How to collect a water or food sample for testing
A badly collected sample produces a valid result for the wrong thing. Use a clean container appropriate to the test, fill it correctly, label it with the exact point and time of collection, keep it cool, and get it to the laboratory inside the hold time for that parameter. Bacteriological samples must be taken in a sterile bottle — nothing else works.
Which container should I use?
For chemical analysis of water, a clean food-grade plastic bottle rinsed with the same water is fine. For bacteriological analysis you must use a sterile bottle from the laboratory — a rinsed mineral water bottle carries its own bacteria and makes the result meaningless. Oil and grease samples must go in glass. Residue samples must never go in a reused plastic container, which can leach the very compounds being looked for.
How do I take a water sample from a tap?
Remove any attached filter or aerator. Let the water run for two to three minutes so you are sampling the supply rather than the water that sat in the pipe — unless you are specifically testing for lead leaching, where the first draw is what you want. For a bacteriological sample, do not rinse the sterile bottle, do not touch the inside of the cap, fill to the shoulder leaving a small air gap, and cap it immediately.
How do I label it?
On the bottle itself, not just the bag: sample description, the exact point of collection, the date and the time. "Tank water" is not a sample point; "Block B terrace tank, outlet tap" is. If you send several samples, the labels are the only thing keeping them apart once they reach us.
How should I transport it?
Upright, in a cool box or insulated bag with ice packs, not in direct sun and not in a car boot on a hot afternoon. Chilled and frozen food samples must arrive at their storage temperature — a thawed sample is a different sample and will fail on microbiology regardless of its original quality.
How quickly must it reach the laboratory?
Bacteriological water samples within 24 hours, pool micro samples within 6 hours, effluent within 24 hours chilled. Chemical panels on water and dry food commodities are far less urgent. If you cannot meet a hold time, tell us — we will either arrange pickup or tell you the result will not be defensible.
How much should I send?
More than you think. A parameter cannot be run twice on a sample that was only just enough the first time, and a repeat on a fresh sample means a fresh collection. The table below gives the quantity per test; when in doubt send extra.
Container, quantity and hold time by test
| Test | Quantity | Container | Get it to us within |
|---|---|---|---|
| Drinking water — full panel | 2 L + 250 ml | Food-grade bottle + sterile bottle | 24 hours |
| Coliform / E. coli only | 250 ml | Sterile bottle only | 24 hours |
| Packaged drinking water | 3 sealed packs | Original unopened retail packs | 24 hours |
| Swimming pool water | 1 L + 250 ml | Sterile bottle with neutraliser | 6 hours |
| Effluent / STP | 2 L + 1 L | Plastic + glass for oil and grease | 24 hours, chilled |
| Construction water | 2 L | Clean plastic container | A few days |
| Packaged food — micro | 3 retail units | Original sealed packaging | At storage temperature |
| Nutritional label panel | 3 units, 250 g total | Original sealed packaging | Not urgent if shelf stable |
| Edible oil | 500 ml | Glass, filled to the top | Away from light and heat |
| Milk | 500 ml | Clean bottle, filled nearly full | Few hours at 4 °C |
| Spices | 200 g | Sealed dry pouch | Not urgent, keep dry |
| Metals / pesticide residues | 500 g or 1 L | Glass or new food-grade pouch | Fresh produce promptly |
The five mistakes we see most often
- A bacteriological sample collected in a rinsed mineral water bottle. The result is invalid no matter what it says.
- A pool micro sample without a chlorine neutraliser — the chlorine keeps working in transit and a dirty pool passes on paper.
- One sample for a problem with many possible entry points. Reservoir, tank and tap is three samples, and only three samples locate the fault.
- Oil and grease decanted from a larger bottle. It sticks to the first container, so the result reads low.
- A chilled food sample sent at ambient temperature and arriving warm. It will fail on microbiology, and the failure tells you about the courier, not the product.