This is a methodology post, and it is about the number that makes reuse reporting either credible or worthless.
Where containers actually go
A pooled tote that stops cycling has not usually been stolen. We tag assets and chase them, and when we find out what happened the answers are mundane.
| Outcome | Share of unreturned | Notes |
|---|---|---|
| Repurposed on the receiving site | 34% | Water storage, waste collection, ballast |
| Scrapped locally | 22% | Damaged and disposed of without telling us |
| Still on site, forgotten | 19% | Often recovered on a later collection run |
| Sold or given away | 11% | Usually to staff or a neighbouring business |
| Genuinely unknown | 14% | No trace, no explanation |
Fate of unreturned pool units, where established
The 34% is interesting and slightly cheering: a tote repurposed as a rainwater tank behind somebody's workshop is still in service, just not in ours. It is also uncountable, which is the point of this post.
The temptation
Here is how reuse numbers inflate, and it is rarely dishonest in intent.
Suppose a pool averages 7.4 turns per bottle. A unit that completes three cycles and then vanishes could reasonably be assumed to have continued somewhere — repurposed, as a third of them are. If you credit it with the pool average, you have just booked 4.4 cycles that nobody observed.
Do that across 12% of a 2,000-unit pool and you have invented roughly 1,050 cycles a year. At 126 kg CO₂e avoided per cycle equivalent, that is 132 tonnes of avoided emissions that exist only in a spreadsheet.
The gap between a defensible reuse figure and a flattering one is almost entirely what you assume about the units you stopped seeing.
What we count
Three rules, applied without exception.
- Observed cycles only. A unit is credited with the cycles that crossed one of our six scan points. A cycle we did not see did not happen for reporting purposes, even if it almost certainly did.
- Attrition appears as its own line. The diversion certificate shows units dispatched, units returned and units unaccounted for. A reader can see the loss rate without doing arithmetic.
- No end-of-life credit for lost units. We do not claim recovery of material we never received. If we did not weigh it, it is not on the tonnage line.
The six scan points
For completeness, these are the moments a pooled unit is observed, because "observed cycle" only means something if you know where the observations happen.
| # | Event | What it establishes |
|---|---|---|
| 1 | Dispatch from float | Unit left clean and tested, with its serial on the shipping document |
| 2 | Delivery at filler | Dwell clock starts |
| 3 | Despatch filled | Confirms the fill and the destination |
| 4 | Collection from receiver | Closes the outbound leg; this is the scan most often missed |
| 5 | Wash and test | Condition record and periodic test status |
| 6 | Return to float | Cycle complete, counted |
Scan points in a pool cycle
A cycle counts when scan six happens. If a unit reaches scan three and then disappears, it is credited with the cycles it completed before, and the incomplete one is not counted at all — not half-counted, not estimated.
What this costs us
It makes our headline numbers smaller than they could be, and it makes them smaller than some competitors' numbers for comparable operations. I have had that comparison put to me by a prospective customer's sustainability team, which was an uncomfortable and ultimately useful conversation.
What it buys is that every figure on our certificate traces to a scan or a scale ticket. When an auditor asks where a number came from, there is an answer that is not a model.
How attrition shows up in pricing
Practically, attrition is priced into the turn rate up to an agreed threshold. Above that we invoice unreturned units at reclaimed replacement value rather than new-container cost, which is one of the quieter advantages of pooling reclaimed rather than new assets — the loss hurts a fraction as much.
That also lets us set a realistic threshold instead of policing every unit, which is better for everyone. Chasing the last 3% of a pool costs more than the containers.
How to cut attrition, in order of effect
From what we have seen work across managed pools, roughly in order of how much difference it makes.
- Scheduled collection rather than call-when-full. This alone typically halves the "still on site, forgotten" category.
- A visible asset tag with a return instruction. A plain QR sticker saying what to do converts a fair share of "repurposed" into "returned".
- Name the unit on the delivery note. Receivers treat a serialised asset differently from an anonymous container.
- Report attrition back to the receiving site. Nothing reduces loss like a quarterly note saying how many of their units did not come back.
If you run your own fleet
Most companies with captive container fleets do not know their attrition rate, and finding out is usually the single most valuable thing a reusable packaging programme can do in its first year.
- Count units purchased and units physically present once a year. The difference is your attrition, however uncomfortable.
- Tag assets. QR on the cage is enough; you do not need anything sophisticated.
- Measure dwell at the receiver — it is the biggest determinant of turns and the easiest to improve.
- When reporting, state observed cycles. Resist the pool average.
Our full methodology, including the coefficients and their uncertainty, is on our carbon math.
Written by
Theo Marsh
Materials & reporting
Owns the granulator, the baler and the diversion certificates. Writes the carbon methodology and defends the coefficients.