What Does It Actually Cost to Run an Old Phone as a Security Camera for a Year? The Electricity Arithmetic

Short answer: A phone on camera duty is one of the very few things in a house with no duty cycle at all — it is working for all 8,760 hours of the year. That sounds expensive and it is not. At a continuous 3 W and 20¢/kWh it comes to 26.28 kWh, about $5.26 a year. The number that actually decides the argument is the other one: for the electricity to match even a modest $36/year camera subscription, the phone would have to pull 20.5 W continuously — more than most phones draw while fast-charging, let alone while sitting still on a shelf. What I cannot give you is a measured wattage for your phone, and nor can anybody else, including the people selling you things. Below is the arithmetic, the three reasons the charger's rating does not answer it, and the roughly $15 way to settle it for your own setup. Apps like Background Camera RemoteStream run with the screen off, which matters here for a reason that turns out to be about arithmetic rather than marketing.

Almost every article about turning an old phone into a camera uses the word free. It is a word about the software. The setup still has a plug in it, and the plug is in the wall for a year.

Nobody publishes what that costs. So let's do it properly.

The one appliance in the house with no off cycle

Everything else with a plug spends most of its life idle. A kettle runs for perhaps five minutes a day. A washing machine runs a few hours a week. A television that feels like it is "always on" is realistically on for three or four hours a day. Even a fridge — the household byword for something that never stops — only runs its compressor for a fraction of each hour.

A camera has no duty cycle. That is the entire point of buying one. It is on at 3am on a Tuesday in February.

24 hours × 365 days = 8,760 hours per year
24 hours × 366 days = 8,784 hours in a leap year

Which means two opposite intuitions are both defensible. "It's only a phone, it barely uses anything" is true about the wattage. "It's running every hour of the year, that has to add up" is true about the hours. Only the multiplication settles which one wins, and the multiplication is not hard.

How do you calculate the yearly electricity cost of an always-on device?

Two lines:

kWh per year  =  watts × 8,760 ÷ 1,000
cost per year =  kWh per year × the rate on your bill

Here is the whole space of plausible answers, so you can find your own row rather than take mine. Rates are per kWh; use the one printed on your own statement, because it varies more between two houses than the wattage does between two phones.

The useful thing about a table rather than a single figure is that it shows you the shape of the answer before you know your own inputs. Across the entire plausible range — every wattage a phone could sustain, every domestic tariff I could think of — the answer is between one dollar and thirty-five dollars a year. No measurement you take is going to move it outside that box.

Run it the other way and it gets sharper. If you want to know what draw would make the electricity as expensive as a subscription:

watts = annual subscription cost ÷ rate per kWh ÷ 8,760 × 1,000

A phone would have to hold 20.5 W continuously, forever, to make its electricity cost as much as a $36/year plan at 20¢. Phones do not do that. They do not do that while gaming, and a phone with the screen off, sitting still, encoding a video stream, is a long way below the number it hits while gaming.

That is the real finding, and it is a boring one: the electricity is not the line item that decides anything. Which is worth knowing precisely so you can stop thinking about it and go and worry about something that matters, like whether the camera is pointed at the right thing.

Why you cannot read the answer off the charger

Almost every estimate of this on the internet is wrong in the same way — someone reads "20W" off the charger brick and multiplies. Three reasons that does not work.

A charger's rating is a ceiling, not a consumption. "20 W" is the most that adapter is willing to deliver under negotiation with a device that is asking for it. It describes the maximum of a range. A phone sitting at a full charge is at the bottom of that range, not the top, and the adapter delivers what is asked for rather than what it is capable of.

Charge state changes the regime entirely. Filling an empty battery and maintaining a full one are two different jobs with two different power profiles. If you measure for twenty minutes after plugging a flat phone in, you have measured the first job. A camera spends 8,758 of its 8,760 hours doing the second.

The adapter is on the meter even when nothing is attached. Converting mains AC to low-voltage DC is not free, and a plugged-in adapter with an idle cable draws a small amount continuously. It is genuinely small. It also runs for all 8,760 hours, which is exactly the multiplication this article is about, and it lands on your bill whether the phone wants anything or not.

There is a fourth problem, which is that the phone's own draw is not a constant either. A hallway at 3am and a driveway at 4pm are different amounts of encoder work. Somebody watching the live view is different from nobody watching. A weak Wi-Fi signal costs more radio power than a strong one. Any single instantaneous reading is a sample of a distribution, and you want the mean.

