Data Center Viability Screen
19 parcels · 2,499.3 acres · Denton County, Texas
The nearest 345 kV transmission is 8.8 miles away, which caps what the grid will give you at roughly 135 MW without new construction. But three high-pressure gas pipelines cross this property — including a 36-inch Enterprise line — which makes on-site generation the realistic path to a large load. The land is not the constraint here. The power strategy is.
What this means, in plain English
A data center needs land and it needs power. You have plenty of land — at roughly 50 acres per 100 MW, these 2,499 acres could physically host well over 4,000 MW. Land is not what limits this site.
Power is. A 138 kV line already crosses the property, and the Rector Road substation that feeds it can deliver about 135 MW today with no new construction. That is a real, immediately usable number and it is more than most rural land can claim. But the 345 kV backbone that large campuses connect to is nearly nine miles off, so scaling past a few hundred megawatts on the grid alone would mean paying for transmission that does not exist yet.
The offset is underneath the property. A 36-inch Enterprise natural gas line and two Atmos lines physically cross this land. That opens the door to generating power on site instead of waiting in the interconnection queue — which, for a large load, is often the faster and cheaper path.
Six-factor screen
Infrastructure map
Power — detail
Transmission
| Asset | Distance from boundary | Note | Source |
|---|---|---|---|
| 138 kV line | Crosses site | ~2.0 mi of line inside the parcels, Rector Road path | OSM / OpenInfraMap |
| 345 kV line | 8.8 mi | No viable near-term interconnect at this voltage | OSM / OpenInfraMap |
| Rector Road Substation (Brazos) | Feeds site | ERCOT_672 · 134.6 MW available at zero upgrades | ERCOT capacity screen |
| Aubrey Substation (Brazos) | 3.8 mi | Not present in the capacity screen | OSM |
| Sanger North (Brazos) | ~6 mi | 138 kV distribution class | OSM |
| Krugerville (Oncor) | ~5 mi | 138 kV distribution class | OSM |
Every substation within reach of this site is distribution class, not transmission class. That is the reason for the roughly 100–300 MW practical grid ceiling. A larger grid-served load would require new transmission, which is a multi-year, multi-million-dollar undertaking and is not assumed anywhere in this report.
Gas — the on-site generation case
| Pipeline | Diameter | Distance | Status |
|---|---|---|---|
| Enterprise · T-4 03883 | 36″ | Crosses site | RRC verified |
| Atmos | 24″ | Crosses site | RRC verified |
| Atmos | 12.75″ | Crosses NW parcels | RRC verified |
| Energy Transfer | 36″ | 0.44 mi | RRC verified |
| Atmos | 30″ | 0.45 mi | RRC verified |
| Explorer Pipeline | — | 2.07 mi | Refined product — not usable |
Physical proximity to a pipeline is not the same as being able to buy gas from it. Volume availability, interconnect cost, and tap feasibility all require a direct conversation with the operator. Proximity establishes that the conversation is worth having — nothing more.
Water, flood, and air
| Factor | Finding | Implication | Source |
|---|---|---|---|
| Surface water | Culp Branch crosses the site | On-site water present; volume and rights not assessed | OSM waterways |
| Flood | Predominantly Zone X, with Zone A fingers and one AE reach | Buildable envelope needs to route around the creek corridor | FEMA NFHL |
| Air quality | Inside the DFW ten-county ozone non-attainment area | Major-source New Source Review risk for on-site generation | EPA designations |
Indicative value framing
Land that can host power trades on a different basis than farmland. Data center capacity has recently traded in the range of $650,000 to $1,000,000 per megawatt. Applying that only to the 134.6 MW this site can serve today, with no new construction and no on-site generation, implies a capacity value on the order of $87M to $135M — against a current appraised value of $24.0M.
If the on-site generation path clears permitting, the ceiling is meaningfully higher, because the constraint stops being the grid and starts being the gas volume.
What we could not determine — and what to do next
A desktop screen using public data answers roughly eighty percent of the viability question. Here is the twenty percent it cannot, in the order the money should be spent.
- Air permitting path for on-site generation. Non-attainment status is the single largest threat to the thesis for this parcel. An air permitting consultant can scope this in weeks. Do this before anything else.
- Gas volume availability. Contact Enterprise regarding T-4 03883 and Atmos regarding the 24-inch. Proximity is confirmed; deliverable volume and tap cost are not.
- Confirm the capacity number. The 134.6 MW figure comes from a published ERCOT screen. Verify current standing directly with Brazos Electric, since queue position changes.
- Zoning and jurisdiction. Determine parcel by parcel whether authority sits with the city, the ETJ, or the county. Zoning, not power, is usually the true critical path on schedule.
- Title, easements, and setbacks. Three pipelines crossing the property means three easements constraining the buildable envelope. A survey will show how much of the 2,499 acres is genuinely usable.
Method and sources
Parcel geometry, ownership, acreage, and appraised value are pulled from the county appraisal district GIS. Transmission lines and substations come from OpenStreetMap and OpenInfraMap. Gas pipelines are cross-verified against Texas Railroad Commission T-4 filings rather than relying on crowd-sourced tags. Flood zones come from the FEMA National Flood Hazard Layer. Air quality designations come from EPA. Capacity figures come from published ERCOT screens.
All distances are measured from the parcel boundary, not the centroid, so that figures remain comparable across sites. Every finding is labelled by confidence: verified against a primary source, drawn from public data, or requiring field verification.
This report contains no proprietary or paywalled data. Everything here is public record. What it contains is the assembly and the judgment — six databases that do not talk to each other, read by someone who knows which numbers matter.