A new study into the energy efficiency of Bitcoin-based blockchain networks has found that Bitcoin SV (BSV) is orders of magnitude more efficient than the Bitcoin Core (BTC) and Bitcoin Cash (BCH) protocols.
In its report titled ‘Blockchain technology and energy consumption: The quest for efficiency’, leading Canadian accounting, tax and business consulting firm MNP examines these three Bitcoin-based blockchain protocols and finds the BSV blockchain to be far more energy-efficient.
BSV’s greater energy efficiency is apparent even if the hash power and transaction volume of all networks were equal, as it is a function of a core design difference between itself and the other protocols: the removal of arbitrary restrictions on maximum block size.
The BSV blockchain is built on the protocol most aligned with the original vision of Bitcoin laid out by creator Satoshi Nakamoto. Subsequently, it places no limit on the maximum size of blocks processed or ‘mined’ by the network, and therefore offers unbounded scaling and the ability to process an unlimited volume of transactions.
‘BSV is more efficient due to block size and number of transactions (throughput) currently available on the network and limitations of other protocols. So long as the size or number of transactions on the BSV network exceeds the limitation of the other protocols, BSV is the most efficient in this group,’ the report states.
‘As more businesses and consumers adopt blockchain technologies, and regulatory requirements towards green and renewable energy continue to become more stringent, it is important to understand the impact blockchain has on the environment — especially related to the energy consumption issue.’
Comparing the energy efficiency of the BSV, BTC and BCH protocols requires creating a model to estimate the efficiency of the entire network. That is exactly what MNP has done in its study, building on the work done by previous research into blockchain energy efficiency to estimate the efficiency of each protocol across measurements that include the average energy consumed per transaction and per megabyte (MB) of data.
The Bitcoin protocol and energy efficiency
As each protocol included in the comparison uses a SHA-256-based proof of work consensus mechanism, their relative energy consumption can be measured and projected using MNP’s model. The proof-of-work system requires miners to expend computing power (measured in hashes per second) to solve a cryptographic puzzle in order to validate and add blocks of transactions to the blockchain. The difficulty of this activity can be adjusted dynamically based on the hash power of the miners, who are rewarded for mining blocks with the transaction fees the block contains and a block reward subsidy.
‘The total supply of Bitcoin is limited to 21 million coins, with fresh coins added to circulation via the winning node operator in the amount of the fixed subsidy portion of the block reward. The subsidy portion of the block reward is halved every 210,000 blocks (approximately every four years) first to 25, then 12.5, 6.2, 3.125 and so forth until the full 21 million supply of fresh coins is circulated,’ the report explains.
MNP notes, therefore, that as the number of fresh coins in the network grows, transaction fees will eventually need to grow and serve as a replacement for the regularly halving block rewards.
The energy consumption of mining activities on Bitcoin blockchains has grown significantly as competition for block rewards increases and new hardware is developed to output more hash power. Mining has graduated from an activity completed by consumer-grade CPUs to that of large entities controlling masses of hash power produced by Application-Specific Integrated Circuits (ASICs).
‘Like the difference between CPUs and GPUs, some ASICs are more efficient than others. The earliest ASICs were power-hungry compared to those available today,’ the report states.
‘The equipment decision is the key determinate in how much energy a Bitcoin network will consume in a given time period. This equipment mix also impacts the difficulty metric of a Bitcoin blockchain protocol and the network supporting and running this protocol.’
MNP reached an appropriate approximation of the SHA-256 mining equipment used across all networks by citing previous studies and working with mining companies and publicly available data. From this approximation, it was able to build a model to measure the energy efficiency of each protocol.
Transaction throughput is key to efficiency
Applying this model to the BSV, BTC and BCH protocols, the study found that when it comes to the pertinent measures of energy consumed per transaction and per megabyte of data, the BSV network offers far greater energy efficiency than the others.
This greater energy efficiency is not a product of different equipment or proof of work mechanisms – all three protocols use the same consensus mechanism. The simple reason BSV is more energy efficient is because it can process transactions with a far higher throughput and while expending the same amount of energy per block.
‘Transactions are the ultimate measure of throughput. The number and size of the transactions in a block will affect the size of the block. BTC has a strictly limited block size approaching 4MB. BCH has a much more permissive limit of 32MB. BSV is unbound by block size. Since mining is what consumes energy, and blocks are the product of mining: the more transactions in a block, the lower the energy consumption per transaction. Similarly, the larger a block can be (measured in megabytes), the lower the energy consumption per megabyte,’ the report states.
MNP modelled the energy efficiency of each protocol from Q2 2020 through to Q2 2021, finding that BTC used the most energy over this period by a significant margin.
‘Results indicate BTC consumes orders of magnitude more power than either of the other two protocols tested. The estimated consumption difference between the protocol with the most consumption (BTC) and the protocol with the least consumption (BSV) is at its most (16,041.24 GWh) in Q1 2021, and its least (11,343.25GWh) in Q2 2020 as per our model’s output. Our estimates indicate BTC consumes between 60 and 250 times the power of BSV per quarter.’
Energy consumption per transaction
The study found that the energy consumption per transaction on the BTC network increases significantly over the modelling period, and it remains by far the least efficient blockchain in terms of energy consumption per transaction (measured in kWh/tx).
‘For BTC, the consumption per transaction steadily increases over time. The estimate beginning Q2 2020 has the average consumption at 430 kWh/tx through to 706 kWh/tx in Q2 2021,’ the report states.
For the BCH protocol, the estimated consumption per transaction peaks in the third quarter of 2020 at 183 kWh/tx and falls as low as 6.5 kWh/tx in the first quarter of 2021.
This is significantly lower than the energy consumption of BTC, especially at its minimum, but it is still a far cry from the BSV network’s consistently efficient energy usage.
According to MNP’s model for the BSV network, the energy consumption per transaction ranges between 3.3 kWh/tx in the third quarter of 2020 and 2.4 kWh/tx in the second quarter of the same year.
The differences in energy consumption per transaction for these three blockchain protocols are shown in the graphs below. The second graph excludes BTC to better describe the scale of BSV and BCH’s energy efficiency.
Energy consumption per MB
When it comes to modelling the energy consumption of these blockchain protocols relative to the amount of data they process, the difference between BSV and the others becomes even more clear.
On the BTC network, MNP’s estimate found that the peak average energy consumption per megabyte (MB) is 991 MWh. That is enough energy to power a residential home for almost 100 years.
‘The estimated consumption per megabyte follows the same pattern [as the consumption per transaction], going from approximately 757 MWh/MB in Q2 2020 to 991 MWh/MB in Q2 2021,’ the report states.
The estimated consumption per megabyte for the BCH network is slightly better. MNP’s model estimates a maximum energy consumption of 194 MWh/MB in Q3 2020 and a minimum of 20.5MWh/MB in Q1 2021.
Once again, though, BSV offers a level of energy efficiency leagues ahead of the protocols it is being measured against: MNP estimates that the consumption per megabyte on the BSV network ranged from a peak of 12.63 MWh/MB in Q3 2020 and 0.9 MWh/MB in Q2 2021.
These differences in energy consumption per MB are shown in the graphs below. Once again, BTC has been excluded from the second graph more accurately represent the scale of the difference between BSV and BCH.