Your Sample Came From the Top 300 mm: Stratified Lagoon Sampling Error

The lagoon is four metres deep, holds 9,000 cubic metres, and has not been mixed since the aerators were removed in 2019. The sample that went to the laboratory came off the top, through a hatch, in a clean bottle, at ten in the morning. It read 1,800 mg/L COD and 240 mg/L ammonia. The dose was set from that. The first batch pulled from the bottom drain came out at 6,400 mg/L COD and 890 mg/L ammonia, and set like putty in the line.

Stratification is the quiet error in leachate work. It does not announce itself, it produces plausible numbers, and it biases every decision downstream: dose, cost per cubic metre, and the acceptance test result at the gate.

Why a still lagoon separates

Three gradients form at once. Temperature: surface water warms in summer and sits on cooler, denser water below, and the interface can be sharp enough to feel with a hand. Salinity: dissolved solids concentrate in the lower layer, especially in lagoons that receive evaporator blowdown or RO concentrate. Solids: sludge accumulates at the bottom and releases ammonia and soluble organics back into the pore water.

The result is a stratified column where the top metre can be half the strength of the bottom metre. In one lagoon we sampled at five depths, conductivity ranged from 11 to 34 mS/cm and COD from 1,500 to 7,800 mg/L. Density differences of less than one percent are enough to keep that stable for months in the absence of mixing.

Depth-weighted sampling, without a pump

The correct sample is a depth-integrated one: equal volumes from at least five depths, combined in a single container, thoroughly mixed, and subsampled. A simple weighted bottle on a marked rope does this without a peristaltic pump. Lower it slowly to each depth, let it fill, haul it up, pour a fixed volume into the composite.

Two details make the difference between a representative sample and a random one. Lower and raise slowly, because a bottle dropping fast grabs water from everywhere it passes. And take the bottom sample at least 300 mm above the sludge blanket, measured rather than assumed, so the sample represents the liquid and not the slurry.

Where the lagoon is being drawn down, take the composite at the start of the draw and again at a quarter remaining. The strength rises as the level falls, because the bottom layer becomes a larger share of what is left. A dose set on day one is often 25 to 40 percent low by day four.

The sampling point is part of the process, not an afterthought

Plants that get this right stop sampling from the nearest hatch and install a proper point: a vertical stilling tube with depth markings, or a tap on a recirculation line that runs long enough to purge the line volume. The second option is better where a mixer exists, because a mixed sample removes the whole question. A twenty-minute recirculation before sampling costs a pump run and buys a sample that means something.

There is a counterintuitive consequence worth flagging. Mixing a stratified lagoon before treatment raises the measured strength of every subsequent sample, which looks like the leachate got worse. It did not. The earlier numbers were simply too low. Budgets built on top-of-lagoon samples routinely under-provide by a third, and the shortfall shows up as an overrun in the third quarter.

Documenting it so a regulator can follow

Write the depth on the sample label, not just the date and time. A chain of custody that says “lagoon 2, composite 0.3/1.2/2.1/3.0/3.7 m, 250 mL each” is auditable. One that says “lagoon 2” is not, and the difference matters when a disposal facility challenges a waste characterisation or an inspector asks how the dose was derived.

Photograph the sampling point with the tape measure visible. It takes ten seconds and settles arguments months later about whether the bottom sample came from 3.7 metres or from the sludge.

Before the next batch, add the sample depth and the number of subsamples to the batch header. If the answer is one depth and one subsample, the dose on that sheet is a guess dressed as a measurement. Where the batch is fed from a mixed tank rather than a lagoon, stratification mostly disappears, and the grade selection guide becomes the more useful reference than the sampling protocol.