Insights

Reporting

Avoided CO2 from refurbished IT

Circuit board close-up used as a stand-in for electronics still in the use phase

Avoided CO2 from refurbished IT is the greenhouse gas you did not emit because a wiped, graded device stayed in use instead of triggering a new manufacture. It is not the same as recycling credits, and it is not a licence to invent "cars off the road" without a factor sheet.

Helene had to put a number in a 2025 sustainability appendix. A vendor offered a single heroic total. She asked for the laptop factor, the phone factor, and whether recycled units were in the reuse total. They were. She threw the slide out. We rebuilt from 91 reused laptops, 18 phones, and a conservative factor range. The appendix got a paragraph, not a firework.

You already feel the pressure to report. This article is how we keep ITAD refurbishment numbers defensible: boundaries, device types, exit channels and what we refuse to claim.

What "avoided" is allowed to mean

A refurbished laptop avoids the bulk of manufacturing emissions associated with a new laptop if someone would otherwise have bought new. That if is the whole game.

Safe claims:

Unsafe claims:

We report avoided CO2e by device type and exit channel. Employee channel and B2B channel can use different conservatism. Recycling gets mass and a WEEE destination, not a reuse factor.

If you need a factor table next to a Recover file, ask during the audit. Contact.

Boundaries: company, device, year

Write the organisational boundary like you would for any other environmental metric.

In. Devices that went through Recover and Refresh in the reporting period, with a serial, a wipe result and an exit code (employee, B2B, recycle, harvest).

Out. Devices still in a cupboard. Devices wiped in-house and never seen by us. Consumer take-back you do not operate.

Year. Exit date, not pickup date, if transit straddles December. Pick one and keep it.

Helene's first draft used pickup dates and counted a January 2026 workshop exit in 2025. We moved to exit dates. The number got smaller and the auditor got quieter.

Chain of custody from reverse logistics is what makes the count possible. No serial, no line in the CO2 file.

Factors: be boring on purpose

Laptop manufacture sits in a different order of magnitude than phone manufacture. Mixing them into "units" hides phones inside laptop glory or vice versa.

Publish:

We prefer a range over a fake point estimate when sources disagree. "12 to 18" is adult. "14.732" is not.

Do not convert to "trees" or "flights to New York" unless the reader asked for a cartoon and you still show the tonnes underneath.

Eurostat is context, not your KPI

Eurostat WEEE statistics are useful for the opening of a board memo: 5.2 million tonnes collected in 2023, 11.6 kg per inhabitant, 37.5 percent collection rate versus 14.4 million tonnes of EEE placed on the market.

Those numbers explain why reuse inside a company is worth doing. They are not your avoided CO2. Pasting them into a customer report as if you collected 5.2 million tonnes is how you lose the room.

Your KPI is reused units by class, recycle remainder, and CO2e from the factor table. Keep WEEE as a footnote about the system you sit in.

A worked example you can copy into an appendix

Suppose a 2025 exit file says:

You do not multiply 91 by a phone factor. You pick a laptop manufacture factor from your sheet. Imagine the conservative factor is 250 kg CO2e avoided per laptop reuse, and 45 kg per phone, and you apply a 0.7 discount on B2B because you are one step from the end user.

Employee laptops: 70 × 250 = 17,500 kg

B2B laptops: 21 × 250 × 0.7 = 3,675 kg

Employee+B2B phones: 18 × 45 = 810 kg (apply the same B2B discount if you split channels)

Total in this cartoon: about 22 t CO2e, reported as a range if your factors are a min/max. Recycled 29 + 7 units appear as mass to WEEE, not as 250 kg each.

The numbers above are illustrative for the method. Put your real factors in a table with sources. If legal will not let you publish the factor, still keep it in the working paper.

Priya asked if she could round 22 to 25 "for the intranet." We said 22, or "about 20." Rounding up is how appendices get restated.

Uncertainty, allocation, and the sentence you should refuse

Uncertainty is normal. Allocation is where people cheat.

Refuse:

If two factors disagree by 40 percent, print both and take the lower for external claims. Internally you can show the spread. External claims should be the ones you can still stand next to in twelve months.

Helene's auditor asked who chose the 0.7 B2B discount. We wrote "judgement, documented, conservative relative to employee channel." That is allowed. "Because marketing asked for a round 30 tonnes" is not.

What we put in the client working paper

A usable pack is short:

  1. Exit table: serials aggregated by class and channel (not a dump of every row in the PDF).
  2. Factor table with source and date accessed.
  3. Boundary paragraph (in, out, year rule).
  4. One worked example like the cartoon above, with the client's real counts.
  5. A sentence on what was recycled by mass, without a reuse factor.

That pack sits next to wipe certificates, not instead of them. Climate claims without ITAD evidence are a poster. ITAD evidence without a factor table is still a valid operations report; it just should not shout tonnes.

If a procurement team wants a single KPI for a vendor scorecard, reuse rate is healthier than CO2e. Reuse rate is countable from the same file. CO2e is an interpretation of that count.

Channel mix changes the claim

Employee programmes are the easiest story: a named person took a named serial. You can still overclaim if they would have used a 2016 brick forever. Stay conservative.

B2B lots are one step further from the end user. We still count reuse at exit from our workshop, and we do not invent the downstream retailer's second count.

Harvest and recycle: parts reuse can carry a smaller factor if you actually track the part. We often skip parts-level CO2 unless the volume is real. Honesty beats a parts novel.

Priya wanted to add "potential" reuse for units still in the shop at year-end. We said no. Inventory is not an emission avoidance until it leaves.

Operational KPIs that make the carbon number believable

Finance will not fight a CO2e line that sits next to:

If reuse rate is 40 percent and the slide screams a huge CO2e, someone will ask whether laptops were counted as data centres.

Device grading is not a climate tool. It still decides whether a unit can exit as reuse. A honest C sold to staff is reuse. A pretty fail that should have been shredded is a liability, not a tonne.

FAQ

Can we use market-based electricity for the workshop?

You can report workshop energy separately. Do not bury it inside avoided manufacture. They are different stories.

What if a buyer resells twice?

You count your exit once. Their LCA is theirs.

Do wipe certificates belong in the climate file?

They belong in the ITAD file. The climate file joins on serial. No wipe, no reuse claim, because you should not have remarketed it.

Should we get a third-party carbon opinion?

If the number is going into a regulated report, yes. If it is an appendix next to operational KPIs, a documented factor sheet may be enough. Ask your auditor, not your recycler.

Report a range you can defend

Avoided CO2 from refurbished IT is a conservative multiplication: reused units by class, times a published factor, with recycling in another column. The work underneath is Recover identity, Refresh wipe and grade, Remarket exit codes.

If you want that table next to a fleet you already have in a spreadsheet, send the mix. We will not inflate it. Request an audit.