Chemical Injection and Inline Mixing: Nozzle and Eductor Selection for ppm Accuracy

An 800 m³ service reservoir was dosed at 1.5 ppm free chlorine instead of the 0.8 ppm setpoint. Roughly double the target reached the taps. The plant was metering 3.1 L/h of 12.5 percent sodium hypochlorite. The setpoint called for 1.6 L/h. A corroded injection nozzle had widened its orifice by about 0.2 mm.

The opposite error is quieter and costs more. A 50 m³/h cooling circuit believed it was dosing 2 ppm of scale inhibitor. It was actually running at 1.4 ppm. The injection quill stopped 3 diameters upstream of the first elbow. The stock arrived at the heat exchanger as a concentrated ribbon along one pipe wall. Carbonate scale closed the tubes inside one season.

This guide covers chemical injection nozzles, venturi injectors, static mixers and eductors for inline dosing and chemical mixing. It works the ppm arithmetic and compares the three ways to mix. It then sets out injection point geometry, suction ratio, turndown, materials and the tests that prove a mix is real.

The Snapshot

• 1.0 L/h of 10 percent stock is what a 50 m³/h line at 2 ppm needs, from dose rate = (water flow in m³/h × ppm) divided by (stock percent × 10).
• 10 pipe diameters is the minimum straight run from the injection quill to the first mixer or bend, which on a 100 mm main is 1.0 m.
• 1:1 to 4:1 is the typical entrainment band of a motive-driven eductor, quoted as litres of motive water per litre of stock drawn.
• 5.6 seconds is the residence time in a 10 m length of 100 mm pipe at 50 m³/h, the entire mixing window when no static mixer is fitted.

The Dose Arithmetic: From ppm to Litres per Hour

One ppm is one milligram of active ingredient per litre of water. That identity carries the whole calculation.

Take the treated water flow in litres per hour and multiply by the target in ppm. The product is the active mass in milligrams per hour. Divide that by the stock strength in milligrams per litre. The answer is the stock flow in litres per hour.

Work the 50 m³/h example by hand. The water flow is 50,000 L/h. At 2 ppm the active demand is 50,000 × 2, which is 100,000 mg/h. That is 100 g/h. A 10 percent stock at a density of 1.0 kg/L holds 100 g of active per litre. The stock flow is 100 g/h divided by 100 g/L, which is 1.0 L/h.

Collect the constants and the same sum runs in one line. Dose rate in L/h equals (water flow in m³/h × ppm) divided by (stock percent × 10). The factor of 10 folds the 1,000 litres per cubic metre. It also folds the 10,000 mg/L a stock of percent strength carries at unit density.

Water flow	Target dose	Stock strength	Stock flow
50 m³/h	2 ppm	10 %	1.0 L/h
50 m³/h	2 ppm	1 %	10.0 L/h
200 m³/h	1.5 ppm	12.5 %	2.4 L/h
10 m³/h	0.5 ppm	5 %	0.10 L/h
500 m³/h	0.3 ppm	30 %	0.50 L/h

Read the first two rows together. The same duty needs ten times the pumped volume when the stock is diluted tenfold. Concentrating the stock is the cheapest turndown available. It moves the duty away from the bottom of the pump curve.

One correction decides whether the figure is real. A 10 percent sodium hypochlorite solution has a density near 1.16 kg/L. The shortcut understates its active strength by about 16 percent. Use the datasheet density for brines, strong acids and caustic above 5 percent. Then check the basis of the ppm target. A 12.5 percent hypochlorite quoted as available chlorine differs from one quoted as product. The gap is close to a tenth.

Three Ways to Mix, Ranked by Contact Efficiency

Three devices put stock into a pipe: a venturi injector, a static mixer and an injection quill or nozzle. They are not interchangeable in every duty. A venturi can draw and mix. A static mixer can only mix. A quill can only inject.

Method	What it does	Length and mixing path	Pressure cost	Typical accuracy
Venturi injector (eductor)	Draws stock by vacuum and mixes it in the throat	Throat plus 10 to 20 D downstream	30 to 50 % of motive pressure lost	Within about 5 % of the set ratio
Static mixer	Blends an already-injected stream	10 to 20 D total, element by element	0.05 to 0.15 bar at 1.8 m/s	Within 2 to 5 % with a centred quill
Injection quill or nozzle	Delivers stock into the pipe core	Needs 20 to 40 D downstream to blend alone	0.05 to 0.30 bar at the pump discharge	Set by the pump, within 2 % at steady flow

Most dosing lines need two of the three. A venturi is the only device that can pull stock from a tank without a pump. A static mixer guarantees a blend over a short distance. A quill injects against a high main pressure with a pump behind it.

