The Demand Supercycle: Data Centers, the Grid, and Years of Hydrovac Work

Picture a crew staging two hydrovac trucks on the shoulder of a county road in Loudoun County, Virginia. The work order is unglamorous: daylight several dozen potholes along a new feeder route so a transmission contractor can set foundations for a substation. Nobody on that crew will ever badge into the data center the substation is being built to feed. They are simply the first trade in, exposing live gas, fiber, and water that no one will let a backhoe near.

That scene is repeating across the continent, and it is the visible edge of a structural shift most operators feel before they can name it. After roughly two decades in which US electricity demand barely moved, the load curve has bent upward again. The U.S. Energy Information Administration now forecasts the strongest four-year growth in US electricity demand since 2000, the first stretch since 2007 in which power demand has risen four years running, driven in large part by large computing facilities including data centers.

The thesis of this piece is simple to state and hard to overstate. The same forces lifting that load curve, AI data centers, grid and transmission expansion, electrification, and reshoring, also push enormous quantities of new infrastructure into the ground. Because so much of that build happens beside live utilities, it is daylighting and potholing work, not only open-trench digging. For hydrovac, this is the demand backdrop the industry has waited a long time for. This article is about why that tailwind is durable, where the work concentrates, and what it does to operator economics.

After twenty years of flat load, the curve bent

The starting point is a two-decade plateau. From the mid-2000s into the early 2020s, US electricity consumption was essentially flat, as efficiency gains and the shift from manufacturing toward services offset most growth. Utilities planned around a stable load. That assumption is now obsolete.

The reversal is being led by computing. The International Energy Agency estimates global data-center electricity use at about 415 terawatt-hours in 2024, roughly 1.5 percent of world consumption, and projects it to more than double to about 945 terawatt-hours by 2030, slightly more than Japan's total electricity use today. In the United States specifically, the IEA finds data centers account for close to half of the growth in electricity demand through 2030.

The domestic numbers are just as stark. A 2024 report prepared for the U.S. Department of Energy by Lawrence Berkeley National Laboratory put US data-center consumption at about 4.4 percent of national electricity in 2023, roughly 176 terawatt-hours, with load tripling over the prior decade. The same report projects data centers will reach somewhere between 6.7 percent and 12 percent of US electricity by 2028, a range of about 325 to 580 terawatt-hours.

Those figures stay abstract until you translate them into steel and trench. A percentage point of national load is generation that has to be built, high-voltage lines that have to be strung, substations that have to be sited, and distribution and service that have to reach the building. Almost none of that arrives without excavation.

The mechanism: a gigawatt of load is miles of buried work

Load does not stay in the abstract. Each new block of demand sets off a predictable chain of construction: new generation, then transmission to move it, then substations to step it down, then distribution to spread it, then the fiber that controls and connects it, and, for water-cooled facilities, the water and sewer to serve them. Most of those links have an underground component.

The decisive detail for hydrovac is location. Greenfield trenching in an empty field is backhoe work. But most of this build happens in and around developed corridors, beside energized cable, pressurized gas, pressurized water, and existing fiber. That is precisely the condition that mandates vacuum excavation. You expose the unknown by hand or by vacuum, not by blade.

Damage data underlines the point. The Common Ground Alliance's 2024 DIRT Report analyzed 196,977 unique reports of damage to buried utilities, and its CGA Index, a year-over-year measure of damage trends, rose from 94.0 in 2023 to 96.7 in 2024, meaning the industry moved in the wrong direction. Water-and-sewer and telecom excavation dominated nine of the top ten root causes of those damages. As more crews work near more buried assets, the case for daylighting before disturbing strengthens, and so does the regulatory and insurance pressure behind it.

Four segments feeding the hydrovac queue

The demand thesis is easier to underwrite when you break it into the four streams of underground work it generates.

Power: transmission, distribution, and substations

This is the largest stream. The DOE's 2023 National Transmission Needs Study concluded that, in its scenario of high electricity load and high clean-energy growth, the country would need to expand interregional transmission more than fivefold by 2035, with within-region transmission more than doubling nationally. That is a generational rebuild of the wires, and wires need foundations, duct banks, and crossings, much of which involves potholing existing utilities.

