The second factor for the energy consumption
is "non-IT" or "(mining)
support" infrastructure, such as cooling and lighting
or energy monitoring devices.
... how much?
The question of how much energy is required for (i) a blockchain
network in total or (ii) a single transaction on a certain
blockchain cannot be answered with absolute certainty for numerous
reasons.
As for the energy demand of a blockchain,
uncertainties concern the wide range of energy
efficiency in the mining gear used (thus, the total
hash rate cannot be translated into energy demand) or the use and
creditability of support
infrastructure.
As for the energy demand of a single
transaction, additional factors must be considered: The
total hashrate of the blockchain and the difficulty of the maths
puzzle. However, since both factors are known, the uncertainties of
the energy demand with respect to a single transaction on a
blockchain also relate to the energy efficiency in the
mining and the support
gear used.
Often-cited estimates (e.g. digiconomist.net)
reveal an annualised electrical energy consumption of Ethereum
of over 50
TWh (comparable to the power consumption of
Portugal) and electrical energy requirements of
almost 90 kWh
per transaction on Ethereum (similar to the power
consumption of an average US household over three days). However,
looking at the methodology of this "Energy Consumption
Index" more closely shows significant (data) gaps. Thus, it
may also be possible that Ethereum requires an annualised
electrical energy consumption of "just" over 10 TWh.
In short, the energy consumption of blockchain remains an –
admittedly sophisticated – assumption.
The question of greenhouse gas (GHG)
emissions (and thus climate
effects) of blockchain networks adds another layer of uncertainty.
In order to determine GHG emissions of a blockchain or a singular
transaction on such a network, one would have to know the exact
electricity mix used in the computing (and supporting) work on the
network, in addition to the exact energy efficiency of the mining
and supporting equipment and the system boundaries. In principle, a
weighted average of 475 g CO2EQ/kWh is suggested for the
electrical energy used to power e.g. the bitcoin network. This
weighted average would lead to an (annual) GHG footprint of
over 24
Mt
CO2EQ for the entire
Ethereum network or
over 42
kg
CO2EQ
per transaction.
According to another calculation, our NFT self-experiment
– the minting of the token and the authorisation of an
eventual sales processing – required more than 100 kg
CO2EQ. However, both of those estimates fall short if we look at
the "true" GHG-emissions of our NFT. In environmental law
we are used to considering alternatives to
environmental impact. Considering alternatives to an
environmental impact (such as the minting of an NFT), think of an
analogue work of art: A painting or a sculpture, for example
requires various resources, some of which may be even more, others
less scarce. The same is true for the admiration of a painting in a
museum (considering the footprints of the museum itself and its
visitors), and even the transaction of an artwork: Distances may
need to be covered by the parties to inspect the piece of art or to
sign the transaction documents. Would anyone dare to regulate the
permissible environmental impact of a piece of art, estimate the
carbon footprint of all museums or the energy requirements of all
art-related transactions in a certain country or even in the world?
In other words: It is impossible to attribute a certain
environmental impact to an NFT like ours without considering
possible alternatives. Thus, the carbon footprint of our NFT
(around 100 kg CO2EQ) may be much lower considering that
potential admirers do not have to travel to Austria to marvel at it
but may easily access it in the museum online.
Admittedly, we do not know what the exact "alternative"
to "our" NFT would be and how a possible comparison of
environmental impact would end up. But we want to shed light on the
fact that a carbon footprint of around 100 kg CO2EQ is
not as finite as it may appear to be at first sight.
Regulatory (and other) instruments
While the exact numbers of the energy consumption and attainable GHG emissions of blockchains may remain a mystery, the main blockchain's hunger for energy is undeniable. As environmental/energy lawyers, we thus ask ourselves the following questions: 1. How is it possible that almost every other sector is occupied with multiple regulations on energy efficiency and GHG-emission restrictions while in crypto-related projects, energy in general, and energy from fossil fuels in particular is used on a big scale? 2. Is there anything that can be done on a global level? And what can we do to make our (or another) NFT "greener"?
1. Differences between the crypto sector and others
As we have established, mining requires a lot of energy. This is
– at least in the case of the "Proof-of-Work"
method – partly intentional because the maths puzzle must be
complex and the computational work intense to solve it. Also, in
the end it is all about making profit. From a strictly economical
perspective, it is thus preferable for miners to use cheap (fossil)
energy instead of the expensive (green) one.
Moreover, blockchain technology is a true manifestation of
globalisation and decentralisation: Contributors from all over the
world generate a network of trust independent from a fixed location
via the internet and by consensus. In order to regulate a globally
decentralised network, an equally global regulation would be
required.
This brings us to the practical problems with international law.
International law (and especially international treaties) is driven
by its subjects (or potential parties). Since most of those
subjects are sovereign countries or international organisations,
their interests vary immensely. Additionally, the duration of
decision-making processes usually correlates directly to the size
of the parties involved and – of course – the
importance of the decision. Those factors lead to international law
being extremely slow in adapting to new technologies. With respect
to blockchain, international law is in its infancy – and the
environmental impact of this technology is not the first priority
on the list.
At the same time, existing instruments – such as the Paris
Agreement – only provide for targets but leave the means to
achieve them to its Member States. Since the interests of the
Member states still vary immensely, so do their measures. Thus,
miners may still choose a Member State with less rigorous measures
(or even a Non-Member State) to conduct their business.
2. What can be done?
The energy efficiency of a global, decentralised network may
only be increased, and its carbon footprint only be decreased
through equally global regulation. Possible measures may thus
include incentives to produce/use energy from renewable sources,
incentives to use blockchain networks with low energy requirements,
the termination of mining subsidies, global standards for energy
efficiency of mining and support equipment, GHG-emission
restrictions for electricity generation or certificate trading
schemes. If you are a legislator, implementing such measures might
aid the attainment of the abovementioned targets. If you are not a
legislator, possible means to increase the energy efficiency and to
decrease the carbon footprint of blockchain networks include, for
instance, exercising your right to vote, (eventual) popular
initiatives or (eventual) climate actions in your country.
As previously mentioned, most NFTs are created and traded on the
Ethereum blockchain which uses the energy-intensive
"Proof-of-work" concept. However, there are alternatives,
such as the "Proof-of-stake" protocols, which are much
more energy-efficient (like Cardano, Tezos, EOS, Tronix, or Steem).
Ethereum too is currently undergoing a transformation process and
will soon operate as Ethereum 2.0, a "Proof-of-stake"-
blockchain. Moreover, when reflecting on how to make NFTs (even)
greener, the advantages of this technology and the alternatives
must be considered as well. This leaves us with two main roads to
take to a "greener" future of NFTs: The reduction of
energy consumption (and thus GHG emissions) and the substitution of
processes that are currently less "green" than NFTs.
Conclusion
The minting, trade, even the sheer existence of NFTs currently consumes a lot of energy, most of which stems from fossil fuels. Since NFTs run on decentralised networks, regulation may only be effective on an equally global level. However, international law is quite slow to adapt to disruptive technologies. We therefore expect that current technological advancements and individual choices (e.g. of the blockchain used) will outpace regulatory initiatives in the pursue of energy and greenhouse gas emission reductions.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.