Back to top

Data centers need more than just large-scale power

Growth in the U.S. data center market is being driven by hyperscale cloud services providers, artificial intelligence (AI) workloads, and digital transformation. In 2024, the United States data center market was valued at approximately $121 billion; in 10 years, this valuation is expected to explode to over $358 billion. 

The U.S. power grid and utilities will be challenged to meet the growing power demand. Some projections indicate that data centers may account for up to 60% of total U.S. load growth between 2025 and 2030. These power-hungry developments already account for about 3% of total U.S. power consumption, with planned local concentrations that range from tens of megawatts up to multi-gigawatts.  

For perspective, power demand has remained relatively unchanged for the past decade; only the generation mix is different. It is believed that data centers will create a paradigm shift in how generation is developed, forcing hybridization of technologies to meet the rapid increase in demand. 

Managing large-scale power

The need for large-scale power is only part of the story. For utilities and data center developers, the management of this scale of energy is where some of the most complex challenges lie. 

Mitigating potential impacts to the power grid is not necessarily compatible with delivering large amounts of energy to sites that are sized in acres and containing sensitive technology requiring high quality and reliable power.

Imagine power needs that are sized on the order of a large industrial load -- think as large a standard U.S. city -- placed at a single transmission point of interconnection and contained within less than half-a-square mile. The engineering challenges are not just formidable, they are daunting. 

The design and implementation of effective electrical protection and control systems are essential when response times are truncated and physics become improbable. 

Energy resilience is essential to protect sensitive computing operations both technically and economically when power must always be available. 

Equally as important are the risks these new data center loads can pose to the security of the larger electrical grid. 

AI-related computing loads can be notoriously variable. It is possible to find load reductions of up to 90% of full demand occurring within seconds – or even milliseconds – many times every minute with no warning. Without proper planning and design, these variations can pose unacceptable risks to stability of the transmission grid and any on-site generation resources. 

Additionally, the contractual risks to data center operators can be significant when load stability violates agreed limits.

Alternative energy resources

Increasing numbers of data center developers are incorporating a complex, hybridized mix of energy resources to support their operations with power and stability enhancement. 

An example of these resources include battery energy storage, which brings into focus a variety of solution options with each carrying different costs, performances, and risk profiles. To add to the complexity, data center developers are actively deploying on-site generation scaled far beyond traditional diesel backup generation for the occasional critical building. 

One option, large-scale gas-turbine generation adds gas utility interconnections, investment analysis for technologies, various permitting requirements, and operational concerns to the challenges of incorporating these generation assets into a complex power environment. Sourcing natural gas has become another consequent issue for many data center developers, adding the capital and operating costs of supply pipelines to the investment stack. 

Interwoven into the energy solutions context are communications capacity, design, and performance specific to the implementation of the site energy management platform.

Significant investment requires informed decisions

The scale of investment involved in the development of any data center facility is substantial. Depending upon size, location, and intended use, costs can generally be estimated in the range of $8 million to $11 million per megawatt, with large AI facilities costing at or above the top of that scale. The largest facilities easily exceed $1 billion in total cost. 

All this to say, in addition to the technical complexity, comes substantial financial risk. 

With speed to market being paramount in this new data center "gold rush" era, the consequences of hurried, poorly informed decisions can have cascading impacts to construction, operations, and eventual contractual performance and returns on investment. The confluence of market urgency, technical complexity, investment scale, and risk all require solutions and expertise that test even the heartiest of investors and developers. But the rewards of a well-executed project have the potential to be unparalleled. 

Innovative solutions insights

Tackling these daunting challenges requires a mix of expertise, experience, innovative thinking, and rock-solid decision support. Protection and control engineering integrated with a responsive, effective control system, and communications infrastructure are needed to overcome the scale of power, performance requirements, and risk for fault protection. Innovative solution insights must support the deployment of immediate, uninterruptible power with battery energy storage configured to serve use cases that require guarding the grid against sudden load and frequency drops while also providing resilience to the data center when power reliability is a concern. 

Site and power system master planning, design, and construction must be coordinated to deliver cost-effective results. Operational efficiency and safety standards are critical once the data center site is operational, particularly when customer-owned high-, medium-, and low-voltage assets are activated. 

Tying it all together, the technical insights of a broad industry perspective honed with years of investment grade analysis must be relied upon to ensure smart, efficient decision-making when millions of dollars are at stake and the consequences of a misstep manifests not only in terms of dollars and cents, but in safe operation and liability mitigation. 

The right mix of bold expertise and prudent thinking are essential in a data center development team.

Leidos delivers

More so than any other company, Leidos provides the precise mix of experience and knowledge needed to support data center development projects of any scale or nature. 

Our work with electric utilities and regulatory agencies across the U.S. provides unparalleled engagement, planning, interconnect, and engineering capabilities. We are engaged in advanced site-level power systems research and demonstration projects for customers in the Federal, local, and commercial spheres. Most importantly, we are partnered to deliver results addressing the exact issues described for the largest data center projects underway today. 

Our engineers and consultants support our data center clients in solving the most impactful challenges they face. We help these clients chart the road ahead by identifying risks and advising them on decisions so they can meet the urgency of the moment and achieve technically superior results and improved returns.  

Author
Joseph Blackwell
Joseph Blackwell Management Consultant

Joseph Blackwell is a business-oriented consulting executive with over 40 years of experience in the information technology (IT) industry and strategic management consulting. He focuses on helping customers define and execute vehicle electrification plans, develop strategic technology plans, address changes brought about by emerging technology and business needs, converge informational and operational technologies, and develop their leverage of data through improved practices and analytics.

Posted

June 26, 2025

ESTIMATED READ TIME