The factor table
Every assumption behind these figures.
Circularity · July 14, 2026
Half-year 2026: what changed
Six months of figures, a new bottle supplier, and a problem we finally made progress on.
Order desk
Talk to the people who wrote this
Questions about anything on this page reach the same yard it was written in. We answer them whether or not you are buying.
Circularity6 min read
Twenty-one thousand containers through the gate in six months, an 83.1 % reuse rate, and the first movement in three years on the landfill fraction.
Six months
| Measure | H1 2025 | H1 2026 | Change |
|---|---|---|---|
| Containers recovered | 19,640 | 21,080 | +7.3% |
| Returned to service | 16,090 | 17,517 | +8.9% |
| Reuse rate | 81.9% | 83.1% | +1.2 pts |
| Tonnage diverted | 1,144 t | 1,231 t | +7.6% |
| Tonnage landfilled | 48 t | 47 t | −2.1% |
| CO₂e avoided | 1,556 t | 1,684 t | +8.2% |
| Fresh water per container washed | 11.2 gal | 10.4 gal | −7.1% |
| Rainwater share of wash feed | 20% | 26% | +6 pts |
| Own-fleet empty miles | 30% | 27% | −3 pts |
| Rebottles completed | 2,910 | 3,720 | +27.8% |
The landfill fraction finally moved
For three years the non-recoverable fraction has sat stubbornly at about 4 % of end-of-life mass — gaskets, adhesive labels, valve elastomers and residual product — and two attempts to find a route for it failed. This half it came down, in absolute terms, on higher volume.
The cause is unglamorous: we changed how labels come off. A steam-and-scrape station ahead of the granulator, which we built in February largely to speed up the line, turns out to remove adhesive labelling cleanly enough that the residual adhesive no longer contaminates the bottle stream. That means less material rejected at the regrind stage, which means less of it going to landfill.
- Label and adhesive to landfill: down about 31 % per container processed.
- Regrind rejected for contamination: down from roughly 3.1 % to 1.4 % of output.
- Net effect on the landfill fraction: from 4.0 % to about 3.7 % of end-of-life mass.
- Gaskets and elastomers: unchanged. Still no route. Still looking.
Rainwater at 26 %
A wet spring. The array supplied 26 % of the wash line's feed water over the six months against 20 % in the same period last year, and that is almost entirely rainfall rather than anything we did. It will fall back through the summer as it does every year.
Worth restating the awkward part: the array contributes least in drought years, which is exactly when municipal supply is most stressed and the contribution would be most valuable. Storage is the constraint, and we have not yet found a way to justify the capital for a materially larger array on a benefit that appears in the wrong years.
Method, unchanged
Unit counts from the grading station log. Tonnage from scale-certified weigh tickets, retained seven years, available to any customer for their own containers. Carbon calculated against a stated virgin-purchase baseline using the published factors — estimates, labelled as such. Water from the line's meters, read monthly. Empty miles from fleet telematics.
Full-year figures in December, including whatever the summer does to the water numbers.
Written by
Priya Raghavan
Operations
Routes the trucks, prices the freight and keeps the diversion figures honest enough to publish.
reportinghalf yeardiversion
»Read next