Even if AI is not as much in the form of an energy hog, people are excluding it, it can still face trouble for the power grid across America.
Tech companies are already burning in the increasing amount of electricity to train and run the new AI model. And they are asking for too much electricity because they try to get each other out. This increasing demand is already starting to reopen the energy system, with utilities to build new gas plants and pipelines.
But all these plans to re -shape the US energy system can be based on the AI bubble. Ovexed investors are afraid to remember Bandwagan with pumping money in tech companies, but still at the risk of developing AI tools that eventually flop, utilities also face a wave of speculation over data center energy needs.
“Uncertainty is unnecessary”
Uncertainty is unnecessary given the costs that can pay on the payment of Americans Higher utility bill More pollutionrecently Report Warns. A transition to cleaner and more affordable energy sources is slowly progressing in the US. It is in Peril as long as the tech companies and utilities demand more transparency and choose for more renewal such as solar and wind energy.
“AI Boom provides exciting opportunities, there are several risks not to get closer to energy needs with a deliberate and informed response, which takes into account long -term effects,” Kelly Polay, the lead author of the report published this month published by the shareholder advocacy group, as you said by the shareholder advocacy group.
The country’s fleet of gas-powered power plants will almost increase if all the new gas projects proposed between January 2023 and January 2025, as the generic AI industry was heated, to flourish. The amount of new gas capacity which was proposed by Utilities and Independent Developers, increased by 70 percent during the time frame, which operated in large stakes due to increase in data center power demand.
Before the generous AI Boom, the demand for electricity was very high Over a decade With energy efficiency gains. But soup-up new data centers for AI are much more energy than being in the past. A rack of computers in a traditional data center can use 6-8 kW power-equal to the power used by three houses in America, Dan Thompson, a leading research analyst of S&P Global, stated in briefing. However, AI requires more powerful computer chips to run more complex tasks. According to Thompson, the power required to run one of those high-gradation racks is equal to about 80 to 100 houses, or above 100 kW.
“Essentially what you are seeing is being deployed the power of a small town,” he said.
Why does he talk? Power grids originally act as an uncertain balance act. If the power supply cannot meet the demand increase, it can lead to high utility bill and potential outage. On the other hand, overworting new capacity risks, who make trapped property who pay to the utilities and their customers, pays out whether they really need them in the long term. Therefore, it is so important to try to achieve accurate forecast of future demand.
“Essentially what you are seeing is being deployed the power of a small town.”
And while AI uses a lot of energy, the estimates for the future get marki. The report said, “The bookie market is flooding.” The power has long been trying to get ahead of the time to connect to the grid, some of them are requesting power to the speculators, before they line the capital or customers, to ensure that they can bring a project into a finish line. Some double or triple counting (or more) may also occur when it comes to forecast AI energy demand as developers get closer to more than one utility to receive many quotes.
In the south -east, a major center for data centers, utilities are estimating more and more in the form of four times more demand increase than industry trends, According to a report by Institute for Energy Economics and Financial Analysis (IEEFA) earlier this yearAt the national level, the utilities are preparing to increase 50 percent more demand than the technical industry, one is different. Report From December 2024 states.
Utilities have recently identified this risk on earning calls. Proposed projects attempted to join the grid “can be oversted anywhere from three to five times that can actually be physical,” said Jim Burke, CEO of Texas -based Vustra Energy, said in a Q1 earnings. call out This year.
Despite uncertainty, they are still building new gas power plants and pipelines to meet that demand. Eventually, the construction of new infrastructure is one The most attractive ways for a utility to increase profitAnd right now, the Trump Administration – which was a campaign Surrounded by contribution of oil and gas – Fossil is encouraging dependence on fuel. In Louisiana, for example, local utility entry Proposed three new gas plant To give strength to a huge new meta data center. The data center estimates to consume more electricity as 1.5 million houses and lead 100 million tonnes of carbon emissions in 15 years.
This is an opposite opposite from the biden administration Target Getting a power grid to run 100 percent carbon pollution-free energy by 2035. The only way to prevent climate change in its tracks is to get rid of planetary fuel from planetary fuel. The construction of a congestion of new gas infrastructure clearly leads the nation in the opposite direction.
There are solutions to reduce all these risks, as you sow and the Sierra club explains in its report. Utilities may require developers to disclose the number of other utilities that they have brought their data center proposal and how far they are in finalizing a project. When adding contracts, they may also require long-term service agreements, increase non-individual deposits, and can increase the fee for cancellation of a project.
Tech companies clearly have a large role to improve the energy efficiency of their technologies and invest in renewable. Over the years, there have been technical giants including Amazon, Meta and Google Top corporate buyer Of renewable energy. Adding those types of long -term agreements to build new solar and wind farms can now have even more impact, while combating the rollback of financial incentives for renewal of Trump administration, if companies are ready to prefer their own stability goals, as much as their AI ambitions.
0 Information
