Hydroelectric dam turbine propeller illustrating hydropower’s role in Bitcoin mining

Cambridge Preview Pairs Cleaner Bitcoin Mining Mix With Higher Emissions

By Isaac • July 26, 2026 10:28 am • Comments

Bitcoin mining’s reported power mix became cleaner while its estimated appetite for electricity climbed sharply, leaving annualized emissions higher despite a larger low-carbon share. That tension is the central finding in preliminary Cambridge data presented in Dallas.

The figures were presented by Alexander Neumueller of the Cambridge Centre for Alternative Finance at the Energy Investors Forum on July 23, 2026. They are preliminary annualized estimates, and the full next edition of the Cambridge Digital Mining Industry Report is expected later in 2026.

At the December 2025 measurement point, annualized electricity consumption was estimated at about 190 TWh, 38% above the 138 TWh estimate from June 2024. Annualized means a run rate at that date, not exactly 190 TWh consumed during calendar 2025.

Trending: BitMart’s Shutdown Clock Starts July 26 and Ends in January

The Energy Mag reported that Neumueller presented the early research at the Dallas forum, giving the industry a preview of the Cambridge Centre’s next mining study. The reported snapshot pairs a rising low-carbon share with a much larger annualized electricity estimate, so improvement in the mix did not translate into lower estimated emissions.

Low-carbon sources accounted for 59.4% of miners’ reported energy mix in the preliminary data, up from 52.4% in the previous comparison. Hydropower overtook natural gas to become the largest single reported energy source, changing the top position within the mix as well as the broader low-carbon share.

Estimated greenhouse-gas emissions nevertheless increased about 20%, moving from roughly 40 million to 48 million tonnes of CO2 equivalent. The preliminary explanation is arithmetic at system scale: estimated electricity consumption rose faster than the reported mix became cleaner.

The same reporting said only 10% of surveyed miners had allocated power to artificial intelligence despite broad interest in diversification. Every figure comes from preliminary research presented at the forum, and the finalized Cambridge Digital Mining Industry Report is expected later in 2026.

The two consumption figures describe annualized conditions at different measurement dates, June 2024 and December 2025. They should be read as comparable run-rate estimates, not as meter readings that assign an exact amount of electricity to either calendar year.

That distinction is especially important around the 190 TWh figure. It estimates what a full year would look like at the December 2025 rate; it does not say the global Bitcoin mining network consumed exactly 190 TWh between January and December 2025.

The 38% rise therefore measures the change between annualized estimates, with 138 TWh serving as the June 2024 comparison point. It captures the direction and scale presented by Cambridge without converting a dated run rate into a finalized historical total.

The energy-mix result also reflects what miners reported in the preliminary research. A 59.4% low-carbon share is evidence of a cleaner reported composition than 52.4%, but it is not an exact, continuously metered inventory of every mining operation worldwide.

Hydropower’s move ahead of natural gas gives that composition shift a concrete center. The presentation identified hydro as the largest single reported source, while the wider low-carbon category grew from 52.4% to 59.4% between the compared snapshots.

A cleaner percentage can coexist with more emissions when total electricity consumption expands more rapidly. In these preliminary figures, estimated consumption rose enough that estimated greenhouse-gas output increased despite the larger reported low-carbon share.

The emissions estimate moved from roughly 40 million tonnes of CO2 equivalent to 48 million tonnes, an increase of about 20%. Those are modeled, reported estimates from the preliminary presentation, not exact audited emissions totals for every miner.

The electricity estimate grew from 138 TWh to about 190 TWh over the stated measurement interval. Cambridge’s preliminary presentation attributes the higher emissions alongside the cleaner mix to electricity consumption rising faster than the composition improved.

The AI finding adds a narrower view of miners’ diversification activity. Although interest was described as broad, only 10% of surveyed miners had allocated power to AI, a survey result that does not establish how the remaining operators will use their capacity later.

That 10% figure belongs to surveyed miners, just as the energy mix is a reported mix. The data remain a research preview and cannot serve as a finalized census of global facilities or a set of exact metered totals.

The presentation still provides a coherent picture of the direction Cambridge observed. Reported low-carbon sourcing gained share and hydropower moved into first place, while the network’s estimated annualized electricity use and greenhouse-gas output both rose.

The full Cambridge report expected later in 2026 will provide the completed edition behind this preview. Until then, the Dallas figures are best identified by their actual status: preliminary annualized estimates presented at a forum on July 23.

That status does not erase the tension in the early findings. It defines the confidence boundary around them, keeping 190 TWh as a December 2025 run-rate estimate and 48 million tonnes as an approximate emissions estimate.

For miners and energy investors, the preliminary mix data show why source shares and total electricity consumption need to be read together. A higher low-carbon percentage can accompany higher estimated emissions when the overall electricity estimate rises by 38%.

The timing also prevents the presentation from being mistaken for Cambridge’s finished 2026 publication. The research was previewed at the Energy Investors Forum, with the full next edition of the Cambridge Digital Mining Industry Report still expected later in the year.

The early result combines measurable improvement in the reported power mix with increased estimates for electricity use and emissions. Its most responsible reading preserves both sides of that finding and the preliminary, annualized nature of every headline figure until Cambridge releases the full report.

Join the conversation!

We have no tolerance for comments containing violence, racism, profanity, vulgarity, doxing, or discourteous behavior. If a comment is spam, instead of replying to it please click the icon below and to the right of that comment. Thank you for partnering with us to maintain fruitful conversation.

 

Search ProCoinNews