On this page
Ozone treatment performance does not end at the generator outlet. A generator may deliver its specified ozone mass and concentration while the treatment process receives less effective gas-liquid contact than anticipated. The missing link is often hydraulic: the relationship among sidestream flow, pump head, venturi operating conditions, gas delivery, and downstream mixing.
This distinction matters in municipal drinking water, wastewater reuse, industrial process water, and aquaculture systems. An injection skid must operate not only at its design point but also as plant flow, pressure, water temperature, and ozone demand change.
A venturi can draw ozone into water without proving that the intended ozone mass was transferred or that the required process residual was achieved.
Three Different Questions Engineers Must Answer
- Ozone production: How many grams per hour of ozone are generated at the specified gas concentration?
- Gas introduction: At what gas-flow and pressure conditions can the injector accept the ozone-bearing stream?
- Process transfer: How much ozone crosses from the gas phase into the water, where does it react, and what residual or treatment result remains?
These questions require different measurements. Generator output alone does not establish injector performance; injector suction alone does not establish mass-transfer efficiency; and dissolved-ozone residual alone does not reveal how much ozone has reacted with contaminants.
An Injection Skid Is a Hydraulic System
In a common sidestream arrangement, part of the process water is withdrawn, pressurized by a pump, passed through an injector, then returned to the main process flow for mixing and downstream contact. The pump must overcome piping and equipment losses while producing the inlet and outlet conditions required by the selected injector. The process must also provide sufficient downstream pressure, mixing, and contact to complete the intended treatment.
The engineering boundary should include the withdrawal location, suction piping, pump, injector, ozone gas connection, discharge piping, return nozzle or mixer, contactor, instrumentation, and protective controls. Designing only the venturi or only the pump can miss the interactions that govern full-system performance.
Start With a Defensible Design Basis
Before selecting hardware, establish the peak and minimum water flow, intended treatment objective, water temperature and chemistry, target applied ozone dose, generator gas concentration range, and pressure at both sidestream tie-ins. For pressure-driven injection, identify the highest expected downstream pressure, not only the normal condition. Specify which performance quantities must be proven by calculation, factory test, or field verification.
Where disinfection is the objective, dissolved ozone exposure and validated contact hydraulics must be assessed separately from the injection equipment. High transfer efficiency is useful, but does not by itself establish CT compliance or oxidation performance.
Venturi Differential Pressure: Necessary, Not Sufficient
A venturi accelerates water through a constriction, creating a lower local static pressure at its gas port. Depending on injector geometry and the liquid-side operating point, this pressure can entrain ozone-bearing gas. The injector then recovers some pressure in its diffuser, with a permanent hydraulic loss.
The practical design variables are the actual injector inlet pressure, outlet pressure, water flow, gas suction or feed pressure, and gas composition. Two installations with the same nominal pressure differential can have different gas capacities. The local throat pressure, not merely the difference between two pressure gauges, governs gas entry.
Manufacturer performance curves should therefore be reviewed at the complete operating envelope. A suction-capacity rating at one inlet/outlet pressure combination must not be carried into a different plant condition without verification.
Why the suction connection matters
An ozone injector is also a boundary between pressurized water and ozone gas equipment. Loss of pump flow, downstream pressure transients, or failed valves can drive moisture toward the gas line. Design needs more than a single check valve: piping geometry, pressure monitoring, appropriate isolation and drainage/condensate strategy, and generator interlocks must be reviewed together. The specific protective arrangement depends on the injector and generator configuration.
Sidestream Flow Is a Design Variable, Not a Rule of Thumb
Sidestream flow is sometimes expressed as a percentage of the main process flow. That may help during early concept development, but there is no universal percentage that ensures ozone transfer. The necessary flow depends on injected gas volume, injector geometry, pump pressure, downstream mixing, water quality, and the required transfer efficiency.
Increasing sidestream water flow may improve gas handling and contact, but it can also raise pumping power. Reducing sidestream flow lowers pumping duty only if the injector still operates within its verified capacity and the downstream process finishes the transfer.
Pump Selection Requires the Full System Curve
A pump must be selected where its head-flow curve intersects the hydraulic resistance of the actual skid. Engineers should account for suction conditions and net positive suction head (NPSH), strainers, piping losses, injector pressure requirements, fittings, control valves, static mixers, return elevation, and variable plant pressure.
The highest process pressure is often the hardest condition for gas induction, while low suction pressure or warm water may create a separate NPSH limitation. The design review should check both, as well as minimum-flow operation and pump protection.
Do not specify an injector by nominal pipe diameter, or a pump by motor horsepower, without verifying their combined hydraulic operating point.
