The principles of the first cryptocurrency Bitcoin were published in 2008; along with them the enabling
blockchain technology was developed. Blockchains build on several pre-existing computing concepts
(such as cryptographic hashes, hashpointers, and consensus mechanisms in distributed systems),
combining and adding to them in an innovative way to provide secure and immutable transactions
without the need of a trusted central authority. Quickly after their inception, it became clear that
extending blockchains with the capability to automatically execute small pieces of code called smart
contracts, extends their potential application domains far beyond cryptocurrencies.

While in the beginning a notion for specialists, blockchain technologies have recently received much
public and media attention due to price surges of several cryptocurrencies, but also due to their ability
to prove and transfer ownership through non-fungible tokens (NFTs), a special type of smart contracts.
At the same time, it quickly became clear that the trustless consensus mechanism needed to register
new transactions and produce new cryptocurrency, called mining, needs a large, continuously
growing, and possibly unsustainable amount of energy.

Given growing concerns about this energy consumption, but also the diversifying application domains
for blockchain technology, this study set out to analyse the different factors that affect the energy
consumption of one blockchain, and to identify the best levers to mitigate this energy consumption as
a consequence. The results confirm that the trustless consensus mechanism based on what is called
proof-of-work dominates the energy consumption of a blockchain by a margin: While it can be
responsible for over 100 TWh per year, the storage of the blockchain requires over 4-6 orders of
magnitude less energy (30 MWh – 3 GWh) and the coordination messages sent across the Internet
over 7 orders of magnitude less (6 MWh). Energy conservation measures must thus address the
proof-of-work consensus mechanism: Individual blockchains can switch to alternative consensus
mechanisms that do not rely on proof-of-work, while company and public policies can aim at
discouraging proof-of-work-based blockchains, and encourage the uptake of blockchains with
alternative consensus mechanisms.