Can a centuries-old technology like hydropower meet AI’s rising energy demands?

Surging demand for the information and processing energy of synthetic intelligence is placing a hidden pressure on U.S. electrical grids.

Generative AI techniques like ChatGPT can every eat as a lot electrical energy in a single day as 180,000 typical U.S. households. The vitality required to coach these massive language AI fashions is much more staggering: Coaching GPT-4 required over 50 gigawatt-hours, about 0.02% of California’s annual electrical energy output and 50 instances greater than what was used for GPT-3.

As AI adoption grows, California’s largest utility, PG&E, forecasts demand may double by 2040.

Shon Hiatt, an affiliate professor on the USC Marshall Faculty of Enterprise and director of the Enterprise of Vitality Transition initiative, sees hydropower as a promising resolution. This centuries-old, clear and renewable energy source is underutilized in the USA, says Hiatt, however may play a vital position in powering the AI revolution—and will assist ease the strain on nationwide electrical grids.

Throughout Inexperienced Week, USC Information spoke with Hiatt to discover how hydropower may assist meet AI’s rising vitality wants and assist a extra sustainable future.

What challenges does the AI revolution current when it comes to vitality consumption?

U.S. electrical energy demand is projected to surge over the following 5 years, with progress charges doubling from final 12 months’s estimates. This sudden improve is pushed by three essential components: the rise of AI data centersfederally backed manufacturing vegetation and widespread electric vehicle adoption.

As a result of information facilities require fixed, dependable energy, renewable sources like wind and photo voltaic cannot present with out large battery backup. Because of this, utilities might want to rely extra on pure gasoline, coal and nuclear plants to satisfy the growing demand.

Wanting forward, electrical energy demand for information facilities is projected to extend by 13%-15% yearly by means of 2030. There’s not sufficient deliberate electrical energy era growth to accommodate projected AI information middle progress.

That is driving Large Tech firms into the vitality sector, as seen in Amazon’s latest $650 million deal to amass a Pennsylvania information middle powered by an on-site 2.5 gigawatt nuclear plant. Because the demand for electrical energy continues to develop, it is clear that the vitality panorama is present process a big transformation.

How can hydropower assist tackle these challenges?

A comparatively fast resolution is repowering present hydropower vegetation and placing generators on present reservoirs.

Hydropower can present baseload vitality not like wind and photo voltaic, that are intermittent resulting from clouds, climate, and so forth.

We’re not constructing any new reservoirs (dams) on this nation. That doesn’t imply we can not improve hydropower from present reservoirs, nevertheless. The U.S. Division of Vitality estimates that as much as 10 gigawatts of vitality may be created by upgrading present powered services. That is one thing that may be finished inside months with capital.

Furthermore, lower than 3% of the greater than 90,000 reservoirs in the USA produce energy. Putting in generators and mills on these reservoirs may present an extra 12 gigawatts of energy. Placing generators on present reservoirs may also be finished in a well timed method—in some states, a matter of months.

What are the benefits of investing in hydropower over different renewable vitality sources to assist AI’s rising demand?

Each vitality supply has a tradeoff. Photo voltaic scales linearly and has thus the biggest land footprint of present energy sources. Wind can have an effect on birds and sea life. Nuclear and mixed cycle pure gasoline have the smallest footprint for vitality output.

Run-of-the river hydropower has a lot decrease environmental impression than reservoir hydropower because it requires no reservoirs. Run of the river diverts some water from a watershed to energy a turbine downstream whereas the remainder of the water flows down the mountain.

The U.S. Division of Vitality estimates that the U.S. has 65 gigawatts of unexploited hydropower vitality that may come from ecologically pleasant run-of-the-river services. Nonetheless, growth of run-of-the-river services can take years to develop resulting from authorities licensing and allowing obstacles.

Frankly, it’s possible that the demand for electrical energy for AI data centers over the following 5 years might be met by new mixed cycle natural gas services. They’re fast to place up, require a small footprint, and the U.S. at present has a big abundance of low-cost gasoline. Small modular nuclear reactors are one other potential resolution. Nonetheless, they don’t seem to be prone to come on-line till 2030 on the earliest.