When a Venturi Injector Beats a Metering Pump

An eductor needs no electricity at the dosing point. A motive stream at 2 to 4 bar accelerates through a converging nozzle. That creates a low-pressure throat, which draws stock from a day tank. The diffuser recovers part of the pressure. Suction vacuum at the throat runs 0.3 to 0.6 bar below atmosphere at design flow.

State the ratio convention before you buy. Most suppliers quote motive litres per litre of stock drawn. A 4:1 unit passes 4 L of motive water for each litre pulled in. Entrainment of 1:1 to 4:1 covers most water-driven units. The ratio falls as back pressure rises. Double the discharge back pressure and the drawn flow can drop by a third.

Pressure loss is the hidden cost. A unit holding 4:1 at its rated flow may shed 30 to 50 percent of motive pressure. The downstream system then sees 1.4 to 2.8 bar from a 2.8 bar supply. Size the motive pump for that loss, not for the injector inlet.

An eductor holds a ratio, not a ppm. Vary the motive pressure and the ratio drifts with it. The vacuum follows the square of the throat velocity. For steady duty that is harmless. For variable flow, drive the motive line from a flow-proportional signal. The alternative is a metering pump for the chemical injection duty.

A venturi also doubles as a tank mixer when it discharges below the liquid surface. It entrains two to five tank volumes per unit of motive flow. The eductor range carries both the pumping and the mixing variants. The eductor sizing calculator converts motive pressure and back pressure into an entrainment figure. The water eductors guide works the throat pressure balance.

How Long Should a Static Mixer Be?

A static mixer is a bank of fixed elements. They repeatedly split, rotate and recombine the flow. Length is quoted in pipe diameters. Each helical or cross-flow element adds about 1.5 D. A 6 element mixer is roughly 9 D long. A 12 element bank is about 18 D.

Six to twelve elements handle a miscible, low-viscosity stock at 1 to 3 m/s. Twelve elements bring the outlet coefficient of variation below about 5 percent. Below 1 m/s the elements stop generating the vortices they depend on.

Pressure drop is the price. Use ΔP = K × ρ × v² / 2, where K counts the element losses and ρ is the density. At 1.77 m/s in a 100 mm line the dynamic pressure ρv²/2 is 1.56 kPa. That is 0.016 bar.

A typical K runs about 0.7 to 0.8 per element at turbulent flow. Six elements give K of 5 and a drop of 0.078 bar. Twelve give K of 9 and 0.14 bar. Double the flow and the velocity doubles. The pressure drop then quadruples to 0.56 bar.

Keep the velocity above 1 m/s. Keep 10 D of straight pipe between the injection point and the first element. Leave the bank removable so a fouled chemical mixer can be cleaned.

Where to Put the Injection Point

Put the injection point 10 pipe diameters upstream of the first mixing element or bend. Closer than 10 D and the stock arrives as a slug the pipe cannot smooth. On a 100 mm main, 10 D is 1.0 m.

Stop the quill tip at the pipe centreline, half a diameter in. A tip that stops short of centre leaves the stock in the fast core. A tip that reaches the wall wets it with full-strength product and seeds a deposit. Point the tip downstream and bevel the opening.

Give the mixture a straight run after the injection point as well as before it. Turbulent eddies need about 20 D to even out a transverse concentration gradient. A flow meter 3 D downstream of a quill reads whatever passes its port. It does not read the line average.

Separate incompatible injections by at least 10 D, and preferably by a static mixer. Sodium hypochlorite and an acid injected at one point can drive local pH below 4. That pocket evolves chlorine gas while the bulk pH still reads 7. The pocket corrodes the pipe from the inside.

The same geometry governs a spray injection nozzle feeding an open channel. A full cone nozzle must cover the cross-section at its mounting standoff. The width law gives the angle: W = 2 × standoff × tan(angle / 2). A 100 mm flume with the nozzle 30 mm above the surface needs 2 × 30 × tan(angle/2) = 100. The half angle is 59 degrees, so the full cone is about 118 degrees.

Why Droplet Size Sets Contact Efficiency

The physical form of the injected stock sets how fast it reaches the bulk liquid. A bare quill releases a stream 1 to 3 mm across. That stream blends by shear and diffusion alone. An atomizing injection nozzle breaks the same flow into 100 to 500 micron droplets.

