As the use of real-time data analytics and high-speed storage evolves, so too do data centers. With this evolution, utility grid-interactive microgrids are emerging as a strategic solution for resilient and flexible power.
To explore the critical factors shaping microgrid design for data centers, Jake Grove, sales engineer on PowerSecure’s data center team, brings his hands-on expertise in designing and deploying resilient energy solutions, while Todd Jackson, vice president of strategic development and leader of the company’s commercial strategy, shares his insight to identify key markets and growth opportunities.
The power industry has evolved in recent years to meet rising demand, driven in part by the growing sophistication of AI, with utilities increasingly relying on real-time data analytics and high-speed storage.
A key trend is the shift from solely centralized power plants toward distributed energy resources, allowing utilities to move generation assets to where they are most beneficial for customers. Jackson emphasizes that this evolution is not just about adding capacity, but about redesigning the grid itself:
“They’re moving to more of a distributed infrastructure for power generation and storage, so that it’s more localized to where their customer base can benefit the most.”
This distributed approach to power helps utilities relieve capacity and infrastructure pressures within their service territories. At the same time, by decentralizing energy, utilities can better manage peak demand, reduce the risk of bottlenecks, and meet their increasingly strict uptime and reliability requirements.
To support the transition, many utilities offer incentives to their commercial and industrial customers, encouraging them to install backup power solutions at their facilities. These customer-owned assets can also be integrated with the wider grid, allowing utilities to draw on this capacity during peak grid loads. As Jackson notes:
“Continuing to build out that distributed power is the way they’re addressing this central challenge, so they can access the distributed energy their customers are putting in place.”
"Demand from the data center industry is just surging, where we’re seeing interconnection queues for some utilities now exceeding dozens of gigawatts" - Jake Grove, PowerSecure
It’s an exciting time to be an engineer in a fast-moving power industry. With no single solution that can solve all the distribution challenges facing data centers, this makes designing bespoke systems for customers both a rewarding and complex task.
“You see utilities challenged with acute capacity constraints all across the US, with interconnection timelines stretching from three to over 10 years, which is unprecedented. At the same time, demand from the data center industry is surging, and we’re seeing interconnection queues for some utilities now exceeding dozens of gigawatts,” says Grove.
Grove points to grid-interactive microgrids as one promising answer, offering utilities and their customers a more flexible and resilient way to manage power. For example, multiple utilities have adopted a ‘bring your own device’ model, giving higher interconnection priority to customers deploying microgrid solutions that can serve as capacity or flexible demand resources.
Once a permanent utility connection is available, customers can earn financial incentives by reducing their load or running alongside the grid during peak demand.
Still, the approach has its challenges. With the industry changing so rapidly, standard deployments and standby-only systems are no longer sufficient. New interconnection requirements must be addressed, along with the need to optimize an increasingly diverse mix of resources. Grove says:
“It’s not just traditional, standard backup generation anymore. You see everything from PPAs with renewable integration, to grid-interactive microgrids using cleaner diesel technology, and natural gas in the mix – you name it.”
Bringing these varied sources together creates a unique energy landscape that hasn’t existed before in this space. The key now lies in balancing and optimizing them to best serve end users’ needs.
With so much growth in the industry, secondary and emerging markets are beginning to take shape, often in locations that were historically overlooked or considered undesirable. This shift reflects a new rush for powered land. At the same time, development in larger cities and Tier I markets continues, creating momentum across multiple fronts.
Across the US, Jackson has observed pockets of various markets expanding their energy sources, driven less by the need to meet everyday, steady-state demand, and more to handle the peak capacity challenges utilities and grid operators face.
The rise of AI-driven data centers is a prime example, as they create short-term spikes in power consumption due to their intense compute requirements.
“You’ll see utilities moving toward their carbon reduction goals by retiring coal-fired fleets, and at the same time aligning with new large-load customers like data centers,” says Jackson.
“For instance, tied to recent US government policy changes, one of our customers prioritized capacity for a semiconductor chip manufacturer reshoring its facility. So there’s no single answer for where growth is happening – it’s really concentrated in pockets where capacity constraints and new demand overlap.”
"You’ll see utilities moving toward their carbon reduction goals by retiring coal-fired fleets, and at the same time aligning with new large-load customers like data centers" - Todd Jackson, PowerSecure
As data centers face growing demand for high-performance computing, developers are rethinking traditional expansion methods, and turning to vertical builds to maximize the limited land in high-demand markets.
“Atlanta, Dallas, Los Angeles, Northern Virginia – the capital of infrastructure – any of those major hubs with skyrocketing price tags for powered land are now seeing developers shift from building outward to building upward,” explains Grove.
