Why U.S. Energy Bills Are Set To Keep Rising
AI DATA CENTERS
- U.S. electricity prices have surged 36% since 2021 and are projected to keep rising, with the EIA forecasting residential prices to hit 17.7 cents per kilowatt-hour by 2026.
- The primary driver of increased costs is a sudden jump in power demand, largely from new data centers in regions like ERCOT and PJM, with overall demand projected to grow by 15% over the next decade.
- The country's heavy reliance on natural gas for 40% of its electricity generation exposes consumers to rising gas prices, though a looming LNG export supply glut could eventually help to moderate costs.
Democrats' pitch on reining in energy prices helped them handily win Tuesday’s elections, with several candidates promising to freeze utility rates. Democrats also notched seats in Georgia’s utilities commission for the first time in nearly 20 years, a clear rebuke of the Georgia Public Service Commission after it hiked electricity rates six times over the past two years. Georgia Power residential customers are now paying an extra $516 per year on their electric bill from two years ago. But Georgia is hardly alone. Surging electricity costs have become a major flashpoint for U.S. consumers across the country, with more than 60% of Americans reporting utility bills as a leading source of financial strain.
After years of relatively stable prices, U.S. electricity prices have climbed significantly, amplifying financial stress for millions of consumers who have been grappling with inflation for years. U.S. electricity prices have surged 36% since 2021, averaging annual increases of about 7% and triple the 12% increase from 2009 to 2020.
But here’s the kicker: U.S. electricity prices are set to continue climbing, with the EIA predicting that residential prices will hit 17.7 cents per kilowatt-hour by 2026, up from 16 cents in 2024.
There’s a method to the madness though, with a sudden increase in power demand largely responsible for higher power prices. After nearly 14 years of stagnation (from roughly 2008 to 2021, with an average annual growth of just 0.1%), U.S. electricity demand rose by 3% in 2024, the fifth-highest annual increase this century.
The surge was largely driven by new data centers in ERCOT (Electric Reliability Council of Texas) and PJM regions. Last year, the power sector consulting firm Grid Strategies published a report titled “The Era of Flat Power Demand is Over,” which pointed out that United States grid planners--utilities and regional transmission operators (RTOs)--had nearly doubled growth projections in their five-year demand forecasts. For the first time in decades, demand for electricity in the U.S. is projected to grow by as much as 15% over the next decade thanks to the proliferation of AI data centers, electrification of the transport and heating sectors, battery production and high-tech manufacturing in semiconductor fabrication.
According to the Electric Power Research Institute (EPRI), data centers will gobble-up up to 9% of total electricity generated in the United States by the end of the decade, up from ~1.5% currently thanks to the rapid adoption of power-hungry technologies such as generative AI.
Second, America’s heavy reliance on gas-powered electricity generation has exposed the country to higher power bills. Natural gas is the most dominant fuel in the country’s energy mix, accounting for 40% of total electricity generation. U.S. gas prices have been surging, with Henry Hub prices up nearly 60% over the past year to $4.33 per million British thermal units (MMBtu). The EIA has predicted that U.S. gas prices will continue rising, with Henry Hub spot prices averaging $4.90/MMBtu in 2026, up from $4.00/MMBtu in 2025 thanks to robust LNG export demand coupled with flat domestic production growth. U.S. LNG exports are projected to continue growing significantly, with capacity projected to grow by about 75% by 2030 from sanctioned projects alone. U.S. LNG capacity is expected to increase from around 17 billion cubic feet per day (Bcf/d) currently to about 30 Bcf/d in 2030.
And, the large increase in U.S. LNG output might eventually help to tame surging power costs.TotalEnergies’ (NYSE:TTE) CEO Patrick Pouyanné recently warned of a looming LNG supply glut in the United States, shortly after Texas-based NextDecade Corp. (NASDAQ:NEXT) announced it has made a positive final investment decision (FID) on Train 4 at its Rio Grande LNG liquefaction plant with a planned total capacity of 48 million tonnes per annum (mpta).
