Ninety seconds on the grading line

Why daylight, and why it is not sentimental

We grade under natural light, and people assume it is a nice touch. It is not — it is a detection problem.

Overhead artificial light comes from a single direction and washes out the shadow that a hairline split casts. Under diffuse daylight the same split throws a visible line. A grader working under a shop lamp will miss cracks that the same grader catches outside.

The second reason is colour. Early mold on pallet wood is a grey-green tint that is genuinely difficult to distinguish from ordinary weathering under warm artificial light. In daylight it is obvious. Missing it means shipping a mold problem into a customer's food or pharmaceutical facility, which is the kind of mistake that ends a relationship.

The sequence, in order

1. Sight the stringers, both faces

The grader looks along the length of each stringer, the way you would sight along a plank for a bow. A split running with the grain shows as a dark line. Previous repairs — plugs, companions — are noted here, because either one caps the pallet at Grade B.

2. Check the four notch shoulders

The most common failure point, so it gets its own check. A crack radiating from the notch corner is the single most frequent reason a core drops out of Grade A.

3. Count and sight the top deck

Board count, gap consistency, and whether any board is split, cupped or missing. A grader does not measure — measuring every pallet would double the cost of the product. They compare against a reference pallet, and consistency between graders comes from working next to each other.

4. Flip and check the bottom deck

Bottom boards take fork strikes and floor impact. Missing bottom boards are common and often acceptable; a bottom board that is split *and* still attached is worse, because it will catch.

5. Hand check for protruding fasteners

Run a gloved hand across the deck. A nail standing a sixteenth of an inch proud will not be seen but will be felt, and it will shred a stretch wrap two states away.

6. Moisture and smell

Wet wood is heavier and sounds different when set down. Mold has a smell. Both are reasons to pull a pallet out of the recoverable stream entirely rather than route it to repair.

What the four routes mean commercially

• Grade A — original unrepaired stringers, seven or more deck boards, clean. The highest-value output and roughly 30% of intake.

• Grade B — repaired or slightly compromised but structurally sound. The workhorse, and the best value for most buyers.

• Repair — one fixable defect. Goes to the bay, comes back as Grade B.

• Dismantle — beyond economical repair. Recovered as component lumber and ground remainder.

The margin of the entire business sits in the accuracy of that split. Grade too generously and customers get pallets that fail. Grade too conservatively and you dismantle inventory that was worth five times its lumber value.

How graders get consistent

Not with a written rulebook, though we have one. With reference pallets and proximity.

Each grading position keeps physical reference examples of the borderline cases — the pallet that is just barely Grade A, the deck gap that is just barely too wide. Written specifications drift in interpretation; a physical object does not.

New graders work beside experienced ones for weeks, and we periodically pull a sample and re-grade it blind. If two graders disagree on more than a small fraction of a sample, the reference pallets get revisited. That is the whole quality system, and it works because the object is right there to argue about.