Russian gas flows long time nonstop. American gas needs constant expensive fracking.
"They used to say in Silicon Valley that software was eating the world. Now it's hardware that's hungry. After decades of preferring bits to atoms, the world's biggest tech firms are pouring hundreds of billions of dollars into building data centers to enable rapidly evolving, energy-hungry artificial-intelligence models.
The country that is best able to harness AI will reap immense benefits in higher productivity and faster innovation. In the U.S., the biggest constraint on AI development isn't ingenuity but electricity. AI models require vast quantities of power. The Trump administration must ensure that its energy policies support the "AI dominance" the president has promised.
Despite tech firms' willingness to pay a premium for clean and reliable energy, the current U.S. grid can't meet the AI challenge. Leading data-center markets in northern Virginia, Georgia and elsewhere are crippled by lack of power. Environmental regulations and grid-permit guidelines mean that major data-center projects have had to be canceled to protect, for example, the habitat of a rare bumblebee. Without more power supply, AI data centers' growing demands could cause a spike in electricity prices, harming American consumers and industrial competitiveness.
The U.S. leads the world in research, innovation and commercialization, but not in building infrastructure -- largely due to political and regulatory barriers. Protectionist tariffs don't fix this.
By contrast, China has both AI expertise and a world-beating capacity for infrastructure development. Beijing already has a plan to build data centers in its sparsely populated interior, and it's investing in power stations at a relentless pace. Despite its mass production of solar panels and batteries, it has no qualms about generating electricity by burning coal. In 2023 China's new coal electricity capacity accounted for two-thirds of the world's total.
Existing power generation technologies can't compete. Solar and wind are cheap and clean but don't provide the round-the-clock power that data centers require -- and they are often deeply dependent on Chinese supply chains, now severed by the trade war. The number of mothballed U.S. nuclear reactors that can be turned back on is low (two, perhaps three). The best sites for hydroelectric dams have long since been developed. Advanced nuclear and geothermal technologies are years away from commercial operation, if they ever prove viable. The only other options are coal or natural gas.
The downsides of coal are well known, despite President Trump's nostalgia for "beautiful, clean coal." Even if you don't believe the climate-change narrative, coal particles impose billions of dollars of cost by spreading respiratory illness. In any case, today's hyperscalers won't go back to coal even if the U.S. government urges them to do so. Gas is different. Thanks to American innovation and carbon-dioxide management techniques [1] developed during the shale revolution of the early 2010s, the U.S. energy industry has learned to manage, transport and capture more than 90% of a natural gas turbine's carbon-dioxide emissions.
The U.S. also has huge reserves of natural gas. Save for a brief spike following 2022 events in Ukraine, U.S. natural gas prices have remained low and stable since the shale revolution.
Natural gas power generation with carbon capture and storage is the only scalable, clean, reliable way to power the data centers the AI revolution requires.
The race to pair natural gas assets and carbon capture for data centers has already begun, bringing in the world's largest oil and gas companies, as well as entrepreneurial new entrants. But the government needs to play its part.
The political will is there. Mr. Trump has promised to "unleash American energy" and "power American homes, cars, and factories with reliable, abundant, and affordable energy." And the Trump administration is stocked with policymakers who understand the challenge and necessity of big power projects, from tech entrepreneurs such as Elon Musk to energy executives such as Chris Wright.
But good intentions and good people aren't enough. There are three steps the Trump administration can take to solve America's electricity challenge. First, it should follow through with its plans to reduce the regulatory burden on natural gas drillers and accelerate the sale of drilling leases on public lands.
Second, Environmental Protection Agency Administrator Lee Zeldin should grant more states -- namely Texas, Ohio and Oklahoma -- Class VI primacy to regulate carbon capture and storage. These states have significant regulatory expertise in their individual geology and infrastructure. But applications to sequester carbon in the ground have been delayed by Democrats' opposition to faster permitting, and by the EPA's inability to review the flood of applications triggered by the 2022 Inflation Reduction Act's new tax credits for carbon capture.
Third, the administration should exempt these new tax credits from its repeal of the Inflation Reduction Act to help scale America's carbon capture infrastructure.
America has the resources it needs to win the AI race with China. By intelligently exploiting its vast reserves of natural gas, it can win and do so much more cleanly than China. The key is to recognize the solutions and execute them. Software is no longer eating the world. The new generation of information technology is eating electricity. If we don't deliver it -- and fast -- China will eat America's lunch.
---
Mr. Ferguson is managing director and Mr. Kumleben director of energy at Greenmantle, an advisory firm whose clients include energy and technology companies.” [2]
[1] To manage, transport, and capture over 90% of a natural gas turbine's carbon dioxide (CO2) emissions, you can use pre-combustion carbon capture combined with an integrated gasification combined cycle (IGCC) power plant, followed by CO2 compression and transport via pipelines or other methods for sequestration or industrial use.
1. Pre-Combustion Carbon Capture:
IGCC Power Plant:
Instead of directly burning natural gas, the IGCC process converts it into a synthesis gas (syngas) primarily composed of hydrogen (H2) and carbon monoxide (CO).
Carbon Capture:
The CO2 is captured from the syngas stream before it's burned, using a process like amine scrubbing (using a chemical liquid like monoethanolamine (MEA)).
Absorption and Desorption:
The MEA absorbs the CO2, forming a CO2-rich liquid, which is then heated to release a pure CO2 stream.
Purification:
The pure CO2 stream is then compressed and transported for storage or use.
2. Transportation:
Pipelines: Pipelines are the most efficient and common method for transporting large quantities of CO2.
Other Methods: CO2 can also be transported via trucks, trains, or ships.
3. Sequestration or Utilization:
Sequestration:
Captured CO2 can be stored permanently underground in depleted oil and gas reservoirs, or saline formations.
Industrial Use:
CO2 can be used in various industrial processes, such as enhanced oil recovery or as a feedstock for chemical production.
[2] Trump Must Deregulate U.S. Energy to Unleash AI Dominance. Ferguson, Niall; Kumleben, Nick. Wall Street Journal, Eastern edition; New York, N.Y.. 09 Apr 2025: A15.
Komentarų nėra:
Rašyti komentarą