Sampling a tanker load of granules: a probe pattern that stands up

Two labs, one tanker, and twelve percent between them. The buyer’s sample came back at 268 g/g, the supplier’s retained sample at 302. Both labs were competent, both used the same standard, and the argument ran for three weeks before anyone asked where in the tanker the buyer’s sample had come from. It came from the top hatch, six inches below the surface, taken with a cup on a stick.

Powder in a tanker segregates. Vibration during the journey moves fines downward and coarse granules upward, and a sample taken from one point tells you about that point, not about the load. A defensible sample needs a pattern, and the pattern needs to be written down before the tanker arrives.

How much segregation is normal

In a pneumatic tanker of 20 to 25 tonnes, it is entirely normal to find the top layer 3 to 6 percent coarser than the bottom, and a fines concentration at the base that is two to three times the average. On an absorption test that difference can translate into 10 to 20 g/g of measured capacity, purely from particle size, because finer granules swell faster and reach a higher figure in a fixed ten-minute test.

That is not a quality defect. It is a sampling problem, and the fix is to sample the whole depth at several positions and composite them in the right proportions.

A probe pattern that holds up

Use a proper sample thief — a slotted tube that takes material along its length, not a cup. Take five positions: four quadrants about a metre in from the tank wall and one at the centre. At each position, take three depths: 300 mm below the surface, mid-depth, and 300 mm above the base. That is fifteen increments.

Composite them in equal parts into one sample of about 2 kg, quarter it down to 500 g, and split that into three: one for the buyer’s lab, one for the supplier’s retained sample, one sealed and held for twelve months. Label each with the tanker number, the lot, the date, the positions used and the name of the sampler. A sample without that label is worth nothing in a dispute.

If the tanker has no top hatch access at the base, say so in the record rather than pretending the pattern was complete. An honest note about the constraint is more persuasive than a clean-looking sheet that a driver can contradict.

A composite also hides things, and that is the point. Some buyers ask for a “worst case” sample from the bottom, on the theory that the fines are the problem. That is useful for a flow trial but wrong for an acceptance test. The composite represents what the site will actually consume over the discharge, and that is what the specification should be applied to.

Keep the bottom-of-tanker increment as a separate labelled bag if you are investigating a feeding problem. It is a diagnostic sample, not an acceptance sample, and conflating the two is how a good load gets rejected.

Timing and the practical constraints

Sample before discharge, not after. Once the tanker is blowing into a silo the load is mixed with whatever was there before, and you are sampling the silo rather than the delivery. If the site has no top access, arrange for the sample to be taken at the receiving silo’s inlet during the first ten minutes of transfer, and note that in the record.

Take the sample before the driver starts his paperwork, not after. Ten minutes of probing at the top hatch is normal and a driver who has done it before will expect it. A site that samples after the hose is connected will always be sampling a moving target.

What to do when the two labs disagree

First, compare the methods, not the results. Absorption tests are sensitive to water temperature, water hardness, sieve fraction used, and soak time. A 5 degree difference in water temperature can move the number by 5 percent. Second, compare the samples — if the buyer’s sample was a top sample and the supplier’s was a composite, the disagreement is explained before any chemistry is discussed.

Third, split a retained sample and send it blind to a third lab. That costs a few hundred and usually ends the argument in a week. In the case above, the third lab reported 279 on the buyer’s sample and 285 on the retained, which told everyone the material was fine and the sampling was not.

Write the pattern into the purchase contract, including who samples, when, and where the retained portion is kept. If you are setting that up for the first time, the incoming inspection notes for super absorbent powder include a sampling sheet you can attach to the contract.