Speed Is Not the Only Line Item: A Wi-Fi 7 Power Model for US Enterprises
US enterprises entering the 2026 wireless refresh cycle face a familiar sales narrative: higher throughput, lower latency, and a cleaner path to 6 GHz. The business case usually stops at access point count, license costs, and installation labor. It rarely includes the wattage that the new design will consume every hour of every day, or the cooling load that wattage creates. That omission matters because Wi-Fi 7 is not merely a faster radio standard. It is an infrastructure decision with a power profile.
The refresh case is incomplete by default
A Wi-Fi 7 design often calls for more access points, not fewer. Higher capacity per radio does not automatically reduce density in dense offices, warehouses, and campuses; it raises the ceiling on what each area can support and encourages more devices, more overlays, and more always-on radios. Add 6 GHz channels, multi-link operation, and PoE requirements, and the facility sees a new baseline load that persists long after the cutover party.
Most vendor proposals show a bill of materials, not a utility forecast. That is the gap. The CFO approves a capital project, but the facilities team inherits an operating cost. The two decisions should be evaluated together.
Where the watts actually hide
The first hidden load is access point density. Even if each new AP is more efficient per bit, the estate may consume more watts at idle because there are more radios listening, beaconing, and maintaining management sessions. The second is Power over Ethernet. Wi-Fi 7 APs can draw more power under load, and PoE switches must deliver that power with conversion losses. The third is cooling. Every watt delivered to an AP eventually becomes heat in the conditioned space, and every watt lost in a PoE switch or power supply adds heat to the telecom closet.
The fourth load is less obvious: the network edge that supports the refresh. Higher-speed uplinks, additional switching, and denser cabling can add small but persistent loads across closets that were never designed for the new thermal profile.
A practical TCO model
A defensible model does not need utility-grade precision at the proposal stage. It needs enough structure to separate vendor claims from facility reality. Start with a simple annual energy formula for the wireless estate:
Annual AP energy in kWh = AP count × blended watts per AP × 8,760 ÷ 1,000.
Use three wattage figures rather than one: idle, typical business-hours load, and peak. A blended figure should reflect actual occupancy, not a marketing average. Then add PoE conversion loss. If a switch delivers 30 watts at 85 percent efficiency, the input load is roughly 35 watts, and the difference is heat in the closet. Multiply that loss across ports and hours.
Next, add cooling. For an illustrative model, apply a cooling factor to the IT load based on the facility's actual HVAC efficiency. A closet with dedicated cooling will have a different factor than an open office served by a central plant. The point is not to guess perfectly; it is to show that the refresh changes both the plug load and the thermal load.
Finally, stage the rollout. A pilot of one floor or one site should be modeled separately from a full estate. Many refreshes are approved on a full-deployment energy assumption but billed during a long pilot-and-wave period when old and new equipment overlap. Overlap is a cost, not a transition detail.
Procurement language before signature
Energy telemetry should be a requirement, not a post-installation discovery. The request for proposal should ask vendors to provide wattage at idle, typical, and maximum load; PoE class and worst-case draw; supported power-saving modes and their operational trade-offs; and telemetry access through documented APIs. It should also require per-AP and per-switch power reporting at a granularity that matches the utility bill, together with retention terms and export rights.
A useful clause is simple: the vendor must supply energy consumption data in a machine-readable format for the life of the contract, at no additional license cost. Another clause should require disclosure of firmware changes that materially alter power draw. Without that, the facility team is modeling a product that may behave differently after the first update.
What the approval package should show
The final business case should show capital cost, annual energy cost, cooling cost, and the cost of overlap during rollout. It should also show the sensitivity: what happens if AP density is 10 percent higher, if PoE efficiency is lower than quoted, or if the utility rate rises. That is not pessimism. It is the minimum discipline for a multi-year infrastructure decision.
Wi-Fi 7 may well be the right refresh for a US enterprise. But speed is not the only line item. The facility bill arrives every month, and the wireless design determines part of it.