Overdosing by 15 Percent: The Annual Cost of a Cautious Operator

A site designed for 22 kg/m³ ran at an average of 25.3 through the year. Nobody decided to do that. The dose crept: a wet month here, a rejected load there, a new operator who added half a bag to be safe, a scoop never checked against a scale. Across 12,000 m³ the excess came to 39.6 tonnes of polymer, about $83,000 at $2,100 a tonne, plus the extra cake volume and the gate fee. The rejection rate that year was zero, which is why nobody questioned the dose.

Overdosing is the most expensive habit in leachate solidification because it is invisible, well-intentioned and defended by the absence of failures. This article looks at why it happens, what it does to the cake rather than just the invoice, and how to replace a habit with a control chart.

Five Ways a Dose Creeps Up

Rain is the first and the most legitimate. A 50 mm event can drop conductivity and raise volume at the same time, and the operator who adds 10 percent is making a reasonable call. The problem is that the 10 percent is rarely taken back out when the weather dries.

Fear of rejection is the second. A single rejected load costs more in hassle than a month of extra polymer, so the operator buys insurance at the bag. That logic is sound for one load and ruinous across a year, because it is applied to every load.

The third is measurement drift. A scoop of polymer is not a fixed mass: the same scoop holds 620 g when the powder is free-flowing and 700 g after a humid week. Over a shift that is 12 percent, and over a year it is the whole overrun.

The fourth is batch volume error. A dipstick read against a nominal tank diameter, with a tank that has bowed or has 80 mm of silt, reads 3 to 5 percent high, and the operator compensates by adding polymer rather than correcting the volume.

The fifth is the absence of feedback. If nobody measures the actual dose against the design dose, the number that runs the plant is the one in the procedure, not the one in the tank.

What Overdosing Does to the Cake

The invoice is the least of it. Above the point where the network is fully formed, extra polymer competes for the remaining water, hydrates partly, and forms stiff lumps with dry cores. Free water sits between the lumps, the mass feels firm in the hand, and it releases that water under fifteen minutes of gravity in a paint filter.

Sites that overdosed their way through a winter often report the paradox of loads that feel dry and fail the test. The gel also becomes brittle, which matters at discharge: a 28 kg/m³ gel shears into fragments at the pump and arrives in the skip as a coarse, poorly consolidated mass with visible voids. And the volume penalty grows: an extra 3 kg/m³ adds roughly 4 percent to cake volume, which is another gate fee on top of the polymer bill.

Set time also extends. A batch that would have been ready in six minutes takes nine, and the discharge schedule slips by twenty minutes a day, which is a real cost on a site running two batches per shift.

A Control Chart in Place of a Habit

Three practices hold a dose within 5 percent. Weigh the dose, do not scoop it: a 20 kg platform scale at the bag dump costs less than a tonne of polymer and removes the largest single source of drift. Measure batch volume with a calibrated dipstick or a flowmeter, and re-calibrate after any tank work.

Run a jar test monthly and after any event above 20 mm of rain, and plot the result against the dose actually used. The gap between the two lines is the insurance premium, and seeing it plotted is more persuasive than any instruction. Set a tolerance band of plus or minus 5 percent on daily dose and require a written reason for anything outside it, not an approval, just a reason, because the act of writing it down removes most of the casual additions.

Finally, track cost per cubic metre and rejection rate side by side. A site that has cut dose successfully and seen no change in rejections has found money. A site that has cut dose and seen rejections rise has found the floor, and that is a useful number too.

The Insurance Premium, Priced Properly

There is a case for running 5 percent above the jar-test knee, and it costs about $28,000 a year at 12,000 m³. There is no case for running 15 percent above it, which costs $83,000 and buys nothing except a slower batch and a bulkier cake.

The way to decide is to price the two risks honestly: a rejected load at $900 to $2,400 including return haul and quarantine, against the annual cost of the margin. At 15 percent you are paying for roughly 35 rejections a year that never happen. The SAP grades we supply for leachate work come with a dose-response curve measured in leachate rather than water, which is the reference a site needs if it wants to argue the dose down without arguing with the operator.