Key takeaways
- Power timing kills more sites than power distance.
- A large tract can still fail once wetlands, flood zones and easements are subtracted.
- Zoning that allows data centers on paper does not remove community risk.
- Title surprises, such as pipelines or severed mineral rights, often appear late.
- Some red flags can be cured; the question is cost, time and certainty.
01The pattern behind failed sites
Failed sites usually share a pattern. Someone saw a large tract near a transmission line and assumed the hard part was done. Then the details arrived one at a time: the substation was full, a creek ran through the middle, the zoning required a public hearing, a pipeline easement split the tract in two.
None of these issues is unusual on its own. What sinks a site is a combination of them, or one that cannot be fixed on the client’s timeline. The red flags below are grouped by how often they decide the outcome.
The useful question is not whether a site has problems. Every site does. It is whether those problems are known, priced and survivable on the schedule the project needs. A site with three known, curable issues is often a better bet than a site that looks clean only because nobody has checked.
02Power: capacity, upgrades and timing
Power is the most common reason sites fail. Being near a high-voltage line is not the same as having capacity. The line may be heavily loaded, the nearest substation may have no room for new transformers, or serving the load may require network upgrades that take years to permit and build.
- The utility indicates service only after major transmission upgrades with no firm date.
- The nearest substation is at or near capacity and has no room to expand.
- The site is near a line at a voltage too low for the target load, such as 69 kV for a large campus.
- Other large loads are already ahead in line for the same capacity.
- The ramp the utility can offer does not match the project’s phasing.
Timing is often the real problem. Our guide to power timelines and interconnection queues explains why a site with less power sooner can beat a site with more power later.
03Land that shrinks once constraints are mapped
Gross acreage is the number on the listing. Buildable acreage is what remains after floodplain, wetlands, streams and buffers, steep slopes, easements and setbacks are removed. The gap can be large, and the remaining land may come in pieces too small or oddly shaped for data halls and a substation. The guide to gross vs. buildable acreage walks through the subtraction.
| Red flag | Why it matters | Often curable? |
|---|---|---|
| Floodplain in the building area | Raises fill cost, complicates permits and affects insurance | Sometimes, at cost and with permits |
| Jurisdictional wetlands or streams | Section 404 permits add time; avoidance shrinks footprint | Sometimes, by redesign or mitigation |
| Pipeline or transmission easements across the middle | Splits the site and limits building and crossings | Rarely; relocation is slow and costly |
| Steep slopes or shallow rock | Large grading and blasting costs | Yes, at cost |
| Karst or poor soils | Foundation risk and added geotechnical work | Sometimes, with engineering |
| Narrow or irregular shape | Cannot fit large rectangular buildings and yards | Rarely without more land |
04Zoning and community opposition
A site needing rezoning or a special use permit carries approval risk and schedule risk. Even where zoning allows data centers, local opposition can slow permits, add conditions or push officials to change the rules. Concerns tend to focus on noise, water use, visual impact, traffic during construction, transmission lines and the scale of the buildings.
Warning signs include homes close to the likely building area, a recent contested data center hearing nearby, a moratorium under discussion, or a comprehensive plan that designates the area for something else. Our guide to community engagement for data centers covers how these concerns play out.
05Title, ownership and assembly problems
Title problems tend to surface late because title work is ordered after a deal is signed. Common issues include severed mineral rights that allow surface use, restrictive covenants that limit industrial use, unclear ownership among heirs, and access that depends on a private easement.
Multi-owner sites add assembly risk. One holdout parcel in the middle of a campus can stop the project or force a redesign. The longer an assembly takes, the more likely word spreads and prices move. See parcel assembly for data centers for how teams manage this.
06Fiber, water and airspace
- Fiber: no long-haul route nearby, a single carrier, or no way to bring in two diverse paths without long new builds.
- Water: limited municipal capacity, constrained groundwater, or no path to discharge cooling water, where the cooling design needs it.
- Airspace: a nearby airport whose surfaces under 14 CFR Part 77 limit building or crane height.
- Gas: no pipeline nearby when the plan relies on on-site generation.
These issues rarely kill a site alone, but they add cost and time, and they compound with power and land problems.
07How to catch red flags early
Most red flags can be spotted before a site goes under contract. A structured screen of mapped constraints, followed by early conversations with the utility, the local planning office and fiber carriers, removes many weak sites in weeks rather than months.
- 01Define the target load, acreage and schedule before looking at sites.
- 02Map constraints and calculate buildable acreage on every candidate.
- 03Ask the utility about capacity and timing before signing anything.
- 04Check zoning and recent local history with similar projects.
- 05Pull a preliminary title report early on sites that look strong.
When we screen sites, the goal is to find the reason a site might fail as early and cheaply as possible. A site that survives that process is worth the cost of full diligence.
Common questions
What is the most common reason a data center site fails?
Power is the most common reason. Specifically, the utility cannot deliver the needed capacity on the project’s timeline, either because the nearest substation is full, the transmission network needs major upgrades, or other large loads are ahead in line. Being close to a high-voltage line does not by itself mean capacity is available, which is why early utility conversations matter so much.
Can a site in a floodplain still be used for a data center?
Sometimes, depending on how much of the site is affected and where. Many developers avoid placing buildings in FEMA Zone A or AE areas and keep critical equipment well above base flood elevation. If only the edges of a large tract are in the floodplain, those areas can often serve as buffers or open space. A site mostly in the floodway is usually a poor candidate.
Can community opposition stop a data center project?
It can delay or stop projects that need discretionary approvals such as rezoning or a special use permit, and it can add conditions that change cost or design. Even by-right projects can face pressure through permit appeals or changes to local rules. Sites with few nearby homes, industrial land use designations and supportive local plans carry less of this risk.
Are red flags always deal-breakers?
No. Many red flags are curable at some cost: grading steep land, mitigating wetland impacts, building a utility extension or obtaining a variance. The questions are how much the cure costs, how long it takes and how certain the outcome is. A red flag becomes a deal-breaker when it pushes the schedule past the client’s needs or makes the site uncompetitive with alternatives.
How can a landowner check for red flags before marketing land?
Start with free mapped sources: FEMA flood maps, the National Wetlands Inventory, county zoning maps and parcel records. Note where transmission lines and substations sit relative to the property and list every recorded easement on the title. Then look at nearby homes and any local debate about data centers. A professional screen adds utility context and buildable acreage, which mapped data alone cannot provide.
Have a site in mind?
Get a straight answer on your land.
Send a parcel number, an address, a map pin or a target load. We’ll tell you what it can support and what it would take.
This guide is general information about data center site selection. It is not engineering, legal, tax or investment advice. Requirements vary by state, utility and county, so confirm the specifics for any site with the relevant authorities and advisors.