Ozonation water treatment uses ozone (O3), generated on site by corona discharge, as a powerful oxidant and disinfectant. Dosed into water it acts through two routes — direct molecular ozone and indirect hydroxyl radicals — to inactivate pathogens such as Cryptosporidium and to oxidise colour, taste, odour and micropollutants. Design turns on gas transfer, applied dose and CT.

What is ozonation and why use it?

Ozonation is the dosing of ozone into water to disinfect it and to oxidise dissolved and colloidal contaminants. Ozone is the strongest common water-treatment oxidant, with a standard electrode potential of +2.07 V — well above chlorine (+1.36 V) — and it leaves no persistent chlorinated residual. Because ozone is unstable (half-life in water is typically only minutes), it cannot be stored or shipped: it must be generated on site and used immediately.

It is chosen where chlorine struggles: primary disinfection against chlorine-resistant protozoa, destruction of taste-and-odour compounds (geosmin, 2-methylisoborneol), decolourisation, iron and manganese oxidation, and breaking down trace organic micropollutants. Ozonation also improves downstream biofiltration by converting refractory organics into biodegradable fragments (assimilable organic carbon).

How is ozone generated?

Virtually all industrial ozone is made by corona (silent electrical) discharge. A high-voltage alternating field (several kV, ~0.5–5 kHz) is applied across a narrow discharge gap containing a dielectric barrier. Energetic electrons split molecular oxygen, and the atomic oxygen recombines with O2 to form ozone:

O2 + e− → 2 O·   ;   O· + O2 + M → O3 + M
The reaction is exothermic and heat-sensitive — ozone decomposes above ~35°C — so discharge cells are water-cooled. Feed gas must be dried to a dewpoint below about −60°C, otherwise nitric acid forms and yield collapses.

The two practical feed gases are dried ambient air and high-purity oxygen (from liquid oxygen or on-site VPSA). Oxygen feed roughly doubles the ozone concentration and the energy yield, at the cost of the oxygen supply:

Feed gasO3 concentrationEnergy yield (g O3/kWh)*Notes
Dried ambient air1–3 wt%~50–100Low gas cost; large air-prep (drying) plant; lower O3 partial pressure
High-purity oxygen6–14 wt%~100–250Higher yield and transfer driving force; needs LOX or VPSA supply

*Yield falls as the target ozone concentration rises; the highest yields correspond to low-concentration operation. Figures include the discharge cell, not always the gas-prep plant.

Molecular ozone versus hydroxyl-radical pathways

Once in water, ozone reacts by two competing routes. The direct pathway is reaction of molecular O3 itself — a selective, relatively slow electrophile that attacks electron-rich sites (double bonds, activated aromatics, amines) and drives disinfection. The indirect pathway begins with ozone decomposition, initiated by hydroxide, that produces the hydroxyl radical (·OH), a far faster but non-selective oxidant.

O3 + OH− → … → ·OH (radical chain)
The balance is set by pH and by radical scavengers. Rising pH accelerates decomposition, so alkaline water shifts toward the ·OH route; carbonate/bicarbonate alkalinity and dissolved organic matter scavenge ·OH and quench the chain, favouring slower molecular-ozone chemistry and lengthening the ozone residual.

This is exactly the lever exploited in advanced oxidation. Deliberately maximising ·OH (for example with O3/H2O2 or O3/UV) moves you from plain ozonation to an advanced oxidation process; see our note on advanced oxidation processes for recalcitrant organics. For robust disinfection you generally want the opposite — a stable molecular-ozone residual to build CT.

How does ozone transfer from gas into water?

Ozone is delivered as a gas and must dissolve before it can react, so ozonation is a gas–liquid mass-transfer problem. Equilibrium at the interface follows Henry's law, and the flux is driven by the difference between the saturation and bulk concentrations:

CL* = Cgas / Hc  and  N = kLa (CL* − CL)
CL* = saturation concentration in water; Hc = dimensionless Henry constant (≈3–4 for O3 at 20°C, rising with temperature so warm water holds less); kLa = volumetric mass-transfer coefficient (s−1); CL = bulk dissolved ozone. Fine bubbles and turbulence raise kLa and thus transfer efficiency.