Interfacial area per litre scales as 1 divided by droplet diameter. Halving the diameter from 400 to 200 microns doubles the available area. That area drives a fast reaction such as chlorine demand or neutralisation. For a reaction complete in under a second, area is the rate-limiting term.

Droplet size follows the nozzle flow law Q = K√P. K folds the orifice area and the discharge coefficient into one constant. The same K factor sets a flat fan’s flow, spray angle and mean drop size. A nozzle held below its rated band runs coarse. One pushed above the band runs fine.

For injection into an open channel, droplet class matters more than pipe mixing. Work to the ASABE S572.1 volume median diameter bands. Hold a coarse class above 326 microns where wind can carry the plume. Drop to a fine class only where the spray is fully enclosed.

Back pressure decides whether atomization is possible at all. A pressure-swirl nozzle needs about 2 bar across the orifice for a 200 micron spray. On a 5 bar main the dosing pump must therefore deliver 7 bar at its discharge.

Which Material Survives the Dosing Line

Material choice is a chemistry decision first and a strength decision second. A wrong choice shows up as a drifting dose, not a burst pipe.

316L stainless suits neutral salts, raw water and clean water treatment. It pits under hypochlorite above about 1 percent. Free chloride attacks the passive film while the low pH removes it. Keep it off the chlorine line.

PVC is the standard body for hypochlorite, mineral acids and caustic to about 60 °C. It is cheap to rework. It softens above 60 °C and many solvents attack it.

PTFE is inert to almost every water treatment chemical and holds to about 260 °C. It is soft and creeps under a threaded load. Use it as a wetted liner or a body insert. PVDF handles oxidisers and hypochlorite to about 140 °C. It machines into quills and small orifices, so most chlorine quills are PVDF.

Holding the Dose Through Turndown

Turndown is the ratio of maximum to minimum flow over which a device holds accuracy. A diaphragm metering pump typically holds 10:1. A good one reaches 20:1, and a control valve reaches 20:1 to 50:1. The valve figure holds only inside its own authority limits.

Size the pump so the lowest duty stays above 10 percent of capacity. Below that the check valves stop seating cleanly. Repeatability falls away. A 10 L/h pump asked for 0.2 L/h is running at 2 percent. It will not hold any figure it prints.

A proportional control valve feeding a fixed orifice has its own trap. Flow through the valve follows the square root of the pressure drop across it. That is not the same as the valve opening. A quarter-open valve passes about half the flow.

PWM control adds a second distortion. Duty cycle sets the mean flow. The instantaneous flow is either on or off. A solenoid with a 40 ms closing time turns a 100 ms pulse into 60 ms of flow. That is a 40 percent error at that duty.

Pulsation: The Hidden Cause of Dose Drift

Every positive-displacement pump pulses. A single diaphragm unit at 60 strokes per minute delivers slugs rather than a stream. Close the injection point to the pump and that pulsation arrives as a concentration ripple.

A pulsation dampener cuts the ripple to a few percent. Precharge the gas at 60 to 80 percent of line pressure. Size the volume at 5 to 20 times the stroke volume. Size it on stroke volume, not on mean flow. The ripple amplitude follows the slug rather than the average.

Residual ripple of plus or minus 10 percent becomes a 10 percent dose ripple. A static mixer on a 100 mm main averages a 5.6 second window. That is longer than a 1 second stroke period. An unmixed quill averages nothing.

How to Prove the Mix Is Real

Measure the mixture instead of trusting the arithmetic. A tracer test is the cheapest proof. It needs no new instrument on most sites.

Run the line at steady flow and step the dose from zero to the working figure. Log a downstream conductivity or colorimeter reading. Note the time to reach 95 percent of the final value. That interval is the mixing time constant for the installation.

Residence time sets the floor for that figure. Pipe volume divided by flow gives the transit time. A 10 m length of 100 mm pipe holds 78.5 L. At 50 m³/h the transit is 5.6 seconds. A longer mixing time means the mixture is still changing as it leaves.

Sample across the cross-section rather than at one port. Take readings at 10 D, 20 D and 30 D downstream. Compare a centre sample with a wall sample. A coefficient of variation under 5 percent means the blend is done. A wall sample 30 percent above the centre means the quill is off-centre.

Inline conductivity is the working instrument for continuous duty. A cell with 0.1 percent of range resolution sees a 1 to 5 ppm step. Trend it against the flow meter and the pump stroke. Divergence between the two is the earliest sign of a worn nozzle.

