Bitcoin has an energy issue. It takes a lot of energy to mine. So much so, that one journalist recently likened mining to Big Oil when looking at future environmental consequences. Miners argue that much of this power is generated by renewable energy, such as hydroelectric power, and because of this, concerns about its energy consumption are overblown. But a recent report titled “Renewable Energy Will Not Solve Bitcoin’s Sustainability Problem”, by Alex de Vries of PwC, takes this idea to task, making the case that the energy costs not only aren’t offset by renewable energy, but there are also other environmental costs we don’t fully appreciate yet.

The report identifies a couple of key issues when it comes to how we think about measuring mining’s energy consumption. The first is that we don’t account for all the connected and related energy costs of bitcoin, such as bitcoin ATMs, as well as third parties like exchanges, wallets, and payment solution providers. Even comparing the standard energy cost of an electronic transaction by a bank versus a bitcoin transaction, standard banking techniques consume far less energy than bitcoin. The electricity footprint per bitcoin transaction can range from 491.4 kilowatts per hour (kWh) to 765.4 kWh.

The banking industry, on the other hand, is processing an exponentially greater number of digital transactions, with the average electricity footprint for these transactions reaching 0.4 kWh at most. (Bitcoin advocates may be quick to point out that these metrics don’t include the energy and material costs that go into the creation of physical fiat currency, though.)

A second issue, and perhaps the most important one that de Vries identifies, is that there are large associated carbon footprints and costs for utilizing the cheap hydroelectric power. The Sichuan province of China, where nearly half of the global mining capacity is, according to Coinshares, has large amounts of cheap hydropower due to poor quality power grids that limit the power export capacity of the region. This has drawn bitcoin miners and “energy-hungry and polluting industries trying to take advantage of the low rates,” writes de Vries.

The report contends that the issue with this is that while cheap and renewable on its face, the mining of bitcoin, and the energy needed to do so, is constant, while the amount of hydropower is seasonal.

De Vries draws from an extensive report by The China Water Risk (CWR), a nonprofit tackling water risks, which found that “hydroelectricity cannot be generated year-round’’ due to ‘‘variations in water availability through rain/floods/droughts.’’ This seasonal variability in hydropower is also rising, with CWR reporting that the variability is expected to increase further because of climate change. In Sichuan specifically, the CWR report found that ‘‘the average power generation capacity during the wet season is three times that of the dry season.’’

The end result of this is that power providers have to turn to other places to make up the difference in power supply during the down months, and the solution they generally turn it is coal, according to CWR, which means these powers sources aren’t as “green” as they’re often purported to be. In fact, rather than producing the low carbon footprint of hydroelectric power, the carbon footprint of electricity purchased in Sichuan is more on par with that of natural gas.

Get the BREAKERMAG newsletter, a weekly roundup of blockchain business and culture.

Finally, de Vries identifies the e-waste generated by the regular turnover of mining machines as new, more efficient ones become available, as a major source of environmental concern. Given the most efficient mining is done with application-specific integrated circuits (ASICs), which are optimized for mining, rather than GPUs, there are few repurposing possibilities when are inevitably retired in favor of more efficient hardware. Using modeling to determine the rate at which hardware turnover is likely to happen, the report finds that “the annualized e-waste generation would amount to 10,948 metric tons.” That is on par with the e-waste generated by a country the size of Luxembourg.

“We’ve now got a more complete story on the active life of bitcoin mining machines and beyond,” said de Vries in an email. “We know they required a lot of energy, and we now get a better perspective of that. We also get a more complete view on how damaging that energy use really is, and lastly what damage is caused after the active lifetime of a machine when it becomes e-waste (which had not been examined before at all).”