3.5 Jar Testing Procedure & Chemical Feed Equipment
Key Takeaways
- JAR TESTING and CHEMICAL FEEDERS are separately scored objectives on the Missouri Drinking Water A, B and C exams.
- A jar test replicates rapid mix, flocculation and settling at bench scale so the optimum coagulant dose and pH can be found before committing plant chemicals.
- Positive displacement metering pumps deliver a fixed volume per stroke and are calibrated by drawdown from a graduated cylinder.
- Dry chemical feeders are volumetric (measuring by volume) or gravimetric (measuring by weight); gravimetric is more accurate because it compensates for density changes.
- Feeder calibration must be verified against actual delivered quantity, never assumed from the dial setting.
The official Missouri test descriptions score JAR TESTING on Drinking Water A (3 calculation items), B (2 plus 2 calculation) and C (2 plus 1 calculation), and CHEMICAL FEEDERS on Drinking Water A, B and C. These are practical, hands-on objectives.
The Jar Test
A jar test is a bench-scale simulation of the entire coagulation-flocculation-sedimentation train. Its purpose is to answer three questions before the plant commits chemicals: which coagulant, what dose, and at what pH.
Standard procedure
- Fill six 1,000 mL or 2,000 mL square gang-stirrer jars with well-mixed raw water. Square jars are preferred because they induce baffling that better mimics basin turbulence.
- Record baseline raw water turbidity, pH, alkalinity and temperature. Temperature matters: cold water flocculates more slowly.
- Dose each jar with a different coagulant volume, spanning above and below the current plant dose. Add all doses simultaneously using pipettes positioned in advance so every jar starts at the same moment.
- Rapid mix at 100 to 300 rpm for 1 minute, simulating the flash mixer.
- Flocculate at 30 to 40 rpm for 15 to 20 minutes, tapering the speed down as a real tapered flocculation train does.
- Settle quiescently for 30 minutes with the paddles removed. Record the time to first visible floc, floc size and settling character.
- Sample from a fixed depth below the surface using a pipette, and measure settled turbidity. This is the number that decides the test.
Reading the result
The optimum dose is the lowest dose producing acceptable settled turbidity — not the dose producing the largest floc. Big fluffy floc that shears in the flocculator or floats over the weir is worse than smaller, denser floc.
| Observation | Interpretation |
|---|---|
| Slow floc formation, cloudy supernatant, high settled turbidity | Underdosed. Insufficient charge neutralisation |
| Excellent floc at low dose, deteriorating at high dose | Charge reversal. Excess coagulant re-stabilises the particles |
| Good floc, but supernatant still turbid | Check pH — likely outside the coagulant's effective range |
| Floc forms then breaks up | Excessive mixing energy; reduce flocculation rpm |
| Every jar poor | Consider a coagulant aid (polymer) or a different coagulant |
Dosing arithmetic. If a 1% alum stock solution (10,000 mg/L) is used and 2.0 mL is added to a 2,000 mL jar:
A convenient shortcut for a 1% stock in a 1,000 mL jar: 1 mL = 10 mg/L.
Scaling to the plant. Once the jar test picks a dose, the pounds formula converts it to a feed rate. At 12 mg/L on a 3.5 MGD plant:
Remember alum's alkalinity consumption of roughly 0.45 mg/L as CaCO₃ per mg/L of alum: 12 mg/L of alum consumes about 5.4 mg/L of alkalinity, which soft Missouri surface water may not have to spare.
Chemical Feed Equipment
Liquid feeders
Positive displacement metering pumps are the workhorse. Each stroke displaces a fixed volume, so output is set by stroke length (how far the diaphragm or piston travels) and stroke frequency (strokes per minute). Output is proportional to both, which makes them linear and easy to pace to flow.
- Diaphragm pumps — a flexible diaphragm isolates the chemical from the drive. Standard for hypochlorite, fluoride, polymer and corrosion inhibitor.
- Piston/plunger pumps — higher pressure capability, but the chemical contacts the plunger, so they are limited to compatible chemicals.
- Peristaltic (hose) pumps — rollers squeeze a flexible tube. Nothing but the tube touches the chemical, making them excellent for slurries and aggressive chemicals; the tube is the wear item.
Calibration by drawdown is the required verification method and a standard exam calculation. Draw suction from a graduated cylinder instead of the day tank, run the pump for a measured time, and record the volume removed.
Worked example. A metering pump draws 480 mL from a graduated cylinder in 5.0 minutes while feeding 12.5% sodium hypochlorite.
At 12.5% trade strength, hypochlorite carries about 1.04 lb of available chlorine per gallon, so this delivers roughly $36.5 \times 1.04 = 38$ lb/day of chlorine.
The dial setting is not the answer. Metering pump output drifts with discharge pressure, diaphragm wear, check valve fouling and chemical viscosity. Only the drawdown test proves what is actually being delivered.
Dry feeders
| Type | Principle | Accuracy |
|---|---|---|
| Volumetric | Measures a fixed volume per revolution — screw, rotating disc, belt or vibrating trough | ±2 to 5%; degrades when material density or moisture changes |
| Gravimetric | Measures by weight, usually a belt feeder on load cells or a loss-in-weight hopper | ±1%; self-correcting for density changes |
Gravimetric feeders are more accurate precisely because they weigh the chemical. If lime arrives at a different bulk density, a volumetric feeder delivers the same volume but a different mass, and the plant dose silently shifts.
Common dry feeder problems: bridging (material arching over the outlet, stopping flow), flooding (fine powder flowing uncontrollably like a liquid), and caking from moisture. Bin vibrators, air pads and dry storage address these.
Solution tanks and safety
- Provide two solution tanks or a day tank so feed continues while one is being mixed.
- Calibrate the tank with a sight glass or gauge so drawdown over a shift verifies consumption.
- Fluoride, hypochlorite and acid feed lines must be clearly labelled and physically distinct — cross-connecting an acid line to a hypochlorite tank releases chlorine gas.
- All feed points into a pressurised main need an anti-siphon device or backpressure valve so plant water cannot siphon back into the chemical tank.
During a jar test, the 8 mg/L jar produces good floc and 1.2 NTU settled turbidity, the 12 mg/L jar produces the largest visible floc and 1.1 NTU, and the 16 mg/L jar produces poor floc and 4.8 NTU. What dose should the operator select?
Why is a gravimetric dry chemical feeder considered more accurate than a volumetric feeder?
An operator performs a drawdown calibration on a hypochlorite metering pump and finds it removes 300 mL from the graduated cylinder in 4.0 minutes. What is the delivery rate in mL/min, and what does this test establish that the pump dial cannot?