Traffic Impact Studies for Data Centers: Trip Generation, Construction Traffic and Road Improvements

A traffic impact study (also called a traffic impact analysis, TIA, or transportation impact study) is an engineering report that estimates how many vehicle trips a project will generate, assigns those trips to the road network, and compares future traffic conditions with and without the project. It ends with a list of improvements, if any, needed to keep intersections operating at the level the reviewing agency requires. It is usually filed with a rezoning, special use permit or site plan, and it is one of the first technical studies neighbors and staff read.

Triggers vary by jurisdiction. Many local codes require a full study when a project will generate about 100 peak-hour trips or 1,000 daily trips. North Richland Hills, Texas, uses 1,000 vehicles a day or 100 trips in the peak direction, and St. James, North Carolina, uses 100 peak-hour or 1,000 daily trips based on the latest ITE Trip Generation Manual.⁷⁸ States can add their own layer. Virginia’s rule treats a rezoning as substantially affecting state highways when it produces more than 5,000 daily trips at its connection to a state-controlled road (in localities where VDOT maintains the secondary system), and the locality must then send the package to VDOT for review.⁹

Because operational data center traffic is light, some campuses fall below local triggers on paper. Staff often ask for a study anyway, especially on large rezonings, because the real question is construction access. The study is one piece of the broader data center permitting process and usually feeds conditions at site plan review.

Traffic engineers estimate trips with the Institute of Transportation Engineers (ITE) Trip Generation Manual. The current 12th edition added data from more than 550 new sites, introduced nine new land use classifications, and removed data collected before 1990.¹¹ Data center is land use code 160. Palm Beach County’s summary of 12th-edition rates lists it per 1,000 square feet of floor area, with daily trips estimated by a logarithmic regression equation rather than a flat rate.⁴

The data are thin. A traffic memorandum filed in Stafford County, Virginia, notes that the ITE data center rates rest on two sites for daily trips, six for the morning peak and five for the evening peak, and that the manual itself urges caution because few sites report tenant and employee data.¹ The same memo observes that data centers have very few visitors, so most trips are employees and deliveries.¹

Large buildings still produce real numbers. A June 2026 traffic report for a proposed Florida data center applied the 12th edition and land use 160 to roughly 3.8 million square feet and estimated 2,793 daily trips.⁵ Engineers and reviewers frequently cross-check ITE output against a first-principles estimate built from staffing, shift times and deliveries. Where shift changes on a 24/7 schedule fall outside the morning and evening commuter peaks, the peak-hour impact can be small even for a large building, which is why reviewers ask for the shift schedule as well as the floor area.

A data center campus has two very different traffic profiles. Once operating, it looks like a lightly staffed industrial use. During construction, it can look like a small city’s worth of commuters plus a steady stream of heavy trucks. In Wood County, Ohio, a $750 million data center was expected to bring more than 1,000 vehicles a day at the peak of construction, and the county built a roundabout and turn lanes to serve the site.² In Caddo Parish, Louisiana, officials said heavy truck traffic for the Stateline Data Center could last five to seven years as buildings are built in phases.⁶

Reviewers have started asking for construction traffic to be analyzed explicitly. In Morrow County, Oregon, the county’s public works staff asked for a separate assessment of construction volumes on Tower Road after reviewing a study that focused on post-construction traffic; the consultant used a week of tube counts to set the baseline.¹² In Hanover County, Virginia, the 2025 study for a proposed 900 MW, 434-acre data center rezoning analyzed existing, 2034 no-build and 2034 build conditions and designated a US-33 emergency and utility access as a temporary construction driveway.¹³

Where construction traffic was not managed well, it became the public story. In Port Washington, Wisconsin, damage to local roads from construction of a large campus became a county issue and ended in a repair agreement with the developer, described below.³ Oversize deliveries for transformers and switchgear add a separate permit track, covered in heavy-haul routes and construction access.

Most traffic impact studies follow the same structure, though each agency publishes its own scoping rules. The scope (study area, intersections, analysis years and methods) is normally agreed with the reviewing engineer before counts are taken.

Skeptical reviewers look for a few things: whether ITE rates were applied to the right floor area, whether shift changes overlap with school or commuter peaks, whether the study counted construction at all, and whether the access plan works for fire apparatus and oversize vehicles. Site access and frontage constraints that shape the driveway layout are covered in site shape, access and frontage.

A study’s recommendations become conditions of approval. For an operating data center, the result is often limited. A 2026 turn lane memo for a 750,000-square-foot campus in Spotsylvania County, Virginia, found no turn lanes warranted at the entrance, on a road carrying about 40 vehicles a day.¹⁴ Larger campuses on busier roads can still face turn lanes, signals or a roundabout, as in Wood County, Ohio.²

The bigger negotiation is usually about pavement. Rural county roads are rarely built for months of loaded trucks. Counties have answered with road use agreements: in Ozaukee County, Wisconsin, the arrangement makes the developer financially responsible, with the county highway department doing repairs and billing monthly for time and materials, and either side able to end it on 60 days’ notice.³ Pueblo County, Colorado’s 2026 road use agreement template requires a pre-construction survey of each road on the haul route, paid for by the company, and a repair plan where the survey calls for one.¹⁰ In Caddo Parish, officials have said the developer will be held responsible for significant road damage.⁶

Road commitments often end up in a proffer, condition set or development agreement, sometimes alongside formula-based transportation fees, which are explained in impact fees and connection fees.

Traffic rarely kills a data center site on its own, but it can add cost, delay and opposition. A short desk screen before an option is signed can surface most of the issues.

• Find the local TIA trigger and whether a state DOT review applies, such as Virginia’s 5,000-trip rezoning criterion.⁹
• Identify the road class and owner for every frontage and likely haul route: state, county, township or private.
• Check posted weight limits, bridges and seasonal restrictions on routes to the nearest interstate or rail siding.
• Estimate construction workforce and truck volumes by phase, not only operating trips.¹²
• Ask the county engineer whether a road use agreement, pre-construction survey or bond is expected.¹⁰
• Look for schools, crossroads communities and commuter peaks that construction shifts would overlap.

Traffic also feeds directly into community engagement: truck traffic and road damage are common complaints once construction starts, as the Port Washington and Caddo Parish cases show.³⁶ If you want a second look at access, haul routes and entitlement risk on a parcel, you can get a site reviewed.