As operators push to scale vertically, several changes are becoming immediate necessities. Liquid-cooled or even direct-to-chip cooling, once considered a nice-to-have, is now required from day one. At the same time, vertical growth adds more equipment, switchboards, and UPS systems, all of which add weight and complexity. Grove notes:
“People are looking at downsizing that equipment, or moving to smaller rack-mounted UPS units, to save space, reduce costs, and better manage weight dispersion as they scale vertically.”
He adds that operators are also exploring battery energy storage and grid-interactive microgrids to handle the sharp, unpredictable load swings created by AI. These solutions not only help manage spiky demand but also allow facilities to isolate specific floors or clusters, since power needs are not evenly distributed across vertical builds.
"Major hubs with skyrocketing price tags for powered land are now seeing developers shift from building outward to building upward" - Jake Grove, PowerSecure
While certain US incentives for renewables were scaled back with legislation signed in July 2025, those for energy storage remain in place. As Jackson explains, this is shaping the types of requests PowerSecure receives from customers, which now largely align with what energy assets can be incorporated into a microgrid using available funding to offset costs.
“We’re beginning to see more discussion around hybrid microgrids – systems that integrate multiple energy assets so that they can dynamically balance their power consumption against the grid,” says Jackson.
In practice, this often means customers rely on energy storage assets for a few hours during periods of high grid pricing, or on reciprocating engines for longer durations, with utilities in some cases paying customers to generate electricity for their sustained needs.
“The load profile of a facility is starting to dictate more of the microgrid configuration than ever before,” says Jackson.
"The load profile of a facility is starting to dictate more of the microgrid configuration than ever before"
- Todd Jackson, PowerSecure
With dramatic changes in data center design, operation, and financing, determining the key factors that shape microgrid configuration and performance can be challenging. The design principles that go into an involved system of distributed power sources require balancing multiple variables while keeping one constant in mind: the industry standard of five-nines of availability.
According to PowerSecure, effectively designing a microgrid means understanding the customer or end-user’s desired level of resiliency and redundancy, their sustainability or ESG goals, reliance on the grid, the stage of their shift toward self-generation, the load profile, and the specific use of the data center – whether traditional compute or machine learning workloads. Grove explains:
“There are a lot of variables that have to be considered, but regardless of the number of inputs – navigating all those changes and different sliders – the bottom line is that the output always has to be five nines of availability. That's what you're solving for.”
Above all, customers are seeking resiliency – how quickly they can recover from a power outage caused by a grid fault. As much as technology advances, it cannot fully predict the impacts of extreme weather or unpredictable human intervention, making a one-size-fits-all solution impractical. Jackson explains:
“More storms, more human intervention with the grid – people shooting transformers, car accidents – these are all things utilities can't foresee or predict. More companies are realizing they need resilient solutions for business continuity.”
Mathematically, the average US customer experiences about two hours of non-major-event power outages per year, which encourages many to invest in energy storage for backup and resiliency. By installing battery or flywheel storage, facilities can endure most outages while also providing fast-acting voltage and frequency support to the grid.
In this way, energy storage transforms into a grid-interactive microgrid, capable of both absorbing and exporting power to help utilities maintain reliability. However, extreme events can last far longer.
“I experienced it with Hurricane Helene here in Greenville – we were without power for 11 days. These facilities need long-term backup power, which is where reciprocating engines come into play,” Jackson says.
Ultimately, microgrid design is guided by the customer’s priorities, and most prioritize resiliency. Systems are configured to handle both short- and long-term outages while maximizing the use of assets, often leveraging incentives or partnerships to make the microgrid financially viable. Grove explains:
“That’s why we’re very technology-agnostic. We start by understanding the customer – their long-term goals, what they want the facility to look like, and the ‘rules of the road’ from a utility and regulatory standpoint. Then we begin with a blank sheet of paper and optimize the resource mix to meet their needs: resiliency, reliability, sustainability, future-proofing, and their microgrid.
“If they want to bring in different energy sources later, we factor that in from the start. The goal is to create an optimized solution that meets their requirements while integrating as seamlessly as possible with existing deployments and models.”
"The goal is to create an optimized solution that meets their requirements while integrating as seamlessly as possible with existing deployments and models" - Jake Grove, PowerSecure
The evolution of data analytics and storage is driving higher demands on data centers, and utilities are responding with grid-interactive microgrids as a flexible, resilient solution. As Jackson summarizes:
“Utilities want the distributed energy they can access to enhance grid reliability and better meet growing capacity demands. When these objectives align, this is what will drive the growth of utility-sponsored microgrids.”
For operators, success depends on microgrid design, where modular generation meets scaling needs, asset selection suits high-density loads, and diversified energy sources guard against obsolescence – delivering resilience today and flexibility for tomorrow.
This article was originally published by datacenterdynamics.com.