Most civil contractors walk a data center site and see a big job. Wide open land, large concrete scopes, and underground utility runs. Familiar work at an unfamiliar scale.
Data center civil work is not standard civil work with a bigger budget. The infrastructure density, the phasing requirements, the coordination layers, and the schedule consequences are categorically different from a logistics facility, a retail pad, or even a large industrial site. Contractors who estimate it the same way find this out the hard way.
Daniel Harrell, Gregory’s lead estimator, has priced enough of these campuses to know exactly where that mistake starts. Here is what his team accounts for that most estimating playbooks miss.
The Underground Scope
On a typical civil job, utility conflicts are manageable. You call 811, you know where the existing lines are, and you work around them.
On a hyperscale data center campus, the underground is a deliberate and highly engineered ecosystem. Power distribution, fiber routing, conduit duct banks, storm and sanitary systems, fire suppression lines, and domestic water all compete for the same corridor space. The power and communication infrastructure alone runs at a scale most commercial estimators have never priced: larger duct banks, larger conduit runs, all of it threading through and around a massive building footprint. Storm drainage and water and sewer lines often have to sit deeper than they would on a standard site just to make room for everything else.
These campuses also tend to carry strict underground utility avoidance plans that are not typical of commercial work. It changes how the corridor gets sequenced, and it is easy to miss if you have not estimated this kind of site before.
The underground density on a mission-critical site is not always what people expect, either. It usually is not as dense as under a roadway in an old part of a city street system, but it is more than most commercial sites, and the real complexity comes from somewhere else: phasing. These campuses get built in stages, which means two lines that will eventually run side by side might get installed months or even years apart, with the first one already energized by the time the second goes in. Gregory works with the client to find ways to sequence that installation more efficiently and more safely, rather than defaulting to whatever order the original drawings happen to show.
The conflicts are not always visible on the drawings, either. Duct bank bends, elevation transitions, and future phases of infrastructure that are not yet designed often occupy space that looks open on the current plan set. Gregory’s estimators price coordination time, build in contingency for field adjustments, and flag the corridors that need resolution before a shovel goes in the ground.
Miss it in estimating, and you will find it in the field. At that point, it is not an estimating problem anymore. It is a change order, a delay, and a conversation you do not want to have with the owner.
Phasing Is Not Flexible
On most commercial or industrial projects, phasing is a scheduling preference. On a data center campus, phasing is tied to energization milestones that the owner’s business depends on.
Civil work on these sites goes in first, and everyone else builds on top of it. If the building slab is scheduled to be poured by a certain date, the civil scope has to be finished and out of the way before then. A delay on the civil side does not stay contained to the civil side. It propagates through every trade behind it, and on a project representing a multi-billion dollar investment, that delay is measured against when the facility is supposed to come online, not just against the construction schedule.
The start date might move. The end date rarely does, because the owner has a market they need to get to. Every civil scope that touches a building’s power infrastructure has to be sequenced, completed, and inspected before the electrical contractors can proceed. That includes generator pad tolerances, transformer yard grading, and switchgear pad installations, and most of these campuses do not have just one generator yard. They have several, built in for redundancy, and each one has to be estimated and sequenced with that in mind.
“The critical path for the work can look different than standard commercial projects, with the need for such large infrastructure like power and telecom,” Harrell says.
Meeting that critical path takes an active coordination effort across trades, so no crew is standing around waiting on another to finish before they can start. Civil estimators who treat these scopes as standard concrete pours miss the coordination overhead, the inspection sequencing, and the cost of working the same footprint as other trades under the same pressure.
What Happens When the Schedule Gets Compressed
It will happen. Budget for it.
Three months into a data center civil contract, a supply chain delay on a piece of electrical equipment can pull significant time out of the back half of the schedule. That compression lands on whoever is still in the ground, which is usually the civil contractor.
From an estimating standpoint, the response starts with the schedule itself. Gregory’s team studies the milestones and builds a sequence that keeps crews working steadily without tripping over each other, which often means running multiple operations in multiple areas at the same time. That takes additional crews, additional management and safety personnel, and additional equipment.
The real work is in the plan: knowing which areas get worked by which crews, with what labor and what equipment, and making sure materials are staged and ready before the compressed timeline arrives. That gets priced in up front. It does not get scrambled together after the schedule moves.
What Gregory Actually Accounts For
Civil estimating for data centers requires estimators who have been on these sites, know where the problems live, and can translate that experience into a number that reflects reality.
Gregory hears the same feedback often enough to know it is not a coincidence: a Gregory proposal tends to include scope that a competitor’s proposal did not. Gregory is relationship-driven, and leaving scope out of a proposal just to win the bid is the kind of shortcut that shows up later as a change order. Delivering the full picture up front takes planning; most estimating playbooks skip. Multiple superintendents and site personnel, along with management, are involved before the project ever breaks ground, building a schedule and a plan for what the site actually needs, instead of sending crews out to start digging and hoping it all works out.
That same instinct shows up in details that a lot of estimators overlook entirely. Hyperscale data center engineers will often call for special backfill requirements around mission-critical utilities, tied to thermal and structural calculations that have not even been finalized by the time the project goes out to bid. An estimator without data center experience treats that as a standard backfill and prices it that way. Gregory’s team knows to flag it, push for clarity on what those calculations will actually require, or price an add-on package that covers the typical scenarios, so the number still holds once the engineering catches up to the schedule.
Mission-critical clients expect a first-in-class product, and there is little room on these sites for delays or disagreements once work is underway. That standard starts before crews ever mobilize, with training and planning built around what these projects demand. A job done right can take more time up front than a rushed one, but falling short of that standard costs far more in rework and incidents than a rushed job would ever save. Gregory’s field teams feed lessons learned back into the estimating process regularly, which is a big part of why the numbers hold up once the work actually starts.
“These are things you don’t know if you don’t know,” Harrell says. The basics of the work are the same as any civil job. What changes is the level of detail in the pricing breakouts, the phasing as funding becomes available, and the budgeting discipline required while a project is still being designed. On mission-critical work, the pace is fast, and the tolerance for error on safety and quality is close to zero. Low bidding and making it up later with change orders does not work in this world.
The Difference Shows Up Early
Owners and GCs who work with Gregory on data center civil scopes tend to notice the same thing: fewer surprises. Not because the work is easier, but because the people pricing it have done it enough times to know where the surprises come from.
That is what experienced civil estimating looks like on mission-critical work. It is not a bigger version of the standard playbook. It is a different playbook entirely.
If you are planning a data center project and want to talk through the civil scope early, we would rather have that conversation before the estimate than after.