The money is already committed. The Edison Electric Institute reports that US investor-owned electric utilities are on pace to spend a record roughly 208 billion dollars on the grid in 2025, and more than 1.1 trillion dollars from 2025 through 2029. Distribution, at about 66.5 billion dollars, is the single largest category. Distribution work happens in streets and easements crowded with other utilities, which is daylighting country.

Fiber and telecom

These facilities are fiber-hungry. Data centers, substations, and reshored plants typically rely on dense, redundant fiber, and that fiber is increasingly run in built-up corridors rather than open right-of-way. The DIRT data showing telecom excavation among the top root causes of utility damage is a direct signal of how much of this runs alongside other live lines. Where the lines are congested, the prudent method is vacuum potholing.

Water and sewer for water-hungry facilities

Cooling is the quiet driver. The Lawrence Berkeley report estimates US data centers consumed about 17 billion gallons of water directly for cooling in 2023, with annual direct consumption projected to rise to between 38 and 73 billion gallons by 2028. Serving that demand means new water mains, reclaimed-water lines, and sewer capacity, often retrofitted into existing systems. Tie-ins to live water and sewer are classic hydrovac jobs.

The interconnection backlog

Demand becomes a queue before it becomes work. At the end of 2024, roughly 2,300 gigawatts of generation and storage capacity was sitting in US interconnection queues, and the median time from interconnection request to commercial operation has roughly doubled to more than four years for projects built between 2018 and 2024. For an operator, a long queue is not bad news. It is a multi-year, visible pipeline of build-out that has to clear, and most of it terminates in physical infrastructure that someone has to put in the ground.

The regional crunch: where capacity gets tight first

The load is not spread evenly, and neither is the work. A handful of regions are absorbing a disproportionate share, which turns a national thesis into local capacity crunches for trucks, crews, and disposal.

• Northern Virginia remains the center of gravity. Dominion Energy connected 15 new data centers adding nearly 1 gigawatt in a single year, expects data-center peak demand to reach about 13.3 gigawatts by 2038 against 2.8 gigawatts in 2022, and raised its 2025 to 2029 capital plan to 50.1 billion dollars. A single planned transmission line is slated to feed about 40 new substations. Building out those substations means daylighting campaigns up and down the route.
• Texas and ERCOT are scaling fast. Large-load interconnection requests in ERCOT surged to roughly 226 gigawatts by December 2025, up from about 63 gigawatts at the end of 2024, with roughly three-quarters of the requests coming from data centers.
• Ohio shows how policy gates the work. AEP faced about 30 gigawatts of data-center interconnection requests against a roughly 9.4 gigawatt AEP Ohio peak load, and state regulators approved AEP Ohio's data-center tariff on July 9, 2025, ending a 28-month interconnection moratorium in central Ohio. When a moratorium lifts, backlogged work releases at once.
• Georgia illustrates Southeast speed. The Georgia PSC approved Georgia Power's 2025 Integrated Resource Plan in July 2025, projecting about 8,500 megawatts of load growth over six years; in December 2025 the commission cleared roughly 10 gigawatts, about 9,885 megawatts, of new generation, most of it to serve data centers.

The pattern matters more than any single market. When this much load concentrates in a metro, local hydrovac capacity tightens before national averages move. That is where pricing power shows up first.

What it means for operator economics

The thesis is already converting into results. Badger Infrastructure Solutions, the publicly traded North American hydrovac specialist, reported record third-quarter 2025 revenue of 237.3 million dollars, up 13 percent year over year, with monthly revenue per truck up 8 percent to 47,921 dollars. The company ran a fleet of about 1,703 hydrovac units and raised its 2025 capital spending guidance to between 115 and 130 million dollars.

Pricing power and a record build are the tells. Revenue per truck rising at the same time the fleet is expanding is the signature of demand outrunning supply. Badger's Red Deer plant built 57 hydrovacs in the quarter against 48 a year earlier, evidence the company is funding a record organic build rather than waiting for the market to soften. For the capital-allocation logic behind that choice, see our analysis of Badger's build-versus-buy decision.

