If you’re a data center operator today, you’re likely feeling a familiar, persistent headache: the desperate search for power. This is especially true with the global rush to deploy AI and high-performance computing workloads. But as you try to scale up to meet this demand, you hit a massive brick wall.
The power grid simply cannot keep up.. Just how bad is the backup? Across the US, the average wait time just to secure a grid connection has stretched to an incredible four years.
To make matters more challenging, utilities aren’t handing out power allocations on a handshake anymore. They’re raising the stakes by requiring massive upfront cash deposits and strict minimum contract payments just to stay in the engineering queue.. At the same time, finding alternative colocation space is nearly impossible.
North American vacancy has plummeted to a record low of one percent and is expected to remain near zero through 2028. The active construction pipeline of 66GW is already 95 percent pre-committed, meaning that any new space hitting the market over the next few years is already spoken for.. – Getty Images.
The easiest solution is to make the most out of capacity in existing facilities. The reality is that you’re probably sitting on a goldmine of ‘stranded’ power right inside your existing walls. By modernizing legacy enterprise facilities, many of which are ten to 20 years old, you can reclaim wasted energy, turn it into usable compute capacity and do so without waiting years for a new utility connection..
The real cost of doing nothing. It’s tempting to take a ‘wait-and-see’ approach, but the financial math of staying passive simply does not add up anymore. If you’re approaching the end of a typical five-year colocation lease and plan to look for space elsewhere, prepare for some serious sticker shock: market rents have jumped by more than 70 percent since 2020..
Data center rent growth has averaged nine percent annually since 2020 and current growth remains in-line with that trend. If you need a mid-scale deployment of one to five MW, you’re looking at average asking rents of $178/kW/month (plus electricity), which climbs to $337/kW/month for smaller workloads under 250kW..
Compounding these real estate costs is the rising price of electricity. After a decade of flat commercial power rates, electricity prices have climbed nearly thirty percent over the last five years as utilities struggle to upgrade aging grid infrastructure. In major digital hubs like Northern California and New York, rates continue to place significant pressure on operating budgets..
When you operate a legacy facility with a Power Usage Effectiveness (PUE) of 2.0 or higher, you aren’t just wasting energy, but also actively burning capital. Every single kilowatt lost to inefficient, outdated cooling is a kilowatt that could be powering a high-density server rack..
Step 1: Let the utility pay to find your blind spots. Before you can fix inefficiencies, you have to find them. But data center operations teams are often hesitant to spend tight budgets on expensive engineering audits..
This is where you can let local utility companies do the heavy lifting. Because grids are under historic strain, utilities are highly motivated to help their largest customers use less power. In fact, many will fully pay for or heavily discount professional energy audits and detailed airflow studies..
These utility-funded studies act as a free, risk-free roadmap. They bring in specialized engineers to analyze your airflow, find leaks and check your cooling systems. By using these studies, there’s a clear, data-backed picture of exactly where your power is being wasted and how much capacity you can claw back – without spending a dime of your own budget..
Step 2: Reclaiming your stranded power. Once you have the audit data, you can start making improvements. Older data centers were usually built for much lighter workloads – around three to five kW per rack.
They rely on massive cooling systems that run at full blast all the time to handle worst-case scenarios that almost never happen.. You can get this trapped power back in a few practical steps:. Airflow containment: Simple physical barriers like hot- or cold-aisle containment, blanking panels, and floor grommet seals stop hot and cold air from mixing.
This allows your cooling units to receive warmer return air, which immediately boosts their operating efficiency.. Adjusting setpoints: Many operators still needlessly chill their server rooms to 65°F (18°C). Following modern ASHRAE guidelines, we can safely raise those temperatures to 75°F (24°C) or higher, drastically reducing chiller runtime and maximizing free-cooling economizer hours..
Variable speed fan retrofits: Upgrading old, constant-speed Computer Room Air Handler (CRAH) fans to Variable Speed Drives (VFDs) or Electronically Commutated (EC) fans allows the cooling system to spin up or down to match the actual, real-time IT load.. Modern UPS systems: Swapping out older double-conversion UPS systems for modern, modular systems with high-efficiency ‘eco-modes’ can cut electrical losses by up to sixty percent..
When you lower your PUE, you are basically shifting power away from background equipment and putting it straight into your IT racks. You get more computing power without needing a single extra watt from the power company.. Step 3: Getting utilities to co-sign your CapEx checks.
Even when the business case is clear, finding the capital expenditure (CapEx) to fund these physical upgrades is always a battle. Internal projects compete constantly for limited corporate funds.. To bridge this capital gap, you must look back to the utility companies.
Once a utility-funded study has identified your energy savings, the utility will often offer incentives and rebates to help fund the actual physical upgrades.. These programs generally fall into two categories:. Prescriptive rebates: Fixed cash-back amounts for upgrading specific equipment, like receiving a set dollar amount for every constant-speed fan motor you replace with an EC fan or VFD..
Custom incentives: Custom payouts based on the total verified kilowatt-hour (kWh) savings achieved by a larger optimization project.. By combining these upfront utility incentives with the ongoing operational savings of a lower power bill, the payback period on modernization projects can drop dramatically – frequently turning a tough-to-swallow five-year payback into an incredibly attractive, easily approved two-year business case..
These strategies help bypass the grid-connection queue. Instead of waiting four years for more power or taking on the high capital costs of temporary behind-the-meter generation like the gas turbines we’re seeing deployed in Chicago or Columbus, we simply make better use of what we already have..
The greenest data center is the one already built. Modernizing a legacy enterprise data center is about keeping your business ready for the future. It turns an aging facility into an efficient, high-performing hub that can easily handle modern hybrid cloud workloads..
It also directly supports your corporate sustainability goals. When you lower your PUE, you naturally shrink your Scope 2 carbon footprint while cutting your day-to-day operating costs.. In a world where power has become the ultimate currency, you can no longer afford to waste it.
By using utility funding to audit your legacy facilities and using their cash incentives to offset your modernization costs, you can tap into massive capacity hidden right in front of you.. Ready to find your hidden capacity? Get in touch with our data center team to start reclaiming your stranded power..
More from JLL. 12 Aug 2026
