Why Charlotte is a market at all
Charlotte is a banking town, and its colocation market is shaped by that more than by any land or power advantage. Bank of America and Truist are headquartered here, Wells Fargo runs one of its largest operations outside California here, and Honeywell, Nucor, Lowe's and Duke Energy itself all keep head offices in the metro. That produces a particular kind of demand: regulated workloads, disaster recovery positions, back office production and the compliance obligations that come with financial services. It does not produce a hyperscale market, and the two should not be confused.
The result is a genuine multi-tenant layer of a modest size, sitting inside a state that has become one of the busiest places in the country for campus-scale development. Those two things are happening in the same electricity system and almost nowhere else in common. The campuses are being built for single tenants in rural counties. The colocation you can actually buy is in and around the city, in buildings that were sized for enterprise cabinets and cages rather than for training clusters.
The scale gap is worth stating plainly, because it is what makes this market easy to misread. Charlotte City Council rezoned 156 acres off Moores Chapel Road in May 2025 for a Digital Realty campus of up to 3 million square feet and, on the company's own figures, up to 400 MW. That single approval is larger than the entire operating multi-tenant base of the metro. It is also years of phased construction away from being anything, which is why the table below is a forecast of pressure on the utility rather than a shortlist.
Charlotte's other structural argument is position. It sits between Northern Virginia and Atlanta, which makes it a credible third site for an organisation already in both, and it is far enough inland to be a different risk profile from the coastal markets while staying inside the same fibre routes.
Buying power inside a vertically integrated market
This is the part that travels least well from other markets, and it is worth being precise about. North Carolina does not have retail electric choice. There is no competitive supplier to shop, no basis and delivery split on the bill to arbitrage, and no wholesale market price that a clever contract can track. Duke Energy generates, transmits and distributes, and the North Carolina Utilities Commission sets what it may charge. South Carolina, which starts twenty minutes south of uptown, works the same way.
The Carolinas also sit outside any regional transmission organisation. There is no PJM or ERCOT here running a capacity auction whose results roll into your rate. What exists instead is a utility resource plan and a rate case, both of them public, both of them slow, and both of them the actual venue where the cost of serving large computing loads gets decided.
For a tenant that has three consequences. Your operator's power cost is a published tariff rather than a negotiated supply contract, so ask which tariff and ask to see it. Any special treatment of very large loads arrives as a new tariff class rather than as a bilateral deal, which means it applies to your operator whether or not your operator likes it, and it is knowable in advance if you read the docket. And because the utility owns generation, the timeline for a new large service is a utility construction timeline, not a market signal: when a substation gets built is a planning decision, and no amount of willingness to pay compresses it the way it might in a market with merchant generation.
Ask an operator here whether the capacity being offered to you is already energised at the building, or is contingent on a utility upgrade that has not been completed. In a market with no retail choice and no merchant supply, that distinction is the whole risk. An operator with energised capacity is unaffected by the queue. An operator waiting on a substation is in a line whose length is set by a commission and a construction schedule, and no clause in your contract shortens it.
Three do not. There is no permit collector pointed at Mecklenburg County or the counties around it, so the public filing record is not shown here. Our facility inventory does not extend to this market, so there is no verified building count. And the interconnection layer counts a building only when its city sits on an explicit list, so Charlotte and its suburbs are counted nowhere rather than counted wrongly. The fourth is here. The announced capacity below is collected by hand, and every figure in it was read on the page it links to before the row was written, which is also why some widely quoted numbers for this market are in the note under the table rather than in it. Announced is not built, and built is not available. See how we verify capacity for what each collection proves.
What has been announced in Charlotte
Announcements are not filings. Nobody has to build what they announce, the figure is the one the developer chose to publish, and the date moves. They are here because megawatts are what a requirement is measured in, and because the size of what is coming changes who will take your call. Every figure below was read on the page it links to.
| Capacity | Operator | Project | What was announced |
|---|---|---|---|
| 400 MW | Digital Realty | Moores Chapel campus | 156 acres off Moores Chapel Road in west Charlotte, zoned for up to 3 million SF across two phases. Charlotte City Council approved the rezoning on 19 May 2025 |
One row, and the two projects quoted most often in this market are deliberately not in it. 5C describes its Charlotte campus as opening at 50 MW of critical IT capacity on a 161 acre site designed to scale to 1,600 MW. That ceiling is a design intention rather than capacity anyone has announced building, and the phase one date the company published has passed without a confirmation, so counting it would flatter this market by a factor of four. Amazon's campus at Hamlet is the largest project in the state at $10 billion and has no published megawatt figure at all, which is why it has no row. It is also 80 miles east of the city and single tenant, so it adds nothing a Charlotte buyer can lease. Read the reporting.
Announced capacity is collected by hand from trade reporting and company statements, and each figure links to the report it came from. It is kept apart from the permit record above and never added to it: one is a document a jurisdiction issued, the other is a plan a company published. Neither is availability.
Where the multi-tenant buildings actually are
North Charlotte and the University area
The University Research Park corridor is where most of the metro's purpose-built colocation ended up, and it is where Flexential and TierPoint both hold positions. It suits enterprise production and recovery workloads, has room in the buildings that exist, and is the first stop for a requirement measured in cabinets and cages rather than megawatts.
South Charlotte
The second cluster, closer to the corporate campuses along the southern arc of the city. A genuinely different product from the north: easier for staff to reach from the office parks, generally smaller floor plates, and worth testing when the deployment needs people to be able to touch it.
The airport and west side
Industrial land, good road access and the utility infrastructure that came with a century of manufacturing. This is where conversions and larger single-tenant footprints tend to appear, and where a wholesale requirement is more likely to find a shell than a fitted suite.
The greenfield ring
Cabarrus, Iredell, Rowan, Gaston and York County across the South Carolina line. This is where campus-scale proposals go, on land the city does not have, and almost none of it is colocation. Treat activity out here as information about the utility and the politics, not as inventory you can shortlist.
When Charlotte is right, and when it is the wrong choice
Charlotte is the right answer when your business is here. For a bank, an insurer, a health system or a manufacturer with people in this metro, a market with real enterprise-grade buildings, straightforward staff access and a utility with a long operating record is exactly what the requirement calls for. It is also a sound third site for an organisation already split between Northern Virginia and Atlanta, because it is independently exposed without being far enough away to complicate replication.
It is the wrong choice if you are buying on the strength of the state's headline projects. The campuses that put North Carolina in the news are single tenant, they are an hour or more from the city, and not one of them adds a square foot you can rent. Read them as pressure on the utility and on local politics, which is real and will reach your bill, rather than as supply.
It is also the wrong choice if your requirement is genuinely network led. This is not a carrier hotel market on the scale of Ashburn, Chicago or Los Angeles, and a workload that needs to sit inside the densest possible interconnection ecosystem should be tested against those markets before it is tested here.
Ask which Duke tariff the building is served under and ask for the rate schedule, because in a market with no retail choice that document is the cost basis of everything you will be quoted. Ask whether the space on offer is energised today or waiting on a utility upgrade, and get the answer in writing with a date attached. Ask what happens to your rate if a new large-load tariff class is approved while your term runs, and specifically whether the operator can pass it through and on what notice. If a proposal leans on a hyperscale project elsewhere in the state, ask what that project changes about the building you would occupy, and expect the honest answer to be nothing. And because this is an enterprise market rather than a wholesale one, ask about growth inside the same room rather than the same campus: the constraint here is usually the next four cabinets, not the next four megawatts.