Check the Operating Envelope, Not Just the Best Case
| Operating change | Likely hydraulic consequence | Design verification |
| Mainline backpressure rises | Less injector discharge margin | Gas capacity and pump head |
| Water flow turns down | Injector moves off nominal curve | Minimum stable gas uptake |
| Suction pressure falls | Less NPSH available | Pump NPSH margin and alarms |
Where energy optimization matters, compare the expected annual pump power at several operating cases rather than a single peak condition. A design with lower pumping energy may require more downstream mixing or contacting capacity, and vice versa.
Worked Example: From Ozone Dose to Skid Duty
Consider a process water stream of 1,000 gallons per minute (gpm) with a preliminary applied ozone dose of 2.0 milligrams per liter (mg/L). The applied dose is ozone sent to the process per unit volume of water, not a dissolved residual or a guaranteed transferred dose.
Ozone applied (g/h) = Flow (gpm) × Dose (mg/L) × 0.2271
1,000 × 2.0 × 0.2271 = approximately 454 g/h of applied ozone
For an ozone-bearing gas containing 10% ozone by mass, the corresponding total feed-gas mass flow is approximately 4.54 kg/h. Gas volume for injector selection must be calculated at the specified reference or actual temperature and pressure, accounting for gas composition. Ozone weight percent, total gas flow, and ozone mass flow must never be used interchangeably.
Illustrative preliminary checks, not equipment selections
| Design quantity | Illustrative value | Engineering meaning |
| Main process flow | 1,000 gpm | Basis for dose calculation |
| Applied ozone dose | 2.0 mg/L | Process design assumption |
| Ozone mass required | 454 g/h | Gas-phase ozone delivered |
| Feed-gas concentration | 10 wt% | Total gas ≈ 4.54 kg/h |
| Candidate sidestream flow | 150 gpm | 15% assumed; must validate |
| Example pump differential | 30 psi | Illustrative total developed pressure |
At 150 gpm and an assumed 30 psi of total pump differential, with 65% pump efficiency, the idealized pump shaft-power estimate is:
Brake horsepower ≈ (150 × 30) ÷ (1,714 × 0.65) = 4.0 hp
This estimate is not a motor-size recommendation. The actual pump must be sized from the complete duty curve, including all losses, minimum and maximum pressure cases, NPSH margin, efficiency variation, and the chosen injector’s required inlet/outlet conditions.
What Does a 90% Transfer Efficiency Mean?
If 454 g/h enters the contacting process and an illustrative 90% of that ozone mass transfers from gas to liquid before off-gas removal, approximately 409 g/h has transferred. For a 1,000 gpm main flow, that is an equivalent transferred dose of about 1.80 mg/L. It does not imply that 1.80 mg/L remains as a dissolved ozone residual: transferred ozone can react or decompose quickly.
If 2.0 mg/L were instead a required transferred dose, the necessary applied dose would be approximately 2.22 mg/L at the assumed 90% transfer efficiency, before additional process-design allowances. Transfer efficiency must be established by testing or defensible design evidence, not assumed from a suction gauge.
A Useful Gas-Flow Cross-Check
With the example assumption that the remaining 90% of the gas mass is oxygen, 4.54 kg/h of ozone/oxygen mixture corresponds to about 3.07 normal m³/h at 0°C and 1 atmosphere. Actual gas volume at an injector may differ materially because of temperature and absolute pressure. Vendor data must therefore state exactly whether gas capacity is normal, standard, or actual volumetric flow and at what suction pressure it applies.
The approximately 4.0 hp pump result and 3.07 normal m³/h gas result are two different calculations: the former is a liquid-side hydraulic power estimate, while the latter is a gas-side molar flow estimate. Neither is a substitute for injector selection from manufacturer data.
Mixing After the Injector Completes the Job
The injector initiates gas-liquid contacting; downstream hydraulics influence whether that contact is completed. A sidestream can carry undissolved ozone and oxygen bubbles beyond the venturi. Pipeline length, pressure, mixing energy, and recombination design determine how that two-phase mixture encounters the main process flow.
Depending on the application, a return nozzle, static mixer, or contactor geometry may increase interfacial renewal and disperse bubbles. Each device introduces hydraulic resistance, so its mixing benefit must be evaluated against pressure loss and pump energy. A design that includes degas separation has different gas-handling and energy requirements from a design that relies on downstream mainstream mixing.
Instrumentation That Separates Cause From Symptom
- Liquid hydraulics: measure sidestream flow and relevant pump/injector inlet and outlet pressures; monitor suction pressure or level when required.
- Gas delivery: monitor ozone concentration and gas flow on a clearly stated mass or standard-volume basis; confirm gas-port pressure and supply conditions.
- Process response: monitor dissolved ozone at appropriate locations, water flow and quality, and off-gas when applicable. A residual is not a direct transfer-efficiency measurement.
- Protection: verify loss-of-flow, high-pressure, water-ingress, ambient ozone, and off-gas-destruct permissives and shutdown behavior appropriate to the installation.
For credible troubleshooting, trend the signals together. A drop in dissolved ozone might be caused by higher water demand, lower applied ozone mass, reduced liquid circulation, unstable gas intake, or incomplete mixing. One trend by itself rarely identifies the cause.
