A Conductivity Probe as a Dose Feedback Loop

Eleven mS/cm, and the dose was 2.4 percent. Two hours later the same probe read 26 mS/cm and the batch was still wet. The operator had been dosing by a table that mapped conductivity to dose, and the table had been built in April. In August, after three months of dry weather, the same conductivity meant a different leachate.

Conductivity is the cheapest inline signal available on a leachate plant, and it is badly misused. It correlates with dissolved solids, which is genuinely useful. It does not measure the thing that drives polymer demand, which is the balance between ionic strength and organic load.

What the probe sees and what it misses

Conductivity responds to mobile ions: sodium, chloride, potassium, ammonium, bicarbonate, sulphate. It is blind to neutral organics, and landfill leachate carries a large neutral organic fraction, especially in young cells where volatile fatty acids dominate. Two leachates at 15 mS/cm can differ by a factor of two in COD depending on waste age, and the younger one usually needs more polymer per cubic metre because the organic load competes for water differently.

Temperature compensation is the second blind spot. A probe compensating to 25 degrees will read a batch at 8 degrees as roughly 20 percent more conductive than it is, which in a winter batch looks like stronger leachate. Sites that do not correct for this over-dose through the cold months and blame the powder.

Using it as a trend, not a table

The reliable use of a conductivity probe is not absolute dose, it is relative change. Set a baseline when a batch is known good: record conductivity and the dose that passed. Then treat any deviation of more than 25 percent from baseline as a trigger to sample and jar test, not as a trigger to change the dose by a formula.

That single change removes most of the damage. A table invites automatic adjustment with no verification. A trend alarm invites a person to look.

Second good use: detecting events. A jump of 5 mS/cm in an hour almost always means an unmixed inflow slug, a tanker discharge, or a condensate line running. A slow rise over a week means evaporation concentration. The probe is a good event detector and a poor meter.

Where to install it so it reads the batch, not the pipe

Install in a flowing line after the batch is mixed, ideally on the recirculation loop, with at least ten pipe diameters downstream of any injection point. A probe in a dead leg reads whatever diffused into it last. A probe in the batch tank floor reads the densest layer.

Clean it on a schedule, not on failure. Leachate scales probes with carbonate and organic film within weeks, and a fouled probe drifts downward, which reads as weaker leachate and quietly raises the dose. A monthly two-point check with a standard solution takes ten minutes and is the difference between a useful signal and decoration.

The failure mode that costs the most

Automated dose control from a single conductivity signal is the expensive mistake. A plant in Jiangsu wired a controller to add polymer proportional to conductivity above a threshold. During a rain event the inflow diluted, conductivity fell, the controller cut the dose, and 40 m3 of weak but high-volume leachate went out under-dosed. The receiving facility rejected four loads and the investigation took a week, because the control logic had no record of what it decided.

If a site wants feedback control, the signal should be a proxy for the outcome, not the input. A torque or power draw on the mixer, a level rise in a settling column, or an inline viscosity reading all respond to whether the batch is actually setting. Those are harder to install and far harder to fool.

For now, the practical package is: probe in the recirculation line, temperature compensation checked, monthly cleaning and calibration, baseline plus 25 percent trigger, and a jar test before any setpoint change. That is a useful instrument. The dose-response data for each grade is what turns a jar test into a setpoint quickly, which is the step that has to stay fast.