A dosing pump specified to ±1% accuracy can deliver poor process results if the injection point is wrong. The pump delivers the correct volume. The chemical does not reach the process in the correct form, at the correct concentration, at the correct time. Engineers spend considerable effort selecting pump accuracy and flow ranges, and comparatively little on where and how the chemical enters the process stream. This is usually where the actual dosing error originates.
The Mixing Problem at the Injection Point
Chemical dosing relies on the injected reagent mixing with the process fluid to achieve a target concentration or reaction. If mixing at the injection point is poor, the chemical accumulates locally at high concentration before gradually dispersing downstream. The result is a mismatch between where the process instrument (pH probe, conductivity sensor, ORP meter) is located and where the chemical is actually distributed.
In a pipe with low velocity — below about 1 m/s — injected chemical stratifies rather than dispersing radially. A pH correction reagent injected at the top of a slow-moving pipe will create a high-pH layer at the top while the bulk fluid at the probe location (typically at mid-pipe or downstream) reads a different value. The control system responds to the probe reading, not to the actual bulk concentration, and dosing becomes incorrect by the degree of stratification.
The standard correction is to inject into turbulent flow: Reynolds number above 10,000 provides sufficient mixing energy for rapid dispersion. In a 50 mm pipe carrying process water at 1 m/s, Re is approximately 50,000 — well within the turbulent regime. In the same pipe at 0.2 m/s, Re drops below 10,000 and mixing becomes diffusion-limited. The injection point should be selected for the operating velocity, not the peak velocity.
Injection Depth and Jet Velocity
A dosing injection quill that terminates flush with the pipe wall injects chemical into the pipe boundary layer — the slowest-moving, lowest-turbulence region of the flow. The chemical migrates from the wall toward the pipe centre by diffusion and shear, a slow process. An injection quill that extends to the pipe centreline injects chemical into the highest-velocity region of the flow, where turbulent mixing is most intense and dispersion is fastest.
Jet velocity at the injection point matters as well. A quill with a small orifice at low flow rate produces a low-velocity jet that is immediately entrained by the surrounding flow without penetrating the boundary layer. Increasing the jet velocity by using a smaller orifice diameter — at the same volumetric flow rate — increases jet penetration into the cross-sectional flow and improves mixing. For chemical injection into large-diameter pipes (above 200 mm), a spray nozzle or multi-point injection manifold is preferable to a single centreline quill, since a single-point injection at any Reynolds number cannot achieve uniform radial distribution across a large cross-section in a short mixing length.
The Distance Between Injection and Measurement
The mixing length required between injection point and measurement point is a function of pipe diameter, velocity, and turbulence intensity. A commonly used rule for well-designed injection into turbulent flow is 10–20 pipe diameters of mixing length. In a 50 mm pipe, this is 0.5–1.0 metres. In a 200 mm pipe, it is 2–4 metres. Many installations place the measurement instrument closer than this, particularly when retrofitting dosing into existing pipework with limited space.
When the measurement instrument is upstream of the minimum mixing length, the probe reads a non-representative local concentration. In pH control, this causes the control loop to chase a reading that reflects injection proximity rather than bulk fluid composition. The result is oscillating pH, excessive chemical consumption, and sometimes pH overshoot severe enough to damage the process.
The correct approach when space is constrained is to use a static mixer between the injection point and the probe. A static mixer achieves in 2–3 pipe diameters what turbulent pipe flow achieves in 10–20. In retrofitted systems where the injection point and probe location are fixed, a short static mixer section inserted between them solves the mixing length problem without pipe rerouting.
Back-Pressure at the Injection Point
The injection point must have higher pressure in the dosing line than in the process pipe at the injection location, or backflow occurs. This is obvious in principle but frequently neglected when process pressure varies. A dosing pump specified for 3 bar discharge pressure will inject correctly into a process line at 2 bar. The same pump, if the process pressure rises to 3.5 bar — during startup, during filter backwash, during valve closure — will stall or allow process fluid to flow backward into the dosing line.
Back-pressure at the injection point should be evaluated at the maximum process pressure, not average pressure. For AODD pumps, the maximum drive air pressure sets the maximum discharge pressure; for diaphragm dosing pumps, the motor rating limits maximum discharge. A check valve at the injection quill is necessary but not sufficient — at very low differential pressure, a check valve with worn seats passes fluid in both directions. The dosing pump’s maximum rated discharge pressure must exceed the maximum process pressure by a comfortable margin, typically 0.5–1 bar minimum.
Injection Point Chemistry
Some chemicals react with the process fluid immediately on contact, forming precipitates or heat that can block the injection quill. Sodium hydroxide injected into a calcium-rich process stream precipitates calcium hydroxide at the injection orifice. Sodium hypochlorite injected into an acidic stream releases chlorine gas locally. Ferric sulphate injected into a high-pH stream forms immediate precipitates.
The injection point should be located where the local chemistry does not cause rapid fouling or hazardous conditions. This sometimes means injecting into a dilution stream before the main process, or using a dilution quill that pre-dilutes the chemical with water before injection into the process. The pump specification has no bearing on these interactions — they are determined entirely by the chemistry at the injection location and the physical design of the quill and its surrounding environment.
Getting the injection point right is not more complicated than selecting the right pump. It simply requires the same systematic attention to the physical conditions at the injection location — velocity, mixing length, back-pressure, local chemistry — that is routinely applied to pump selection.
For help designing a complete chemical dosing system including injection point selection, contact Autoflo at info@autoflotechnology.com.