The Eductor Route, and Where Fertigation Meets This Math

A motive-water eductor is the no-power option in the same family. A chemical eductor draws stock with the pressure energy of the main line, so it costs nothing to run and has no stroke to wear. Ask a supplier for an eductor pump and this is usually the unit meant. The dose arithmetic above holds unchanged; only the entrainment ratio moves, at 1:1 to 4:1 litres of motive water per litre of stock. The same equations answer a second buyer question, the fertilizer injector for irrigation system, where the treated stream is crop water instead of cooling water and the stock is fertiliser. Our fertigation pump selection guide runs the nutrient case end to end, and the greenhouse retrofit case prices what a drift in stock strength costs a season.

Failure Modes and Their Numbers

Injection point too close to a bend is the most common fault. The stock travels as a ribbon along one wall. That wall sees a local concentration several times the target. The bulk sees almost none. The ribbon scales a heat exchanger while the analyser reads normal.

Nozzle erosion moves the dose in one direction. Flow through a fixed orifice scales with the square of the diameter. An orifice that erodes 10 percent in diameter passes 21 percent more flow. A pressure-fed chlorine line drifts 20 percent rich before recalibration.

Nozzle corrosion in the other sense is worse. A quill that loses its tip shape stops throwing into the core. It starts running down the wall instead. Pump discharge pressure then falls and flow rises at constant stroke. The dose drifts 20 percent or more across a season.

Clogging shows up as back pressure before it shows up as a dose error. The rule is a filter screen opening near one quarter to one third of the smallest free passage. A nozzle with a 1.0 mm passage wants a screen of 250 to 330 microns. That is about 50 to 60 mesh. A 0.5 mm passage wants 125 to 165 microns, near 100 to 120 mesh. A blocked screen lifts pump discharge from 3 to 6 bar. The relief opens and delivery turns erratic.

Two injections in one pocket is the fastest route to a process safety event. Hypochlorite and acid within 10 D can pull local pH under 4. That releases chlorine gas into the headspace. Treat the spacing as a design rule rather than a preference.

A Sizing Walkthrough for 50 m³/h at 2 ppm

The duty is 50 m³/h at 2 ppm from a 10 percent stock at unit density. The dose rate is (50 × 2) divided by (10 × 10), which is 1.0 L/h. Round up one pump frame so the duty sits mid-band rather than at its floor.

The main is 100 mm, so velocity at 50 m³/h is 1.77 m/s. That sits inside the 1 to 3 m/s band a static mixer needs. Transit time through a 10 m section is 5.6 seconds.

For steady flow and a single chemical, a 6 element static mixer costs 0.078 bar. It blends within 10 D. For variable flow where the stock must also be drawn, an eductor sized at 3:1 covers both jobs. It loses about 30 percent of motive pressure.

Place the quill 1.0 m upstream of the mixer. Stop it at the centreline, 50 mm in, and point it downstream. Leave 2.0 m, which is 20 D, before the first analyser or sample port.

Fit a 60 mesh screen at 250 micron opening ahead of a 1.0 mm quill passage. Add a pulsation dampener sized at 5 to 20 times stroke volume for a diaphragm pump. Verify with a step test at design flow. The 95 percent mixing time should fall inside the 5.6 second transit.

Frequently Asked Questions

How do I convert a ppm target into a pump setting? Divide (water flow in m³/h × ppm) by (stock percent × 10) to get the dose rate in L/h. For 50 m³/h at 2 ppm from a 10 percent stock, that is 1.0 L/h. Adjust for stock density above 1.0. A 12.5 percent hypochlorite is about 19 percent denser than water.

Do I need both an eductor and a static mixer? Not usually. An eductor draws and mixes in its throat. A steady-flow line can run on the eductor alone with 10 to 20 D of downstream pipe. Add a static mixer when the flow varies or the stock is viscous. Add one when the blend distance must be shorter than the available run.

Why does my dose drift after a few months? The three usual causes are orifice erosion, quill corrosion and a partly blocked strainer. An orifice 10 percent wider in diameter passes 21 percent more flow. Re-run a step test and compare discharge pressure with the commissioning sheet.

The Bottom Line

Dose accuracy lives in the arithmetic and the geometry before it lives in the hardware. Divide flow and ppm by stock strength and the pump setting falls out. Put the injection point 10 D upstream of the mixer and stop the quill at the centreline. Give the mixture 20 D to even out.

Then prove it. A tracer step, an inline conductivity trend and a conservative screen catch the faults early. The eductor range covers the suction and inline mixing units.

Send your water flow, target ppm, stock strength, line size and main pressure to the application team. They will come back with a quill or injector size and a stock flow in L/h. They will also give the mixing length for your line.