The constraint has moved. For most of the past cycle, the question was whether there was enough work. The thesis here flips that. With utility capex at a record and a multi-year queue behind it, the binding constraint is no longer demand. It is supply: skilled operators, available trucks, and somewhere to take the spoils.

• Labor is the first ceiling. A hydrovac is only as productive as the operator running it, and trained operators are scarce. Crews built around daylighting near live utilities cannot be hired overnight.
• Fleet is the second. Lead times on new units and the capital to buy them gate how fast any operator can chase the work, which is why a record organic build at the largest player is significant.
• Disposal is the third and most underrated. More daylighting means more slurry, and slurry has to go somewhere that will accept it at a workable cost. Tightening disposal markets can quietly cap utilization no matter how full the order book is, a risk underscored by recent consolidation in the disposal sector.

The honest counterweight

A real thesis names its soft spots. Two deserve attention so this reads as analysis rather than a pitch.

Reshoring is the wobbliest leg. US manufacturing construction spending hit a record annual average of about 235.6 billion dollars in 2024, peaking near 240.1 billion dollars in August 2024, before cooling through 2025 and into 2026 as semiconductor-fab spending fell back from its mid-2024 high. The factory-building wave is real, but it is no longer accelerating the way it was, and operators who underwrote on fab work alone should temper expectations.

Timelines are long, and that cuts both ways. The same interconnection backlog that guarantees years of work also means the work releases slowly and unevenly. A four-year median queue is a pipeline, not a faucet. Permitting, supply chains, and local opposition can push individual projects right. The durable read is multi-year and structural, not a quarter-by-quarter boom.

What operators should do now

1. Position crews where the load is concentrating. The regional data is a map. Northern Virginia, ERCOT Texas, central Ohio, and Georgia are absorbing outsized load, and local capacity tightens there first. If you can stage trucks and operators near those corridors, do it before the crunch peaks.
2. Lock disposal before you scale. Disposal is the quiet ceiling on utilization. Secure spoils outlets and pricing now, and watch the regulatory direction on slurry disposal, because the segment growing fastest also generates the most material to move.
3. Invest in operators, not just iron. A new truck without a trained operator is parked capital. Build the bench through structured training and certification so added units actually turn revenue.
4. Anchor to utility and telecom capital plans, not headlines. Utility capex of more than 1.1 trillion dollars through 2029 and the transmission rebuild are the durable funding behind the work. Track the resource plans and capital budgets in your region rather than chasing single project announcements.
5. Price for scarcity, but underwrite for the long cycle. Badger's rising revenue per truck shows pricing power is real. Capture it, but plan around a multi-year structural build rather than a short spike, because the queue clears over years.

The macro read

This is the demand backdrop hydrovac waited two decades to see. For twenty years the industry rode infrastructure replacement and a slow drumbeat of locating work. The reversal in electricity demand changes the baseline. When the U.S. EIA describes the fastest four-year load growth since 2000 and the IEA attributes close to half of US demand growth to data centers, they are describing, in energy terms, a continent-scale excavation program that has barely begun.

The work is durable because the spending is committed and the queue is long. A record 208 billion dollars of grid investment in a single year, more than 1.1 trillion dollars through 2029, a transmission grid the DOE says may need to expand more than fivefold between regions in its high-growth scenario by 2035, and roughly 2,300 gigawatts waiting in interconnection queues are not a one-year surge. They are a multi-year obligation, and the underground share of it lands squarely in the hydrovac lane.

The winners will be defined by supply, not demand. Demand is, for once, the easy part. The operators who come out ahead will be the ones who solved labor, fleet, and disposal early, and who priced for scarcity without betting the business on any single project clearing on schedule. The supercycle is real. Capacity, not opportunity, is the thing to manage.

Hydrovac News covers the regulatory, disposal, and operational developments shaping the hydro-excavation industry across North America. For ongoing coverage, subscribe to our weekly newsletter.