Coagulation and flocculation: dosage and the jar test

Two jobs behind one phrase

Coagulation and flocculation are spoken of together because they happen in the same stretch of a plant, one after the other, but they are two different jobs. Colloidal particles in raw water — clay, colour, bacteria, organic matter — carry a negative surface charge, and like charges keep them apart. Plain settling does not remove them at a useful rate, because the repulsion between two particles is stronger than the force gravity puts on particles that small.

Coagulation is the step that removes the repulsion. A coagulant dosed into the water produces positively charged species that compress the electrical double layer around each particle and neutralise part of its charge, so two particles that meet can stay together. Flocculation is the step that grows those first small aggregates into flocs large enough to settle or to be floated: long-chain polymer molecules adsorb on more than one particle at a time and bridge them, and where a metal salt is dosed at a higher rate, a hydroxide precipitate forms and enmeshes the particles in what is usually called a sweep floc.

The distinction has a practical consequence in the mixing. Coagulation wants energy in the first seconds so that the chemical is dispersed through the volume before it has reacted; flocculation wants gentle motion over minutes, because a floc that has already been built is easily broken by shear. A plant that mixes once, hard, does one of the two jobs and undoes the other. The separation step that follows — sedimentation, dissolved air flotation or filtration — is where the aggregates have to come out, and its surface loading is what turns a well-formed floc into a clear supernatant.

The three families of chemicals

Almost every dosing arrangement draws on three families, and most use more than one of them.

• Hydrolysing metal salts — aluminium sulphate (alum), polyaluminium chloride, ferric chloride and ferrous sulphate. They hydrolyse in water to positively charged species that neutralise particle charge, and at higher doses they precipitate as a hydroxide that sweeps particles down with it.
• Organic polymers — polyacrylamide (PAM) flocculants, supplied as anionic, cationic or non-ionic grades. The long chain bridges particles that are already destabilised, and the cationic grades carry a positive charge that does part of the neutralising themselves.
• pH and alkalinity adjustment — lime, caustic soda or an acid, used to move the water into the band where the chosen coagulant works. Alum and ferric salts consume alkalinity as they react, so the pH after dosing is the one the process sees.

The reason more than one chemical appears is that the two duties are best served by different molecules. A metal salt neutralises charge within a pH band that has to be established for the water in hand; a polymer builds a floc with a chain length a metal salt does not have. A single metal salt on a difficult water, or a polymer alone on a stable colloid, is how a plant ends up pushing a dose far higher than it needs to be. Coagulant suppliers publish indicative bands drawn from waterworks practice, but those bands describe the chemical rather than your water. The value that matters is determined on the water the plant will actually treat.

Why the dose cannot be read from a table

The optimum dose of a given coagulant moves with the water. Turbidity and suspended solids, alkalinity and pH, temperature, the organic fraction, colour and ionic strength all shift it, and they do not all shift it in the same direction. The same alum at the same dose can produce a fast-settling floc in one raw water and a fine suspension that carries straight over the weir in another. That is why two plants of similar size on two different rivers can hold doses far apart.

Two variables sit on top of the water itself. Cold water slows hydrolysis and floc formation, so a dose that works in summer can under-perform in winter and a longer flocculation time may be needed to compensate. And a metal salt dose set by a phosphorus limit rather than by clarity is set by precipitation chemistry: it follows the ratio of metal to phosphorus in the water, which is a different calculation from the one that decides what looks clear in a beaker.

The conclusion is that the dose is determined rather than looked up, and the determination is the jar test — a bench-scale copy of the plant, run in six beakers on the actual water, at doses chosen around the expected band.

The jar test, step by step

The jar test is standard practice rather than a single fixed standard, and laboratories differ in the details. The jar test procedure below is the conventional bench method and the figures are the ones usually quoted for it, which is why they should be read as a starting method rather than as a specification. What matters more than the exact values is that the same procedure is repeated every time, so that two results can be compared.

