How we calculate your carbon saving

What we are measuring

When one of your retired devices is refurbished and put back into service, someone somewhere does not buy a new one. The carbon saving is the manufacture that never happened.

Making a laptop is the carbon-expensive part of its life. Roughly four fifths of a laptop's lifetime emissions are spent before it is switched on for the first time, in mining, refining, component fabrication and assembly. Keeping a working machine in service avoids repeating all of that.

We measure the emissions avoided at manufacture. We do not count the electricity your devices used while you owned them, and we do not count the electricity the next owner uses. Reuse does not change either of those.

How the figure is built

We start with the embodied carbon of an equivalent new device. Embodied carbon means everything from raw material extraction to the finished product leaving the factory. Our figures come from manufacturers' own published product carbon footprints and from peer-reviewed research, not from our own estimates.

We then subtract three things.

• Refurbishment. Testing, secure data erasure, replacement parts and repackaging all carry a carbon cost. It is small next to manufacturing a new device, but it is real and we count it.
• Transport. Collecting equipment from your site and redistributing it to its next owner.
• Remaining life. A refurbished device has fewer years of service left than a new one. It does not fully replace a new purchase, it defers one. We apply a factor of 0.75, meaning we credit each reused device with three quarters of the manufacturing carbon of a new equivalent.

That last adjustment is a choice, and it is a conservative one. Many carbon claims in our industry credit the full manufacturing footprint. We do not, because a report you cannot defend is worth less to you than a smaller number you can.

We publish the effect of that choice. Every report shows what your figure would be at 0.6 and at 0.9, so your sustainability team can see how much of the result rests on this assumption.

Your reuse rate

Not every device can be reused. Some arrive already broken, some are too old to have a market, and some have to be destroyed for security reasons.

We apply your actual reuse rate as a separate multiplier at the end of the calculation rather than hiding it inside the per-device figures. You can see it, and you can check it. Our current rolling twelve-month reuse rate is 92.5% (an estimate drawn from Terraguard Technologies' own database over the prior twelve months, refreshed quarterly, not an audited figure).

What happens to the rest

Devices we cannot reuse are dismantled and their materials recovered. Recovering steel, aluminium, copper and the precious metals in circuit boards avoids some of the emissions of producing those materials from ore.

That credit is real but small, and we report it as a separate figure rather than adding it to your headline saving. Combining the two would count the same benefit twice, once as avoided manufacture and again as recovered material. Your recycling credit for this period appears as a footnote to the main figure.

Less than 5% of what we collect goes to landfill.

Where our numbers come from

Every device category in your report carries a source and a confidence rating.

Our primary sources are manufacturers' published product carbon footprint datasheets, including Dell, HPE, Apple and Schneider Electric. For laptops, phones, tablets and monitors we use the study by Lovehagen, Malmodin, Bergmark and Matinfar, published in Renewable and Sustainable Energy Reviews in 2023. Our material recovery credits use the UK Government's DESNZ greenhouse gas conversion factors. The remaining-life adjustment follows the displacement work of Zink and Geyer in the Journal of Industrial Ecology.

Some categories have thinner evidence than others. Projectors, scanners and loose components such as memory and processors have no good published data, so we apply a deliberately low default rather than a confident guess. Large rack servers above 2U are credited at the 2U figure, which understates them, because no manufacturer publishes the number we would need.

Where we cannot source a figure, we say so and we round down. You will not find an invented number in your report.

What changed in version 2.0

We rebuilt this methodology in 2026 and most of our figures went up. Servers, network equipment and uninterruptible power supplies were the largest corrections. Our earlier model understated the carbon benefit of returning that class of equipment to service.

Reports issued before 1 July 2026 used the previous factors. We have left them as they are so your year-on-year records stay consistent. If you need an earlier period restated on the new basis for your own reporting, ask us and we will do it.