Explainer

Why nobody can tell you what your router costs to run

A power adapter's rating is the most a router is allowed to draw, not what it actually draws, and almost every specification sheet stops there. BT is the exception: two of its own documents put a real number on the Smart Hub 2, on top of the adapter rating rather than instead of it — which is the only reason this piece can price a router at all.

Written by TechCritic's automated editorial pipeline. No human author is claimed, and no person reviewed this page before it was published. Every figure below was checked by the automated build: it recomputes each cost from the inputs printed beside it and refuses to publish if the sums disagree. Published 9 August 2026. TechCritic is published by Tireless, which sets out how it makes articles in its editorial policy.

What a router's spec sheet gives you

Electricity is billed by the kilowatt-hour, and working out what a device costs to run needs exactly one more number: how many watts it draws. A router's box, its quick-start guide and its online specification page all look like they should have that number somewhere. What they actually give you, almost without exception, is the rating printed on the power supply — a voltage and a current, describing the most the adapter is built to deliver, not what the router actually asks of it.

TP-Link's own specification page for the Archer AX55, a mainstream dual-band Wi-Fi 6 router sold in the UK, lists the power supply as 12V, 2A.2 Multiply those two together — TP-Link does not — and you get 24 watts. That is not a measurement of the router. It is the ceiling its power supply is built not to exceed, and the specification page prints nothing else beside it: no idle figure, no typical figure, no measured figure of any kind.

Netgear is more explicit about it, if you read the explanation rather than just the number. Its own support article on router power consumption states plainly: "Your router's power consumption does not exceed the power adapter rating."3 That sentence is the closest thing to an official position on the question this piece could find from any manufacturer besides BT, and it amounts to an upper bound offered as reassurance. The same article lists adapter ratings for dozens of Netgear router models — the RAX10, a current Wi-Fi 6 router, is rated at 12V, 1.5A, which multiplies out to 18 watts3 — and not one measured wattage among them.

Treat a rating like that as though it were a real answer and the arithmetic still runs: cost = (watts ÷ 1000) × hours × unit rate. At the price cap for 1 July to 30 September 2026 — 26.11p per kilowatt-hour, direct debit, including VAT1 — running the Archer AX55 continuously at its adapter's full 24W ceiling for a year would come to £54.89. TP-Link has never claimed the router draws that much continuously; nothing sold as a home Wi-Fi router is designed to sit permanently at its power supply's limit. But a specification sheet that offers only the ceiling invites exactly that miscalculation, because the ceiling is the only number on the page.

The one router that gives you more

BT is the exception, and it is an exception in an unusual way: it publishes not one extra number but two, in two separate one-page documents that sit among its support downloads rather than on the product page most buyers ever see.

The first is the Smart Hub 2's power supply, described in BT's own "Power Efficiency info" sheet: output voltage 12.0V, output current 1.5A, output power 17.7W — a figure BT states directly rather than one we have multiplied out ourselves.4 That is the same kind of ceiling TP-Link and Netgear publish and stop at. Except BT does not stop there.

A second document, "Power Consumption info" for the same product, gives what the first document and every other manufacturer's page never do: a measured figure for the hub itself, not its power supply. BT states 14.10W "in operation", 8.48W in "network standby", and 0.04W in "off mode", and adds one further detail neither figure alone explains: the hub drops from "in operation" into "network standby" after 20 minutes without traffic.5 That 20-minute figure says how quickly the hub reacts, not how much of a day it spends in either state, so it narrows nothing about the range this piece goes on to give. The off-mode figure does not apply to a router actually providing internet access, so it plays no further part here.

"Network standby" is not a phrase BT invented for a leaflet. It is a defined term in the UK's retained ecodesign regulation on standby and off-mode power consumption, which sets the condition as equipment able to "resume a function by way of a remotely initiated trigger from a network connection", and which carves out a category of "high network availability equipment" (HiNA) — router, network switch, wireless network access point, hub, modem, VoIP telephone and video phone, named as such — allowed a higher networked-standby ceiling than ordinary electronics precisely because they have to keep listening to the network even when nobody is using them.6 That ceiling has moved twice, not once. From 1 January 2017 it was 8.00W for HiNA equipment and 3.00W for other networked equipment; from 1 January 2019 a further provision tightened the non-HiNA ceiling to 2.00W and left the 8.00W HiNA ceiling exactly where it was.6 The Smart Hub 2 is a router, a modem and a hub in one box, which puts it squarely inside the regulation's own list of HiNA equipment — so the figure that actually applies to it has been 8.00W since 2017, not the 2W most other networked kit now has to meet.

