Reducing packhouse waste: you cannot cut what you do not measure

A weight without a reason is not actionable

Knowing you dumped 300 kg this week tells you that you have a problem; knowing 220 kg of it was decay from one grower's Thursday delivery tells you what to do about it. The fix is different for every cause: undersize points at picking maturity or grading settings, bruising points at line handling, decay points at the cold chain or time-to-cool, and each is a different conversation.

The practical discipline is small: when a pack run records a waste output, tag a reason from a short fixed list. Six options cover almost everything, undersize, damage, decay, defect, overripe, other. Free-text notes rot; a short coded list stays reportable.

Where the loss actually happens, stage by stage

Waste does not spread evenly across the operation, and treating it as one number obscures where the fix needs to go. Field and receival losses show up as fruit that never should have left the block: oversize or undersize picked to a quota, sunburn, and physical damage from bins stacked in the sun before pickup. A receival grade that only logs a percentage rejected, with no reason attached, hides whether the problem is the grower's picking standard or the shed's own handling once the bins arrive. Grading losses are the largest and most visible line for most crops, the fruit a machine or a sorter's eye pulls for size, colour, or blemish, and also the line a packhouse can move furthest with a single settings change or a retrained crew.

Handling damage accumulates in the gap between intake and pack: a drop off a conveyor lip, bruising from an overfilled bin tipped too fast, a chain speed mismatched to fruit firmness. It is the category most packhouses undercount, because fruit that bruises today often does not show a mark until two days into transit, by which point it reads as a cold-chain or transit claim rather than a handling one. Cold-chain losses are actually two different failures wearing one name: fruit that was never cooled fast enough after picking, a field-to-forced-air-cooling timing problem, and fruit that cooled properly but then sat through a defrost cycle or a door left open, a facility discipline problem. The two need different fixes and get conflated when the only record is a temperature log nobody reviews.

Overproduction waste, packing more of a line or a spec than the order book calls for, is rarer in contract packing but common where a shed packs to a forecast rather than a confirmed order. It shows up as finished stock ageing past its window rather than as a grade-out at all, which is why it is easy to miss in a waste report built only around the pack line. Packaging failures, a liner that punctures, a tray that collapses under stack weight, a carton whose board grade was under-specified for the route, convert sound fruit into a claim after it has already left the shed, and belong in the same waste accounting even though the loss lands on someone else's dock.

Set a baseline before you change anything

The instinct on discovering a waste number is to start fixing the loudest cause immediately. Resist it for one full cycle of the crop first. A baseline needs coded waste data, reason and destination both, across a normal run of receival days rather than one unusually good or bad week, because a two-day sample overweights whatever happened to go wrong on those two days. Express the baseline as waste per unit packed, not an absolute weight: a slow week and a peak week are not comparable in raw kilograms, and a raw-weight comparison will read as improvement or decline that is really just volume moving.

Once the baseline exists, changes can be judged against it rather than against a feeling. Less fruit dumped this week than last is not evidence of anything on its own; grading intensity, incoming fruit quality, and even the weather move the raw number more than most single interventions do. Comparing the post-change waste-per-unit rate against the same rate from the baseline period, on a similar mix of varieties and grower lines, is what turns a hunch that something helped into a number a buyer, an auditor, or a lender will accept as evidence.

Destinations are the half buyers now ask about

Where waste goes is becoming a compliance and commercial question, not just an operational one. Retailer programs ask suppliers to measure and publish food loss and waste, GLOBALG.A.P. audits carry a waste-management section, and some jurisdictions now expect records of donated food. Recording a destination per waste output, dump, compost, stockfeed, processing or donation, is what makes those answers a lookup instead of a scramble.

There is money in destinations too. Culls that leave as stockfeed typically return a small fraction of packed value per kilogram; culls sold into juice or processing grade usually return more than stockfeed, though still well under fresh-pack price; both beat the zero return, and the disposal cost, of landfill. Donation returns no revenue but can carry a tax or compliance benefit worth recording on its own terms, and some retailer programs now score a supplier's donation volume alongside its waste percentage. None of these outlets are worth chasing for a single surplus load; they earn their keep as a standing arrangement a packhouse can route grade-outs to without a phone call every time.

The grading line trades waste for price, on purpose

Every grading standard is a waste decision wearing a quality label. Tightening a size or colour tolerance to chase a premium-grade price lifts the reject rate on the same intake, and the packhouse only comes out ahead if the extra margin per unit sold clears the extra volume pushed into a lower-value or zero-value outlet. Loosening the tolerance does the reverse: more units pack to the higher grade, but some of them are fruit a fussy buyer would have rejected on arrival, which converts a packhouse-side waste line into a customer complaint or a rejected pallet, and that costs more than the waste line did.

The decision is worth making deliberately rather than leaving it to whatever the grading crew's habit is that week. A useful check is to run the packout and waste numbers for a week at the current tolerance against a trial week at a one-notch looser or tighter setting, on comparable intake, and compare the blended return per tonne of intake, not the packout percentage alone. A higher packout at a lower average price is not automatically the better outcome, and the only way to know is to run the two numbers side by side.

Rank the fixes by payback, not by drama

Not every cause is worth fixing first, and the fixes that look most impressive on a walkthrough are not always the ones with the best return. Cold-chain timing is usually the highest-payback fix available: moving fruit into forced-air cooling sooner after picking, or closing the gap between receival and cooling by even an hour, reduces decay losses that would otherwise not show up until the fruit is already sold. The fix is scheduling and layout discipline, not capital, which is why it pays back fastest.

Handling damage is next: a conveyor drop height, a bin-fill habit, a chain speed mismatched to a delicate variety. These cost little to fix once identified, but finding them takes a deliberate walk of the line with someone watching for where fruit actually gets knocked, rather than assuming the grading machine is the only source of damage. Grading settings and crew training rank below that: useful, but slower to show a return because they trade against price, as above, so the gain has to be measured against the blended return, not just the waste line. Packaging upgrades, a better liner, a tray rated for the real stack height, a board grade matched to the transit route, sit last on payback for most sheds, because the capital cost is real and the damage they prevent is often already small once handling upstream is fixed. They are worth doing, but doing them before the handling and cold-chain fixes usually means paying to fix a symptom of a problem that is still there underneath.

Make the report fall out of the day’s work

None of this survives peak season as a separate clipboard exercise. The reason and destination have to be captured in the same motion as recording the pack run, two taps at the moment the grade-out is weighed, and the report has to compile itself. Report the number as a rate against volume packed, not a raw weight, for the same reason the baseline has to be: a rate is comparable week to week and season to season, and it is what makes a claimed improvement defensible to a retailer, an auditor, or a lender asking for evidence rather than a story.

Fernable works this way: waste rows on a pack run take an optional reason and destination, and the waste report groups the period by crop, reason and destination with each line's share of the total, ready for a buyer, an auditor, or a grower conversation. Uncoded waste stays visible in the report rather than disappearing, so you can see how complete your own discipline is, and where the next round of coding effort should go.