Pouyanné says the U.S. is building too many LNG plants, which could trigger a long-lasting glut if the projects come online as planned. Pouyanné might have a valid concern. Rio Grande’s Train 4 has LNG production capacity of ~6 mpta, bringing the plant’s total capacity under construction to 24 mpta. Meanwhile, NextDecade has revealed that Train 5 is nearing a positive FID while Trains 6-8 are currently in the development and permitting process. Project costs for Train 4 are expected to total ~$6.7 billion, financed with 40% equity and 60% debt. TotalEnergies holds a 10% stake in Rio Grande LNG.
By Alex Kimani for Oilprice.com
Cheap Power Is the Secret to Winning the Global AI Race
- Nvidia CEO Jensen Huang publicly stated that China is positioned to win the global AI race due to its significantly lower energy costs and less stringent regulation compared to the United States and Europe.
- The pursuit of net-zero ambitions has led to high electricity costs in the UK and EU, which directly threatens their plans to become AI superpowers, as AI development is highly dependent on affordable power.
- The surge in demand for electricity from data centers is driving up power prices even in the energy-rich United States, potentially jeopardizing the country's lead in artificial intelligence.
Not long ago, the UK’s Prime Minister declared his government would turn the country into an “AI superpower”. The EU leadership has similar plans for the bloc—while China and the U.S. race ahead without looking back. Neither the UK nor the EU will catch up soon. Their energy is too expensive. Now, Nvidia’s boss is warning that even the United States’ energy costs may be too high to help it win the race.
“China is going to win the AI race.” Jensen Huang made the blunt statement at the Financial Times’ Future of AI Summit this week, going on to list as reasons for this prediction the fact that China enjoyed lower energy costs and, perhaps somewhat surprisingly, less regulation than the United States. He then proceeded to accuse the collective West of “cynicism” with regard to the AI pursuit, calling for some optimism instead. “Power is free” in China, he said, referring to generous state subsidies for the industry—and likely to the fact that China has a lot of low-cost hydrocarbon power generation, unlike, notably, the UK and the EU.
In fairness, Nvidia’s chief executive has a very good reason to be angry with the U.S. leadership. Beijing this week banned foreign microchips from AI data centers receiving state funding. Per the new regulation, new data centers are to only use Chinese-made chips, Reuters reported, citing unnamed sources. It was unclear from the report whether the regulation had a national or local scope, but in either case, Nvidia chip sales to China would be affected by it.
Personal reasons for disgruntlement aside, Huang’s critique of energy policies and overall regulation are quite on point. The European Union is a perfect case in point: it has for years used regulation as a means of stimulating innovation and competitiveness, only to achieve the complete opposite, drawing strong criticism from industries about it.
The UK, meanwhile, is a perfect case in point on energy costs. The country has some of the highest electricity prices in the world because of the net-zero push by several successive governments. AI development, however, depends on low electricity costs. In other words, the UK’s AI superpower dream will likely remain a dream unless the Starmer cabinet finds a way to bring costs down—even though Nvidia recently announced plans to invest $2 billion in fostering an AI startup culture in the UK.
Yet even in the United States, which is a lot more self-sufficient in affordable energy than either the UK or the EU, the artificial intelligence race is driving electricity prices higher, which may jeopardise the country’s success in that race. Earlier this year, the biggest capacity auction in the country, covering a fifth of Americans, ended with a record-high price of $329.14 per megawatt-day. The number was 22% higher than the final price of last year’s auction and reflected the surge in demand for electricity driven by the data center industry. States with high concentrations of data centers are seeing soaring electricity prices that some households are struggling to pay.
Meanwhile, China is subsidizing its AI industry and building new coal plants to ensure stable, reliable, and affordable electricity even as it remains the largest wind and solar installer in the world. Since 2021, the Chinese state has paid an estimated $100 billion in subsidies to the AI industry, Reuters reported this week, after reviewing government tenders in that sector. Nvidia’s CEO has criticized the Trump administration openly for its export curbs on chips, saying it was smart to keep Chinese AI developers hooked on U.S. chips, which, of course, would also boost Nvidia’s profits, but the Trump administration has refused to listen, focusing on what it sees as a risk of the Chinese government using U.S. chips in military technology.
Corporate interests aside, however, it is true that energy costs will determine the winner of the AI race. The challenge for the West is to choose between bringing those costs down and suffering defeat in AI as it remains focused on net zero ambitions.
By Irina Slav for Oilprice.com
