A school with a few hundred computers is quietly running a small power plant's worth of decisions. Whether the machines sleep when nobody is using them, which models get bought, how long they are kept, and what happens to them at the end all have measurable costs in energy, money and waste. The same is true at the other end of the scale: every search, stream and cloud-saved file runs on servers in a data centre that draws power around the clock. Environmental stewardship in computing means making those decisions deliberately, with numbers, rather than by habit. The good news is that most of the gains come from settings and policies, not expensive new hardware.
Where a device's energy goes over its life
A device's life has five stages: extracting the raw materials, manufacturing, transport, use, and end of life. For laptops, tablets and phones, manufacturing is usually the largest share of the lifetime footprint, because making chips and displays is extremely energy-intensive and the devices themselves sip power once built. For desktops, servers and gaming machines that run for many hours at high power, the use stage grows much larger. Where the electricity comes from matters too. Most of Ontario's electricity is generated from nuclear and hydro, so a kilowatt-hour used here carries less carbon than one from a coal-heavy grid — which makes manufacturing, which often happens elsewhere, an even bigger share for an Ontario user. The practical conclusion is the same either way: the greenest device is usually the one you already own, kept running longer.
From watts to kilowatt-hours to dollars
A power rating in watts (W) says how fast a device uses energy. Your electricity bill charges for the total energy used, measured in kilowatt-hours (kWh): one kilowatt-hour is a 1000 W device running for one hour, or a 100 W device running for ten. So the method is always the same three steps. Convert watts to kilowatts by dividing by 1000, multiply by the hours of use to get kWh, then multiply by the price per kWh. In Ontario the price of the electricity itself depends on the time of day and the price plan — very roughly 8 to 20 cents per kWh on the common plans — with delivery and regulatory charges on top. This lesson uses 0.12 dollars per kWh as a round, approximate rate; always state the rate you assume, because the answer scales directly with it.
| Device and use (approximate power) | Energy per year | Cost per year at 0.12 dollars/kWh |
|---|---|---|
| Gaming PC under load, 350 W, 3 h a day | 350 ÷ 1000 × 3 × 365 = 383.25 kWh | about 45.99 dollars |
| Game console left in standby, 10 W, all day | 10 ÷ 1000 × 24 × 365 = 87.6 kWh | about 10.51 dollars |
| Monitor, 30 W, 10 h a day | 30 ÷ 1000 × 10 × 365 = 109.5 kWh | about 13.14 dollars |
| Phone charging, 5 W, 3 h a day | 5 ÷ 1000 × 3 × 365 = 5.475 kWh | about 0.66 dollars |
Policies that cut energy and waste
| Policy | What it requires | Why it works |
|---|---|---|
| Power management | Displays off after about 10 minutes idle, sleep after about 15 to 30, scheduled shutdown overnight | Sleep draws a few watts instead of tens; most machines are idle most of the day |
| Purchasing standards | Buy ENERGY STAR certified and EPEAT-registered equipment, sized to the actual need | Efficiency is built in for the device's whole life, at no extra effort |
| Longer refresh cycle | Keep devices five or more years; repair and upgrade memory or storage first | Every year of extra service postpones a new device's manufacturing footprint |
| Reuse before recycling | Refurbish and donate working devices | Reuse keeps the whole device in service; recycling recovers only materials |
| Secure, responsible disposal | Wipe all storage, then use an authorized e-waste program and keep records | Protects personal data and keeps lead and mercury out of landfill |
Two labels make purchasing policies easy to write. ENERGY STAR, a program of the United States Environmental Protection Agency that Natural Resources Canada administers here, certifies products that meet energy-efficiency targets, including low power use in sleep and standby. EPEAT, run by the Global Electronics Council, goes further: it rates products across their whole life cycle — energy use, hazardous materials, recycled content, ease of repair and recyclability — in Bronze, Silver and Gold tiers. At the end of a device's life, Ontario makes producers, the companies that sell electronics, responsible for collecting and recycling them. Since 2021 this has been set out in a regulation under the Resource Recovery and Circular Economy Act, 2016, overseen by the Resource Productivity and Recovery Authority. In practice that means free drop-off at municipal depots, participating retailers and producer programs such as Recycle My Electronics, which lists locations across the province.
Greener code and greener data centres
Software decides how hard the hardware works, so programmers have their own lever. An efficient algorithm does less work for the same answer — binary search's 20 checks against linear search's million is an energy saving as much as a time saving, and it is multiplied across every user and every run. Moving less data helps too: compressing images, caching results instead of fetching them again, and not polling a server every second when a notification would do. In the cloud, the biggest waste is often resources nobody is using, such as test servers left running all weekend. Data centres measure their overhead with power usage effectiveness, PUE: total facility power divided by the power reaching the computers. A PUE of 1.5 means half as much energy again goes to cooling and power conversion; the most efficient large facilities report figures close to 1.1. Choosing a provider region with a low-carbon grid, such as Ontario or Quebec, can matter as much as any efficiency measure.
- Computers left on overnight and all weekend
- Replacing devices every two or three years by default
- A brute-force algorithm run millions of times a day
- Full-resolution images sent to every phone
- Cloud test servers that are never shut down
- Sleep and shutdown policies applied centrally
- Repair, upgrade, then reuse before recycling
- An efficient algorithm on sorted or indexed data
- Compressed, right-sized images and cached results
- Idle resources scheduled to shut down automatically