The rise of data center development has revitalized demand for many formerly slow-growing industrial sectors, including power generation, heat/ventilation/air conditioning (HVAC), and electrical equipment. New technologies like artificial intelligence (AI), immersion cooling, and changing regulatory standards have accelerated this demand.
AlphaSense’s Expert Insights boasts an extensive range of firsthand perspectives from manufacturing and energy experts covering the latest trends and topics. These insights can be found in our expert transcript library and enable leaders to identify implications for the manufacturing and energy sectors at a deeper level.
Below, we explore how the fast growth of data centers has led to higher power consumption, higher demand for HVAC, and increased usage of AI-driven equipment.
Surge in Power Consumption
Interest in nuclear energy, immersion cooling, and energy management equipment has spiked in recent months as experts see these technologies as paramount to enabling rapid data center growth.
According to an Enverus report:
“US total load is forecast to grow 42% by 2050 from today because of population growth, increased data center demand and EV adoption, offset partially by BTM residential solar and storage growth.”
– Analyst, Enverus | Report
Analysts estimate that 47 GW of additional power generation capacity will be necessary to accommodate growth for the US’s peak demand by 2030—representing $50 billion in capital investment.
In the fall of 2023, Microsoft hired a nuclear technology manager that could develop a strategy for implementing small modular reactor (SMR) technology for cloud computing and artificial intelligence.
Data center power consumption from 2022 to 2030, according to Arthur D. Little.
If Microsoft’s strategy gains traction among tech giants, it could reverse the challenged outlook for SMR producers like NuScale. Some utility companies are concerned about the arrangements between independent power producers, like Talen and Constellation (which are focused on supplying low-carbon nuclear power) to data centers.
These concerns have led American Electric Power and Exelon to protest the power purchase agreement between Talen Energy and Amazon Web Services (AWS). American Electric Power and Exelon asked the Federal Energy Regulatory Commission (FERC) to review whether Talen and AWS are violating their interconnection service agreements (ISA).
US power generation stack, according to Constellation Energy.
According to a former Vice President of Westinghouse, data centers will require a heavy base load. Ultimately, nuclear power may be the best resource to turn to:
“You’re going to have a pretty strong base load of need from data centers. Nuclear is probably one of the best from a clean perspective to be able to provide baseline or base level power to them.”
– Former Vice President, Westinghouse Electric Company | Expert Transcript
One analyst suggests that big tech companies are entering into agreements with geothermal and nuclear power projects to meet data center demand:
“Post-2030, more data center demand could sit behind the meter, tied to new technologies such as SMR nuclear and next-generation geothermal. We are starting to see big tech companies signing up new power purchase agreements with early-stage geothermal and nuclear projects.
Meanwhile, independent power producers are considering developing data centres at their existing generation facilities – off-grid, causing concern for grid operators.”
– Analyst, Wood Mackenzie | Report
Higher Demand for HVAC
Growing AI and cloud services demand is simultaneously increasing the need for power equipment cooling from legacy HVAC players like Johnson Controls, Carrier, Trane, and Daikin, as well as newer HVAC players like Vertiv, ABB, and Asetek.
Liquid cooling, also known as immersion cooling, is a relatively new technology invented in the last three decades. In this technology, servers are completely immersed in a fluid that cools more effectively than air.
New HVAC players currently have a lead in immersion cooling over legacy players. However, legacy players are bolstering their immersion cooling options by investing in new entrants. For example, Trane recently invested in Liquid Stack, while Carrier invested in Strategic Thermal Labs.
Different cooling technologies for data centers, according to Arthur D. Little.
Although one expert suggests that while immersion cooling is introducing a competitive dynamic to the market, there is no market leader in the HVAC space, currently.
“There are a number of leaders, but there’s no one that has a dominant market position. You’ve got the traditional players like JCI, Carrier, Trane, Emerson. But you’ve got the new players that specialize in edge cooling solutions having some real play like the Vertivs of the world.