• Fill six identical beakers with the same raw water, at the same temperature, to the same volume — about one litre per beaker is the usual bench size.
• Measure the raw water first: turbidity or suspended solids, pH, alkalinity, and COD or total phosphorus where those are the parameters the plant is built to remove.
• Set a dosing ladder for the coagulant around the expected band, leaving one beaker undosed as the control. Wide steps when the band is unknown, then narrow steps once the answer is in sight.
• Rapid mix: stir the beakers hard for a short period — of the order of one minute — to disperse the coagulant before it reacts.
• Slow mix: drop the speed and stir gently for the flocculation period, of the order of 15 to 20 minutes, so the flocs can grow without being broken apart.
• Stop the stirrers and let the beakers stand for the settling period, of the order of 20 to 30 minutes.
• Read the supernatant rather than the whole beaker. Sample at the same depth in every beaker, then measure turbidity and the target parameter in the sample.
• Repeat with a narrower ladder around the best two beakers, and repeat the whole test whenever the raw water changes materially.

The speeds and times above belong to the bench, not to the plant. On site, coagulation and flocculation are designed around the energy dissipated per unit volume and the time the water spends in each stage, and the bench result is what fixes the chemical demand that design has to meet. The tank volume and HRT chart shows how a residence time becomes a basin volume once the dose is known.

Record more than the winning dose. Useful entries are the dose, the pH measured after dosing, the mixing speeds and durations, the settling time, the supernatant result, the size and appearance of the floc, and the volume of sludge that settled out — that last figure is the duty the dewatering stage will be handed.

Common mistakes

• Reading the best beaker as the dose to buy. The clearest supernatant in a ladder often sits above the point of diminishing return, and chemical cost and sludge volume both rise with dose. The dose to carry forward is the lowest one that meets the target with a margin.
• Overdosing past the optimum. Excess metal salt or excess cationic polymer reverses the particle charge and re-stabilises what had already been destabilised, so turbidity climbs again as the dose rises. A ladder that turns worse at the top is demonstrating this, not suffering an experimental error.
• Judging the test on clarity alone where a phosphorus limit applies. Chemical phosphorus removal follows the metal-to-phosphorus ratio in the water, and the visual clarity of the supernatant is a poor guide to it.
• Preparing polymer too concentrated. Powder flocculant wetted into a strong solution forms lumps that never fully dissolve, and the undissolved fraction is both wasted chemical and a source of dose error. Working solutions are made up weak and given time to mature before dosing.
• Cutting the rapid mix short, or running it too gently. The coagulant has to reach the whole volume before it reacts; a slow first stage leaves part of the dose reacting locally and leaves the flocculation stage with little to build on.
• Testing water the plant will not see. A sample left standing, warmed in transit or drawn in a single quiet hour does not represent the load the plant has to treat at the peak, and a dose set on it will be wrong when it matters.
• Measuring pH only before dosing. Metal salts consume alkalinity, and the pH that governs the reaction is the one after the chemical has been added.

Self-check before the dose is fixed

• Is the sample the water the plant will treat, taken at the point and the time that decide the duty?
• Is there an undosed control in the set, so the effect of the chemical is separated from what the water does on its own?
• Was pH measured after dosing as well as before it?
• Is the dose quoted for the commercial product as supplied, with its active content, so that the test and the purchase match?
• Has the dose been checked at the peak load as well as at the average?
• Is the polymer working solution dilute enough for the dosing pump, and given enough maturation time?
• Has the settled sludge volume been recorded, so the downstream dewatering duty is known?
• Has the winning dose been narrowed with a second ladder, or confirmed on a second sample?

From a dose to a dosing system

A jar test produces a number; a plant has to repeat that number every day, at whatever flow and load arrive. Powder polymer needs wetting, maturation and metering, and the equipment that does that work is described on the chemical and advanced treatment page, where the automatic polymer dosing machine prepares and doses the flocculant on site instead of by hand. Two details have to travel with the dose: the commercial product and its active content, so that a laboratory dose and a purchased chemical are the same thing, and the dilution the dosing pump can handle. The metering rate that suits your polymer grade and dewatering duty is confirmed with the factory for your polymer and your sludge duty rather than assumed from a table. The design influent quality chart and the effluent standards chart set the duty the dosing system has to meet, and the sludge dewatering equipment pages cover what happens to the sludge the dose produces.