Put the two numbers next to each other and BT's own measured 8.48W in network standby sits fractionally above that 8.00W ceiling — 0.48W over, or six per cent. We are not calling that a breach. The regulation specifies an exact test method for the figure it sets, BT's document does not state that it used that method rather than one of its own, and a gap this small is well inside what a different test procedure could explain. What the near-miss does establish is that "network standby" is a real, regulated state with a real ceiling, not marketing language, and that BT's own number sits close enough to it to be worth printing rather than rounding away.

Put all three routers side by side and the asymmetry is the whole story:

BT states its adapter's wattage directly. TP-Link and Netgear do not, so the 24W and 18W figures in that row are volts multiplied by amps — arithmetic we performed, not numbers either company printed.

What that costs, on BT's own numbers

BT's own figures are the only ones in this piece that can go through the sum honestly. A router does not have a "typical hours a day" the way a games console does — it is designed to be left switched on — so every row below runs the same 24 hours a day, 365 days a year, at the cap rate. The only thing that changes between rows is which of BT's own published numbers is doing the work.

The top row is what you get by doing to BT's own adapter rating exactly what a reader is forced to do with TP-Link's and Netgear's: treat the power supply's ceiling as though it were the router's draw. On BT's own real figures, that overstates the annual cost by between a quarter and more than double, depending on which of BT's two states the hub actually spends most of the day in — and BT does not say, so neither can we.

Why the others stop at the adapter

Nothing we read requires a manufacturer to disclose a router's actual power consumption to a customer — only, under the regulation described above, that the product itself meet the ceiling that applies to its category. Publishing the adapter's rating is close to a legal minimum: it has to be printed on the adapter for safety reasons regardless, so it costs a manufacturer nothing extra to repeat it on a specification page. Publishing a measured figure for the finished product is a different undertaking — it means putting a real unit on a meter under a defined test condition and standing behind the number — and we found nothing that obliges any company to do that for a customer's benefit rather than a regulator's file.

That is presumably why BT's two-page disclosure is the exception rather than the rule, and it is worth being precise about what that observation is and is not. It is not an accusation that TP-Link or Netgear are hiding something unflattering — we have no measured figure for either router, so we have no basis to say their real numbers would look worse than BT's. It is an observation about what gets published by default when nothing requires more than the adapter's rating: a ceiling, not a cost.

What we could not find out

Two of the UK's other large ISP-supplied routers were the obvious next check, and neither yielded a manufacturer document we could cite. Searches for the Sky Broadband Hub and the Virgin Media Hub 5 return retailer listings, help-forum threads and replacement-adapter pages for both, and nothing from Sky or Virgin Media themselves in the shape of a specification or consumption figure. A retailer listing is not a source under this publication's own rules, and neither is a customer forum post, so neither router appears anywhere in this piece with a number attached, real or implied.

BT's own duty-cycle split is the other gap, and it is the one named in the method box above: BT does not say what share of a day a Smart Hub 2 actually spends "in operation" against "network standby", so the true annual cost of running one sits somewhere inside the £19.40-to-£32.25 range this piece can state, not at a single point inside it.

Article one in this series, what your technology actually costs to run, flagged the router as the worst-documented always-on device in a typical UK home and could not put a figure on it at all. This piece closes that gap for exactly one router, because exactly one manufacturer chose to let it be closed. The other two routers named here are not evidence that TP-Link's or Netgear's routers cost more or less than BT's to run. They are evidence that the question has an answer only when a manufacturer decides to publish one, and that most of them, so far, have not.

Read next

Explainer · Running costs · 8 August 2026

What your technology actually costs to run, in the UK

Explainer · Charging and batteries · 8 August 2026

What it costs to charge your phone and laptop for a year, in the UK

All articles, in the archive