It’s not shaking up the market, but it’s providing some dynamics to a market that was a relatively stable and slow-moving in terms of share migration market. It’s an opportunity out there for the specialist coolers to come in and have a bigger part of the total overall HVAC market.”
– Former Vice President, Johnson Controls | Expert Transcript
Another expert sees demand for liquid cooling outpacing air cooling, even though there is not enough liquid cooling supply to satisfy demand. Consequently, more hybrid facilities will begin to appear:
“As we’re moving into the liquid-cooled environment, those players, the Vertivs, the Schneider Electric, some of the more household names, were not developing a liquid cooling solution very rapidly until the last two years.
It depends on the environment you’re trying to create, and it depends on if you’re using water or refrigerant to do the heat exchange. That will dictate the competitive nature of what brands are competing. I think that what you’re going to see, in my opinion, and studying the market, is you’re going to have hybrid facilities that have traditional air-cooled environments in one data hall.
Maybe next door to that data hall is a liquid-cooled data hall, and then maybe next door to that, you have the immersion facility all under the same roof, but with three different types of architecture to fulfill the need of that particular customer for their compute.”
–Vice President, Service Logic | Expert Transcript
Rising Usage of AI-Driven Equipment
Rapid data center development has boosted growth for the once slow-growing management equipment sector beyond previous gross domestic product (GDP) periods. That in part to AI, which is poised to significantly increase the demand for electrical equipment across various sectors.
As AI technologies continue to be integrated into manufacturing processes, the need for sophisticated electrical components that can handle complex computations, data processing, and energy efficiency requirements will only heighten. This includes high-performance power management systems, energy storage solutions, and robust electrical infrastructure.
Electrical equipment player Schneider Electric expects AI, data centers, and sustainability to generate 7-10% annual sales growth between 2024 and 2027. This is reflected in Schneider Electric’s acquisition of AutoGrid, a software company that helps clients manage flexible capacity for energy grid consumption.
“If the demand in the grid is too high, it would basically signal the energy consumer to turn down their load and accordingly balance out their load with the grid. These are ways that Schneider Electric has built up a technology stack and the consulting and software stack and services to be able to address this market, even with their acquisition of AVEVA. AVEVA is a software company. They are primarily focused in the industrial space.
A lot of their technologies enables further optimization to processes, visualization, enabling AI and machine learning to increase production efficiency and asset effectiveness. All of these eventually lead to more sustainable operations. In the past, Schneider Electric was very hardware-focused. A lot of low voltage, a lot of medium voltage equipment a lot of some metering, some monitoring.”
– Former Managing Director, Schneider | Expert Transcript
An expert suggests that AI has different requirements for data center builds in comparison to cloud computing and presents design challenges for operators:
“From an industry standpoint, a cloud development, we would quote that at, say, 18 kw-20 kW per cabinet which is now considered medium density… That’s pretty standard across the cloud providers in terms of their densities.
Now, with AI you’re talking, I’ve seen the low end of the specs at 44 kW per cabinet, all the way up to 80 kW and 90 kW per cabinet. If you listen to some of the hype, we’re going to get 100 kW-150 kW per cabinet at some point. There’s a lot of physics challenges with that. I think that’s going to be further out. With those densities, it’s a real design challenge for the operator to accommodate.”
– Segment Director, Schneider | Expert Transcript
As businesses and governments invest in AI to enhance productivity, efficiency, and innovation, demand for reliable and scalable electrical equipment will continue to grow, driving advancements in technology and infrastructure development.
Discover key expert insights and predictions for energy and industrials in our report, Key Predictions for Energy and Industrials in 2024.
Track Key Trends in Energy and Industrial Demand with Expert Insights
Accelerated data center development has introduced a dynamic, competitive landscape for industries within the energy environment, from power consumption, to HVAC, and AI-driven electric equipment. To stay ahead, industry leaders must keep a finger on the pulse of these new developments.
AlphaSense Expert Insights can reduce the time it takes investment researchers to find critical market insights across the energy and industrial sectors and allocate more time to strategic decision-making.
Evolve your primary market research across energy and industrials when you start your free trial with AlphaSense today.