Tiegel home Talk to a metallurgist

Sampling system design: cross-stream cutters, increment mass and sample stations

Design of plant sampling systems: cross-stream cutter opening and speed, increment mass, number of increments, sample stations and common faults.

· 5 minute read · By

Every figure in metal accounting begins with a sample, a few kilograms intended to represent thousands of tonnes. A biased sample cannot be corrected by careful assaying, and bias usually arises from the way the sample is taken rather than in the laboratory.

A moving stream is sampled correctly by cutting across the whole stream at a constant speed, with an opening wide enough for the largest particles, so that every part of the stream, coarse and fine, top and bottom, has the same probability of entering the sample. Probes inserted into one side of a slurry line, or scoops that reach only the edge of a belt, collect whatever happens to pass that point.

This article covers the design of cross-stream samplers for belts and slurry streams, increment mass and sampling frequency, and sample stations. The final sample mass required for a given precision is covered in Gy's formula for minimum sample mass.

Cross-stream cutterOre falls from the head of a conveyor as a stream. A cutter box with a straight, parallel-sided opening moves at constant speed through the full width of the stream, then parks clear of it.head pulleyfalling streamparkedparkedwconstant speed vopening w ≥ 3 × top size, and wide enough not to blockspeed v ≤ 0.6 m/s; the cutter crosses the whole stream every time
A correctly designed cross-stream cutter takes a full slice of the stream at constant speed, with parallel lips and enough capacity that it never overflows.

Cutter design rules

w≥3dtopv≤0.6 m/smi=Q3.6×wv

where

w
cutter opening (m); also wide enough not to block
dtop
top size of the particles (the size 95% pass) (m)
v
cutter speed across the stream (m/s), kept constant
mi
mass of one increment (kg)
Q
stream flow of dry solids (t/h); for a slurry use slurry mass flow
RequirementDesign rule
Full streamThe cutter crosses the entire stream and parks clear of it on both sides
OpeningAt least three times the top size of the particles, so that the largest particles are not rejected by the lips, and wide enough not to block. For coarse ore on belts, a minimum of 30 mm is commonly quoted
SpeedConstant across the stream, and no more than about 0.6 m/s for linear cutters, so that particles are not deflected by the lips. Electric drives hold speed better than pneumatic drives
LipsStraight, parallel (or radial for rotary cutters), sharp, and replaced when worn
CapacityThe cutter body and chute hold the whole increment without overflowing or spilling back
Falling streamThe cut is made where the stream falls freely, such as at a conveyor head pulley or the end of a launder, not in a pipe or on a moving belt

Increment mass and sampling frequency

The time the cutter spends in the stream is its opening divided by its speed, and all material falling during that time enters the increment.

Worked example for a mill feed belt: 300 t/h of ore with a 25 mm top size. A 75 mm opening (three times 25 mm) at 0.5 m/s spends 0.15 s in the stream, so the increment is 300 ÷ 3.6 × 0.15 = 12.5 kg.

Msample=n×miinterval=lot durationn

where

n
number of increments taken from the lot
Msample
primary sample mass for the lot (kg)

Thirty increments over a 12-hour shift, one every 24 minutes, give a primary sample of 375 kg. This is too large to send to the laboratory, so it is reduced in a sample station. Many small increments are preferable to a few large ones, because they average out the short-term variation in the stream; the number required is determined by a variography study or from standards such as ISO 3082 for iron ore.

Worked example for tailings slurry: 480 m³/h at 1.45 t/m³. A 30 mm opening at 0.4 m/s spends 0.075 s in the stream and takes 1,450 × 480 ÷ 3,600 × 0.075 = 14.5 kg of slurry, about 7 kg of solids at a solids SG of 2.8. The opening is far larger than three times the top size of fine tailings; in this case the limits are blockage and capacity rather than particle size.