What actually moves the number

In rough order of size:

• The screen. On a phone, the display is generally the single largest consumer of power there is. This is why screen-off operation is not a cosmetic feature — a camera app that keeps a live preview lit is paying the biggest line item in the budget, every second, for a picture nobody is standing there looking at. I went into what that costs in heat and battery chemistry in Can you leave an old Android phone recording 24/7?; the electricity is the same story told in dollars instead.
• Encoding effort. Higher resolution and higher bitrate mean more work for the hardware encoder, more bytes to write, and more bytes to move. Dropping resolution is the single lever most people have and never pull.
• Radio. Wi-Fi on a strong signal is cheap. Wi-Fi at the edge of range is not, because the radio compensates with power.
• Viewers. Serving a live stream is real work, but it only happens while somebody is actually watching, which for most households is a handful of minutes a day.
• Ambient temperature. A warm phone throttles and its charging circuit works harder, so a sunny windowsill is worse than a shaded shelf for the bill as well as for the battery.

The measurement that settles it: a plug-in energy monitor

A plug-through energy meter costs roughly $10–20 and reports cumulative kWh. It is the only instrument in this article, and it turns every estimate above into a number about your actual house.

1. Plug the energy meter into the wall socket.
2. Plug ONLY the charger into the meter, with nothing attached to the cable.
   Leave it 24 hours. Record the kWh. This is your standby baseline.
3. Reset the meter. Now attach the phone, with the battery ALREADY FULL
   and the camera ALREADY RUNNING as it normally runs.
   Leave it 24 hours. Record the kWh.
4. Annual kWh = (step 3 reading) × 365.
5. Annual cost = annual kWh × the rate printed on your bill.

Two details are doing the work here.

Start from a full battery. Otherwise step 3 includes the one-off cost of filling the cell, which is not a running cost and will inflate your annual figure substantially.

Leave it a full 24 hours, not an hour. The draw is not uniform across a day, and a short sample lands you inside one regime and reports it as the average. Twenty-four hours covers a light cycle, a temperature cycle, and whatever normal viewing pattern your household has.

If your meter has a resolution problem at these power levels — many cheap ones do not resolve below about 1 W well — leave it running for a week and divide by seven. The instrument gets more honest the longer you give it.

What else is on the bill, and what isn't

The router was already running, and would be running whether or not there is a camera on the network, so it does not belong in this calculation.

There is no line item after that. That is the actual asymmetry, and it is not about watts.

The comparison I am not going to fake

The obvious next section would be a table comparing a phone's annual electricity against a Nest, a Ring and a Blink. I am not going to write it, because I have not measured any of them, and a table of invented wattages would be worse than no table.

What can be said without measuring anything: a dedicated cloud camera is also plugged in for 8,760 hours, so it has the same kind of small electricity number. Then it has a subscription, which is not small and which is printed on a page you can read. The running-cost case for reusing a phone was never an electricity argument. It is a subscription argument that people mistakenly try to defend with electricity.

An honest limitation, and a number we owe you

This is the part where I should give you our own measured figure, and I cannot, because we have not published one.

We have made public claims about how long a phone lasts running with the screen off versus with the screen on. Those claims are, I believe, directionally right — the display really is the largest consumer, and removing it really does change the picture. But directionally right and measured are different standards, and only one of them belongs in marketing copy. Until somebody puts a meter on it and publishes the method alongside the number, any single wattage or battery-life figure you read about phone-as-camera setups — ours included — is a claim rather than a measurement, and you are entitled to treat it that way.

The plug meter above is the arbiter. It costs about fifteen dollars, it does not care who made the app, and it will not tell you what anybody wants to hear. That is the whole reason to trust it.

What to do this week

If the running cost is genuinely a factor in your decision, do not take a number from an article, including this one. Buy the meter, run the two 24-hour readings, and put your own figure in the table above. It will take two days and settle the question permanently.

And if the running cost is not a factor in your decision — which, having done the arithmetic, is where I would expect most people to land — then the useful outcome of this article is being able to stop weighing it. Somewhere between a dollar and thirty-five dollars a year is not a decision. Where the camera points, whether it survives a power cut, and whether anyone would notice if it quietly stopped: those are decisions.

More at superfunicular.com, and the app is Background Camera RemoteStream on Google Play.