The estimated increase in artificial intelligence and its unprecedented demand for electricity for electricity presents a serious challenge for the financial and technical capacity of the US utility system.
Appreciation for that challenge is lost in the coming decade as a forecast for a huge increase in demand for electricity in the coming decade.
The idea of building a data center that will attract 1 GW power or more, adequate amount to serve more than 875,000 homes, so many data centers are in the developers’ plans and therefore discussed regularly that it no longer seems extraordinary.
The challenge, when seen in the clan, can be heavy.
recently Wood McCenzie Report Currently, 64 gWs of data centers confirmed on books with another 132 gigawatts can be identified. 64 GW is enough to give strength to 56 million houses – more than double the population of 15 largest cities in the US.
The US Electric Utility System is struggling to meet the estimated energy needs of the AI industry. The problem is that many utilities do not have financial and organizational resources to create new production and transmission features on the desired timeline of data and data center developers.
Now the question of public policy on the table is that for whom the risk should be paid for these large-scale mega-energy projects.
Will it be AI developers like Amazon, Microsoft, Meta and Alphabet Joint market price Are the entire S&P 500 utility sector seven times – or residential and other customers of local electric utilities?
The process of answering this and related questions is going on in the Hallway of the US Congress, in the Federal Energy Regulatory Commission and other federal agencies, in the tariff proceedings before the state regulatory officers and in public debate at the national, state and local levels.
Whether they develop federal, state or local level, the following values and objectives should be made the origin of public policy in this field:
- Data center developers who require large amounts of electric power (such as 500mw or any other specified level) must pay for the construction of new generating and transmission facilities. State of Texas recently enacted Legislation This requires data centers and other new large users to fund the infrastructure required to meet their needs. Although it is customary to spread the cost of new features on the user basis of a utility, data center developers are placing on utility systems across the country, those developers are sufficiently extraordinary to allocate the costs of new features. In addition, data center developers have financial resources to cover those costs and include them in the rates charged for users of their AI services.
- Developers of large data centers should bear the risk associated with new utility-made production and transmission features, not utility. As an example of such a policy, Ohio’s Public Utility Commission approved only one. Ohio’s agreement proposed by Ohio’s American Electric Power This will require data centers with loads of more than 1 GW and mobile data centers, which is more than 25 MW to be committed to a 10-year electric service contract and pay minimum demand fee based on 85 percent of their contract capacity, which is more than 60 percent under the current general service tariff of utility. Another option includes Texas legislation The planning process requires significant up-front payment and suggests that data center developers reveal that they can simultaneously keep demands for power. This is not uncommon to withdraw data center requests for service after deciding on the best place and encouragement package once. Data center developers have the financial ability and ability to manage this risk, not utilities.
- Facilities that generate co-located features at large data centers should be integrated with local utility electric grids with appropriate cost allocation.Although some projects have examined the option of a co-located power generation “island”, however, the grid is completely independent of the grid, most of the projects intend to interconnect with the back-up power and the grid system for related purposes. Properly managed, this interconnection can be beneficial for both data centers and utility systems, provided that the cost is appropriately allocated throughout the system.
- The US government should continue to support the development of nuclear technology including small modular reactors.American utilities do not have financial resources to assume the risk of building new nuclear-managed features. The emergence of a new set of customers, the data center developers with heavy needs for electric power and deep pockets, change the equation. The US government has supported billions of dollars for new nuclear technologies and should continue to do so with the aim of bringing down its cost.
- The US government should continue to support the improvement in energy efficiency in data centers. Data centers use huge amounts of electricity for server, cooling system, storage systems, networking equipment, backup system, safety system and lighting. National renewable energy laboratory Has developed A “handbook” of measures that can apply data centers to reduce energy use and to achieve savings. In addition, now there are strong market forces to develop new super-skilled chips that will reduce the unit costs of training and use AI models. The US government should take advantage of American electricity demand and help accelerate the development of these chips.
This public policy debate on our energy future may not be much. If we find these policies correct, AI has the ability to remake the US economy and the energy of this country.
If we consider it wrong, pushing to provide new production and transmission facilities to provide electric gigavat has the ability to overwhelm the financial and operating capacity to our electric utility system, increase the cumbersome rate on homeowners and businesses, which increases the use of fossil fuel to reduce the use of parlokikas and reduce the use of fossil fuel and reduces our electricity.
David M. Claus is an advisor to energy issues, who served as a deputy underscore of the US Energy Department during the Obama administration and as a political appointment for two other democratic presidents. Mark McCarthy is the author of “Regulating Digital Industries” (Bruckings, 2023), an assistant professor in the Communications, Culture and Technology Program of the University of Georgetown, is a Nonscent Senior Fellow at Institute for Technology Law and Policy at Georgetown Law and a Nonscent Senior Fellow in Brokening Institute.
