Operators running returnable containers usually describe their losses as theft. In eight years of running loops for other people we have seen very little theft and an enormous amount of something else.
Where they actually go
Dwell — by far the biggest category
A customer receives a container, empties it, and puts it in a corner. Nobody asks for it back. Six months later it is still there, and it is not lost — it is parked. In the fleets we manage, dwell accounts for something like two thirds of units that are unaccounted for at any moment.
The reason is almost never obstruction. It is that returning packaging is nobody's job at the receiving site, and until somebody asks, nothing happens.
Repurposing
The container becomes a water butt, a wash station, a grit bin. Whoever did it does not consider this taking something — the container was sitting empty and they had a use for it. This is a large and completely invisible category until you start tagging units.
Scrapped by a third party
A site tidies up, a contractor clears a yard, and containers with somebody else's name on them go to a scrap merchant with everything else. Recoverable if you catch it fast; usually you do not.
Site closures
A customer relocates or goes under, and whatever is on the floor goes with the building. This one is genuinely difficult to control and it is why we flag customers holding disproportionate numbers of units.
Actual theft
It happens, mostly for scrap value, and it is a small fraction of the total. A container is worth $18–$95 as scrap, which is not enough to motivate much.
The report that changes behaviour
Not a loss report — a dwell report. Units currently held, by customer, with days held. Nearly every operator we onboard finds one or two customers sitting on a disproportionate share, and nearly every one of those customers is genuinely unaware of it. A polite email listing serial numbers recovers containers that a loss process never would.
What actually recovers a fleet
- Serialised tagging. You cannot manage what you cannot count, and "about forty out there" is not a count.
- Scheduled geographic runs rather than ad-hoc collection. Customers can have containers ready on a known day; they cannot drop everything for a same-day request.
- Naming a person at the customer's site. Fleet recovery is a relationship with a specific human, not a clause in a contract.
- Asking early. A container asked for at four weeks comes back. One asked for at eight months has usually become something else.
What does not work
Deposits, in our experience. They generate billing disputes out of all proportion to the containers recovered, and they make the commercial relationship adversarial over an item worth less than the argument.
Penalty clauses have the same problem. The containers that go missing are mostly not going missing through bad faith, so punishing bad faith recovers very little.
Realistic targets
A well-run loop recovers 85–92% of its fleet annually. Below 80% something structural is wrong — usually a customer segment where retrieval is not economic, which should be served with one-way packaging rather than fought over.
Above 92% is achievable but usually costs more in chasing than the containers are worth. There is an optimum, and it is not 100%.
Priya Raghunathan is the compliance & documentation at IBC Tote Recycling NY in Astoria, Queens. We publish the figures we work from — if you think one of them is wrong, tell us.