Sample stations

Sample stationThe primary cutter takes increments from the process stream; a secondary cutter splits the primary sample; a tertiary splitter or rotary divider reduces it again; rejects return to the process; the final sample goes to the laboratory.Primary cutteron the streamSecondarycutterRotary divideror tertiary cutterSample can,to laboratoryRejects pumped orconveyed back to process
Each stage of a sample station must itself be a correct sampler. A good primary cutter followed by a poor splitter still gives a biased sample.

The primary sample is reduced in stages by secondary and tertiary cutters or a rotary divider. Every stage must follow the same rules, and the mass retained at each stage must be sufficient for its top size, which is checked with Gy's formula at each step; otherwise the sample is crushed before it is split further. Rejects are returned to the process. For slurries, the final sample is often collected in a closed container to prevent spillage and evaporation, and is filtered and dried at the laboratory.

Common faults

FaultSource of bias
Cutter does not cross the full streamCoarse and fine particles segregate across the stream, so part of the stream is never sampled
In-line or pipe-wall samplers on slurryThey take what passes one point; coarse, dense particles travel lower in the pipe
Worn, bent or caked lipsThe opening changes along its length and some particles bounce off
Cutter too fast or speed variesCoarse particles are deflected; increments vary in size
Cutter overflowsThe last part of the stream crossed is under-represented
Spillage or contamination at the stationMaterial is lost or added after the cut
Manual cup samples from the side of a launderOnly the surface and edge of the stream are sampled

Samplers should be inspected on a schedule covering lips, speed, park positions, chute condition and spillage. A sampler that appears correct but is not maintained drifts into bias without detection.

Glossary

Increment
One cut taken by a sampler from a stream.
Cross-stream cutter
A sampler that moves through the whole falling stream at constant speed to take an increment.
Top size
The size that about 95% of particles pass.
Delimitation error
Bias from a cut whose shape does not give every part of the stream an equal chance of selection.
Extraction error
Bias from particles that should be in the cut bouncing off, or ones outside it falling in.
Sample station
The series of cutters and dividers that reduce a primary sample to a laboratory sample.
Rotary divider
A splitter that feeds material onto rotating containers or a rotating cutter to divide it into equal parts.
Lot
The quantity of material a sample represents, such as one shift of mill feed.

Frequently asked questions

How wide should a sample cutter opening be?

At least three times the top size of the particles, and wide enough not to block. For coarse ore on belts, a minimum of 30 mm is commonly quoted.

How fast should a cross-stream cutter move?

At a constant speed, no more than about 0.6 m/s for linear cutters. Faster cutters deflect coarse particles.

How is increment mass calculated?

Stream flow in kg/s multiplied by the time the cutter spends in the stream, which is the opening divided by the speed. For 300 t/h, 75 mm and 0.5 m/s, this is 83.3 kg/s × 0.15 s = 12.5 kg.

Can a slurry pipe be sampled through a side tapping?

Not for metal accounting. Particles are not evenly distributed across a pipe, so a side tapping takes a biased sample. The stream should be sampled where it falls freely, across its full width.

Discuss your plant with a metallurgist

Tiegel is a Perth metallurgy company. We help with testwork, plant optimisation, metal accounting and investigations, and we build the Tiegel Suite software.

Talk to a metallurgist

References and further reading

  1. Gy, P.M. (1979). Sampling of Particulate Materials: Theory and Practice. Elsevier.
  2. Pitard, F.F. (1993). Pierre Gy's Sampling Theory and Sampling Practice, 2nd edition. CRC Press.
  3. Holmes, R.J. (2004). Correct sampling and measurement: the foundation of accurate metallurgical accounting. Chemometrics and Intelligent Laboratory Systems, 74, 71–83.
  4. 911 Metallurgist. Introduction to sampling for mineral processing, part 2 (PDF). Link
  5. Madima, T. and Slabbert, W. (2024). The case for using five-times particle nominal top size cutter apertures. World Conference on Sampling and Blending 11, SAIMM. Link

Worked examples and charts marked illustrative use invented figures, not data from any operation.