A visitor observes a computer bay of the PA10 Data Center, operated by Equinix Inc., in Paris, in France, Thursday, February 6, 2025.
Bloomberg | Bloomberg | Getty images
In certain advanced economies, the electrical infrastructure and the cost of public services are under structural changes due to demand for artificial intelligence data centers.
In the process, American consumers could pay higher public service bills due to the transfer costs of the sector to consumers, warned a final article by the Harvard Electricity Law initiative.
Meanwhile, in the United Kingdom, residents can experience larger wholesale prices in the light of a reform offered on the electricity market which would promote data centers that exploit renewable energies.
As the price concerns emerge, the regulation and reform of the energy network will take the front of the scene in the management of energy prices and the satisfaction of changing energy needs.
“Complex” special contracts
Special contracts between public services and data centers companies are one of the ways in which the higher costs associated with data centers can transfer to everyday consumers, identified a report by the Harvard Electricity Law initiative In March.
These contracts “allow an individual consumer to take services under conditions and conditions that are not accessible to someone else”. In other words, they can be used to move the costs of data centers to consumers due to the subjectivity and complexity of the accounting practices of these contracts, the report indicates.
In addition, special contracts are approved by the Public Services Commission, but tend to undergo “opaque regulatory processes” that make evaluation difficult if the costs have been moved from consumer data centers.
To remedy this, the report recommended regulators to strengthen the surveillance of special contracts or to delete them completely and to opt for existing tariff practices.
“Unlike a unique special contract which provides unique terms and conditions at each data and conditions, a rate guarantees that all data centers pay in the same terms and that the impact of new customers is addressed by considering the complete image of the costs and income of the public service”, according to the report.
Jonathan Koomey, researcher in energy and information technology, combined with the need for data centers to pay according to their use of the energy network.
“The key point, in my opinion, is that very profitable companies that impose costs on the network with large new charges should pay the costs created by these new charges,” Koomey told CNBC.
Beyond public service companies and regulators, “stakeholders in the public service regulation process also play an essential role,” said Koomey.
Stakeholders Allow the commissions of public services to hear a range of stakeholders. They may include a specific group of voters or a large commercial customer who participate in procedures and raised problems on the affordability of services.
“They can often dig deeper than overloaded regulators in projections and technical details and reveal key problems that have not yet surfaced in regulatory procedures,” added Koomey.
Too constructed infrastructure?
Another factor affecting public services prices is the excessive development of energy infrastructure.
Public services and pipelines of the States of Virginia, North Carolina, South Carolina and Georgia provide for one “Major-renforcement of natural gas infrastructure over the next 15 years,“Potentially based on a overestimation of data centers load forecasts, highlighted a report by the Institute for Energy Economics and Financial Analysis in January.
Proactive decisions on the part of public services and regulators are necessary to prevent taxpayers from being “to the catchphrase” for an overly constructed infrastructure, the IEEFA report said.
The political decision -makers of the States have adopted a series of measures to encourage, slow down and regulate the influx of development of data centers, from tax loss to legislative invoices, by emphasizing the guarantee of non -given centers, consumers do not support industrial costs, according to a report by the report by Gibson Dunn Data Centers and Digital Infrastructure Practice Group Group.
Zonal price
In the United Kingdom, data centers and consumers are faced with a different price challenge in the midst of government plans to transform the country's electricity market into a decarbonized, profitable and secure electrical system.
The zonal pricing regime which is explored under the government Examination of the electricity markets Would mark a shift in uniform pricing to a divided electricity market. As part of the new framework, consumers in different geographic areas are subject to different electricity prices depending on the marginal cost of demand there.
Modeling From the consulting company, Lane Clark and Peacock, suggests that the north of Scotland would experience a drop in wholesale prices due to their high renewable penetration and their relatively low demand.
The rest of the United Kingdom, representing 97% of the national demand for electricity, is about to see an increase in wholesale prices of the current national pricing model.
The impact on retail prices is still troubled.
“It is not clear how it can have an impact on retail prices, because wholesale prices are only part of the overall electricity invoice for consumers, and Desnz must still make various decisions”, according to the joint comments of Sam Hollister, responsible for the energy economy, policies and investment and Dina Darshini, commercial and industrial manager at Lane Clark Peacock's Energy Transition Division, LCP Delta.
The Desnz is the British energy security and zero net department.
Will data centers benefit from this?
While technological companies appear on board with the reduce the costs that zonal prices should offer, depending on Search for reflection groups According to Hollister and Darshini, supported by Amazon, Openai and Anthropic, if the data centers would in fact benefit from the zonal pricing of their type of operations, according to Hollister and Darshini.
Those who are potentially well suited to zonal pricing include installations of data centers that manage workloads that can be offset in time or location, they said.
AI training for in -depth learning models is an example. These workloads can be scheduled during off -peak hours when electricity prices can be lower and synchronized with periods of wind energy or solar energy, which would reduce costs and attenuate the congestion of the network.
Likewise, data centers that do not need to be close to the main urban centers or end users – such as those supporting IA hyperscal training, the cloud and large -scale data storage facilities or scientific computer centers – could also benefit from cheaper electricity when located in regions with a high renewable generation and a low local demand, Hollister and Darshini.
However, “all AI workloads are not flexible – real -time inference tasks, such as those used in chatbots, fraud detection or autonomous vehicles, require immediate treatment and do not benefit from time lag”, they added.
Applications sensitive to latency such as financial trading and real -time streaming which require proximity to users would also find “less viable” zonal prices.
Grid infrastructure boost
Partisans of zonal pricing indicate the advantages of reducing the need to move energy over long distances.
But with the plans of the national operator of the energy system to increase network capacity and connect more offshore wind turbines, focus on network infrastructure is important “and zonal prices will not eliminate these requirements”, according to Hollister and Darshini.
“It is not only the data centers that will need this additional capacity on the network, they are probably the most highly publicized, but the EV load will change the grid. National Grid as a organization has spoken of the change in the request profile of electric vehicles for a very long time,” said David Mytton, a sustainable computer researcher, in CNBC.
The requirements on the energy network posed by vehicle electrification is a shared challenge in the United States and the United Kingdom
In the United States, electric vehicles will constitute more than half of all new cars sold by 2030 and should exert considerable pressure on an aging energy network system.
While the electricity consumption of American data centers increases at an increasing pace, a report of National Laboratory of Lawrence Berkeley Published in December noted that this takes place against a “much more important electricity request which should occur in the coming decades from an adoption combination of electric vehicles, the imposition of manufacturing, the use of hydrogen and the electrification of industry and buildings.”
Given this, infrastructure and regulatory reforms that emerge from data centers management would be useful for an imminent era of electricity demand, said, said Mytton and colleagues researchers.