Five Failure Modes Worth Designing Out
- Rising return-line pressure shifts the injector off its validated curve, reducing gas intake while the generator still reports normal production.
- A clogged sidestream strainer lowers pump suction performance, reducing water flow and causing unstable injection or cavitation risk.
- An oversized injector or incorrectly chosen control range operates poorly at minimum plant flow even though the maximum-flow case passes.
- Gas pockets persist after injection because the return geometry or static-mixer arrangement does not provide sufficient redistribution.
- A pump trip or pressure transient drives water toward the ozone gas system because protection depends only on one check valve.
Commissioning Must Exercise More Than Full Load
Acceptance testing should include low, normal, and maximum water flows; representative minimum and maximum downstream pressures; ozone gas setpoint changes; pump start and shutdown sequences; loss-of-flow responses; and instrument/communications faults. The delivered records should show which operating points were actually demonstrated and which remain assumptions pending process verification.
What Mass-Transfer Theory Adds
At its simplest, gas-liquid ozone transfer is driven by interfacial area and the difference between equilibrium dissolved concentration and bulk dissolved concentration. It is often represented as a volumetric transfer term proportional to kLa(C* − C). Higher pressure can increase the equilibrium driving force, while efficient dispersion can increase effective interfacial area. Ozone reaction and decay reduce bulk concentration and complicate interpretation; an apparently low residual may coexist with substantial ozone transfer and consumption.
For this reason, transfer testing should rely on a defensible mass balance using gas-in/gas-out information and appropriate measurements, with correction for sampling conditions. Relying only on the water-side residual can underestimate transferred ozone when demand is high.
Questions for the Engineering Design Review
- What is the required applied ozone mass at minimum, normal, and peak water flow, and what process objective sets that requirement?
- At each operating case, what are the injector inlet pressure, outlet pressure, liquid flow, and gas-port pressure?
- Do vendor performance data support gas capacity with the specified ozone-bearing gas at the design conditions?
- Where are the system-curve calculation, pump curve, efficiency basis, and NPSH review?
- How does the design accommodate changing plant backpressure and ozone-generator turndown?
- What is the downstream mixing strategy, and where can undissolved gas be separated and safely destroyed?
- Which field measurements distinguish inadequate gas delivery, incomplete transfer, and increased water-side ozone demand?
- What happens to gas isolation and liquid flow after an unplanned power loss?
The Pinnacle Engineering Perspective
At Pinnacle Ozone Solutions, the engineering objective is not simply to connect an ozone generator to a venturi. It is to create a controllable interface between ozone production and the water-treatment process. That requires a common design basis across process engineering, gas handling, hydraulics, mechanical equipment, instrumentation, electrical controls, and commissioning.
A strong injection-skid specification identifies the required performance across an operating envelope, rather than describing a single design-point flow and horsepower. It also separates what can be verified at the factory from what must be demonstrated with the actual plant hydraulics and water chemistry.
The generator defines the available ozone mass. The hydraulic and contacting system determines whether that mass can reach the water under real operating conditions.
Conclusion
An injection skid is not a minor accessory to an ozone generator. It is part of the core process equipment. Venturi suction, pump selection, sidestream ratio, gas pressure, downstream mixing, and controls must be evaluated as one hydraulic system. A well-designed skid remains stable when plant conditions change and supplies the evidence operators need to understand treatment performance.
Define the Acceptance Evidence
A strong skid submittal includes a hydraulic calculation, vendor injector selection curve, pump performance curve, P&ID, instrument list, cause-and-effect matrix, and commissioning test procedure. Field acceptance should record as-found pressures, flow, gas conditions, instrumentation checks, protective shutdown responses, and the water-quality basis for process performance. A pass/fail requirement is credible only when its measurement method and operating condition are stated.
Technical Source Notes
- U.S. Environmental Protection Agency. Alternative Disinfectants and Oxidants Guidance Manual, EPA 815-R-99-014 (1999), Chapter 3: Ozone, especially Section 3.2.2 on injector contacting, contactors, off-gas and instrumentation. nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=2000229L.TXT
- Wert, E. C., Lew, J., and Rakness, K. L. “Design Considerations for Cost-Effective Ozone Mass Transfer in Sidestream Systems.” Ozone: Science & Engineering (2018), DOI: 10.1080/01919512.2018.1424532. Covers sidestream configurations, pressure, G/L ratio, and water-ingress protection.
- Mazzei Injector Company. Venturi Injectors; Injector Performance Data/Drawings; Technical Bulletin No. 11, Installation Recommendations for Gas-to-Liquid Mixing Applications. mazzei.net
- Rakness, K. L. Ozone in Drinking Water Treatment: Process Design, Operation, and Optimization. American Water Works Association (2005), section on sidestream injection and gas-liquid contacting.
