Six Jar Tests to a Full-Scale Setpoint: Calibration Arithmetic and Its Traps

The bench at a 300 m³/day leachate plant holds six 500 mL beakers and a rack of doses: 8, 12, 16, 20, 24 and 28 kg/m³. The run is 400 rpm for 60 seconds, then 60 rpm for 120 seconds, graded at five and thirty minutes. The curve knees at 18 kg/m³. Production runs at 21, the knee plus a 15 percent margin: 4.2 tonnes a day at 200 m³/day. Every number in that sequence came from six beakers, and every one can be wrong.

Jar testing is the cheapest control a leachate plant has and the easiest to do badly. This article covers how to bracket the knee rather than guess at it, how to grade a gel without a rheometer, how the arithmetic scales from 500 mL to 12 m³, and the four traps that make a technically correct test produce a wrong setpoint.

Bracketing the Knee Instead of Guessing at It

Most jar tests use a geometric series: 4, 6, 9, 13, 20, 30 kg/m³. That spacing finds the knee within one step across a wide range, which is what you want on an unfamiliar leachate. A linear series of 8 to 28 in steps of 4 is better once you know roughly where you sit, because it resolves the region that matters. Run the geometric series first, then a linear series around the apparent knee, and you will have a setpoint in two hours of bench time instead of a week of plant trial and error.

Two details matter more than the series. Jars must be the same diameter, because the shear profile at 60 rpm depends on diameter and a mismatched beaker will read one step high. And the leachate must be at sump temperature, not room temperature, or the whole curve shifts left in winter and you will set a dose that fails in February.

Grading a Gel Without a Rheometer

Three bench tests cover what production needs to know. The inversion test: cap the jar, invert it, and hold for ten seconds. A gel that does not move scores three, one that creeps scores two, one that flows scores one. The rod test: draw a 12 mm rod through the mass at 100 mm per second and measure the length of the thread before it breaks, with 100 mm or more indicating a firm set. The load test: place a 200 g weight on a 40 mm disc laid on the surface and measure penetration after 30 seconds, with less than 8 mm indicating a gel that will survive a skip.

Grade at five minutes for set speed and at thirty minutes for final strength, and record both. A batch that reaches strength at five minutes but drops by thirty is showing syneresis, and it will bleed in the skip no matter how good the early number looks.

Scaling the Arithmetic From 500 mL to 12 m³

Dose in kg/m³ scales directly with volume, so 21 kg/m³ in a 500 mL jar is 0.0105 kg, and in a 12 m³ batch it is 252 kg. That part is simple. What does not scale is shear. A magnetic stirrer at 60 rpm in a 70 mm beaker delivers a very different energy per unit mass than a 1,400 rpm paddle in a 12 m³ tank, and the tank generally delivers more, which produces a firmer gel at the same dose and encourages operators to cut back. The 10 to 15 percent margin over the knee exists to absorb that difference, not to cover analytical error.

Volume also drifts. Leachate density at 1.5 percent solids sits near 1.01 to 1.02 t/m³, so a batch measured by dipstick can be 2 percent light, and at 252 kg per batch that is 5 kg a day of quiet under-dosing. Weigh the batch or use a calibrated flowmeter, and record the actual volume rather than the nominal one.

The Four Traps That Produce a Wrong Setpoint

Temperature is the first: a 22 °C bench against an 8 °C sump will understate dose by 8 to 12 percent. Ageing is the second: leachate held in a sealed jar for three days continues to ferment, its pH drops and its ionic strength changes, so a stale sample reads easier than the real thing. Operator bias is the third, and it is real, because the person grading knows which jar holds which dose unless the samples are coded. Rain is the fourth: a 50 mm event can change conductivity by 20 percent and reset the whole curve.

None of these traps requires new equipment to close. Code the jars, test within four hours of sampling, hold the samples at sump temperature in a water bath, and re-run the curve after any rainfall event above 20 mm or any change in which cell is contributing to the sump. Do that, and six beakers will hold a setpoint through a season. If you want the bench work to translate cleanly to the plant, the SAP grades we supply for leachate work come with absorption curves run at both 20 °C and 5 °C, so the winter step in your jar data can be checked against the material rather than guessed at.