Energy Considerations at the Dawn of Strategic Manufacturing

A new era of U.S. industrial policy has begun thanks to three pieces of recent federal legislation: the Infrastructure Investment and Jobs Act (IJJA), the CHIPS and Science Act (CHIPS act), and the Inflation Reduction Act (IRA). A wide range of industrial and manufacturing sectors have been targeted for public support based on an emergent strategic consensus on the need for geoeconomic competition with China, increased emphasis on the security of supply chains, and a belief in the broad social benefits of renewed U.S. industrial and manufacturing capacity. Already, this new environment is driving a surge in private sector investment.

Two sectors stand out. The semiconductor sector has attracted about $200 billion in private investments for the construction of new or expansion of preexisting fabrication facilities (“fabs”), while the EV battery sector has drawn almost $75 billion in investment announcements since Q3 2022. Both are manufactured goods that deliver strategic benefits beyond their market value. Semiconductors are the backbone of almost all modern technology, but U.S. import dependence for this critical technology has left the United States vulnerable to increased national security risks and supply chain disruptions, as seen in the automotive industry chip shortage in 2021–2022. EV batteries are a key component in reaching U.S. net-zero emissions goals and supporting the U.S. military in both its tactical and operational demands. As the U.S. automotive industry eyes a transition to electrification, its competitiveness is challenged by low-cost, high-performance EVs from China, where the EV battery supply chain is most developed. This creates political, economic, and security challenges.

An overlooked but critical question is the energy requirements of these new industries. Semiconductor fabs and battery manufacturing plants require vast volumes of energy and directly compete for energy resources alongside other growing sectors such as minerals mining and processing, hydrogen production, and, above all, AI data centers.

In ascertaining the energy bill of this growing strategic manufacturing, the lack of public and contemporary data is a major hurdle. For both fabs and EV battery manufacturing factories, for example, there is limited public data on energy consumption and estimations in academic literature vary due to differing methods and data sources. What information is available highlights the large amounts of energy necessary for these critical industries. The first phase of TSMC’s new semiconductor plant in Arizona will consume roughly 200 MW; however, the planned expansion of the plant could grow to consume over 1 gigawatt (GW) of electricity according to filings for transmission system expansion from Arizona Public Service, the electric utility for the state. Put in perspective, that is roughly equivalent to the electricity production capacity of the new Vogtle 3 nuclear reactor.

In order to estimate the amount of energy needed for EV battery manufacturing, the authors conducted an academic literature review and averaged several studies’ estimates—including Florian Degen et al., Florian Degen and Marius Schütte, Q. Dai et al., and Simon Kurland—finding that there is an average energy requirement of 44 kWh to produce 1 kWh of battery capacity. Given that new U.S. battery manufacturing plants average 23 gigawatt-hours (GWh) of capacity production a year, these facilities may consume roughly 115 MW. With 45 domestic battery manufacturing plants already announced, the energy consumption of these facilities is almost 5,200 MW, or over 45,000 GWh per year. However, the academic studies from which these numbers come are largely over 10 years old, so contemporary energy consumption likely differs due to the increased scale of the industry and technological improvements.

The U.S. Energy Information Administration surveys manufacturers to understand the energy consumption of the manufacturing sectors, and the most recently available data, from 2018, is shown in the table below. The table shows that the sectors covering semiconductors and lithium-ion batteries, part of “Other Computers, Electronics, and Electrical Equipment,” were relatively small energy consumers compared to the top five most energy-intensive sectors. Nonetheless, either could be among the largest sectors by energy consumption as both private investment and policy support push continued expansion.

In March, Secretary of Commerce Gina Raimondo announced that U.S. investments will put the United States “on track to produce roughly 20 percent of the world’s leading-edge logic chips by the end of the decade,” a dramatic increase given that the country currently produces none. The domestic lithium battery market could grow by “another factor of five to ten by 2030,” as the Biden administration pursues its goal of having “50 percent of all new vehicle sales be electric by 2030.”

Remote Visualization

These strategic ambitions in both sectors imply an energy system that can deliver increased volumes without undue delay, while maintaining system-wide affordability and progress on decreasing emissions intensity. Most of the energy demand from these two industries comes in the form of electricity, which figures into the broader nationwide electricity demand growth challenge. Delivering the additional generation resources and transmission system investments to meet this electricity demand growth is a major policy challenge, but one which primarily sits at the state level.

At the federal level, neither the IRA nor the CHIPS act imposes energy use or emissions intensity requirements for tax credit eligibility on these sectors, although the latter law strongly encourages the use of renewable energy. For example, renewable energy usage plans are included in the environmental questionnaires for both the “Commercial Fabrication Facilities” and the “Facilities for Semiconductor Materials and Manufacturing Equipment” guidance. This stands in stark contrast to the strict requirements imposed by Treasury Department guidance on electricity use by the IRA-supported green hydrogen sector. Though achievable, these requirements undoubtably impose additional direct financial and time costs on projects and industry expansion.

Given the nation’s abundant supply of affordable natural gas and deep delivery networks, a manufacturing expansion necessarily implies a significant role for natural gas–fired electricity generation. Natural gas provided 43 percent of the energy consumed in the power sector nationwide in 2023, though this figure varies significantly on a state-by-state basis. As states and utilities seek to meet rapidly expanding electricity demand, transmission constraints and resource availability limit wind and solar expansion, continued cost and execution risk limit options in the nuclear sector, and the ongoing need for firm dispatchable capacity often leaves system planners with few viable options beyond expansion of the gas-fired generation fleet.

Outside of the electricity sector, direct natural gas by industry, representing some 32 percent of the total U.S. gas consumption, is critical for applications such as onsite generation, process heat, and feedstock. For example, direct use of natural gas plays a large role in battery manufacturing to provide the high temperatures necessary for drying the solvent used in the electrode manufacturing process and for the drying rooms critical for cell assembly. Studies on the relationship between the boom in U.S. energy production and manufacturing suggest that the effect of reduced U.S. natural gas prices, even in comparison to pre-2022 European gas prices, has been small but positive. However, in the context of a new supportive policy environment for industry, the broad availability and affordability of natural gas is likely a key enabler and competitive edge in both the global commercial and global strategic landscape.

The strategic manufacturing sectors should prioritize collecting and sharing more energy data—qualitatively and quantitatively—as this data will be key to aiding their own success. Better understanding of energy usage estimates and their economic implications can put local governments, utilities, and energy regulators in a position to plan for meeting regulatory and infrastructure needs and mitigating potential strains on the power capacity. It would also inform future energy system investments and policymaking as the nation seeks to seize on the unprecedented investment growth and what it can mean for U.S. leadership both economically and strategically for decades to come.

Cy McGeady is a fellow with the Energy Security and Climate Change Program at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Hatley Post is an intern with the Energy Security and Climate Change Program at CSIS. Jane Nakano is a senior fellow with the Energy Security and Climate Change Program at CSIS.

Cy McGeady
Fellow, Energy Security and Climate Change Program

Hatley Post

Research Intern, Energy Security and Climate Change Program
Jane Nakano
Senior Fellow, Energy Security and Climate Change Program