Contact is achieved with fine-bubble diffuser columns, side-stream venturi (Pinch/eductor) injection, or turbine mixers. Because ozone both dissolves and decomposes, three doses must be distinguished:

  • Applied dose — ozone mass fed to the contactor per unit water flow (g/m³).
  • Transferred dose — applied dose × transfer efficiency (typically 85–95% in a good deep contactor); the difference leaves in the off-gas.
  • Absorbed (consumed) dose — transferred dose minus the measured dissolved residual; this is what actually reacted.

Selecting the contactor and injection method is core to any ozone contactor and generator package, and the hydraulic detail is worth confirming at the process and contactor design stage.

Worked example: ozone dose, mass and generator power

Size the ozone supply for a surface-water plant treating Q = 100 m³/h that needs a transferred dose of 5 mg/L for combined oxidation and disinfection. Assume a transfer efficiency of 90% and an oxygen-fed generator with a specific yield of 120 g O3/kWh.

  • Transferred ozone mass: ṁt = Q × Dt = 100 m³/h × 5 g/m³ = 500 g/h.
  • Applied (fed) ozone mass: ṁa = ṁt / TE = 500 / 0.90 = 556 g/h. The remaining 56 g/h reports to the off-gas.
  • Generator power: P = ṁa / Y = 556 g/h ÷ 120 g/kWh = 4.6 kW at the discharge cells.
  • Oxygen demand: at ~10 wt% ozone, feed-gas O2 ≈ 556 / 0.10 = 5.6 kg/h (about 3.9 Nm³/h of oxygen), plus cooling water and cell power supply.
P = (Q · Da) / Y   where Da = applied dose, Y = specific energy yield.
The same duty on a dried-air machine at Y ≈ 70 g/kWh would draw ~7.9 kW — roughly 70% more electrical power for the discharge alone, which is why higher-dose plants tend to use oxygen feed.

Round up to a generator with output margin (say 700–800 g/h) to cover peak demand, ageing dielectrics and off-spec feed gas.

How is ozone dose set for disinfection (the CT concept)?

Disinfection credit is earned through CT — the product of the dissolved ozone residual C (mg/L) and the contact time T (min). Inactivation follows Chick–Watson first-order kinetics, so log-removal scales with CT:

ln(N/N0) = − Λ · C · T
N/N0 = surviving fraction; Λ = Chick–Watson rate constant (organism- and temperature-specific). Regulators publish required CT (mg·min/L) per log of removal; T is credited as T10 (the time 90% of water is retained) to penalise short-circuiting.

Ozone's standout property is efficacy against Cryptosporidium, which chlorine cannot practically inactivate. Using US EPA LT2 values, roughly 10 mg·min/L achieves 2-log Cryptosporidium inactivation at 10°C (values rise sharply as temperature falls). Sizing the contactor for that target: with a residual C = 0.5 mg/L we need T10 = 10 / 0.5 = 20 min. If the contactor baffling factor is 0.5, the actual mean hydraulic residence time is 20 / 0.5 = 40 min, so at 100 m³/h the required contact volume is 100 × (40/60) ≈ 67 m³. Viruses and Giardia need far less CT, so Cryptosporidium usually governs. For non-chemical primary disinfection, ozone is often paired with or compared against UV disinfection system design.

What can ozone oxidise, and what about bromate?

Beyond disinfection, ozone is dosed to oxidise specific contaminants. Typical transferred doses:

DutyTypical transferred doseNotes
Disinfection (Crypto/virus)CT-driven, ~1–3 mg/LMaintain a measurable residual
Taste & odour (geosmin, MIB)1–3 mg/LRadical route helps
Colour / DOC decolourisation2–5 mg/L~0.5–1 g O3/g DOC
Iron & manganese0.4–1 mg/L~0.43 mg O3/mg Fe; 0.88 mg O3/mg Mn
Micropollutants (pharma, EDCs)0.5–1 g O3/g DOCThen biofiltration polish

The key by-product risk is bromate (BrO3−). In bromide-bearing waters ozone oxidises Br− via hypobromite and radical routes to bromate, a regulated carcinogen (WHO / EU / UK limit 10 µg/L). Bromate rises with dose, contact time, pH and bromide, so control means minimising over-dosing, depressing pH, or adding a trace of ammonia to sequester the intermediate. Finally, the ozone-rich off-gas (typically 5–15% of applied ozone) is toxic and must pass through a thermal or catalytic off-gas destructor before venting, cutting the stack concentration to well below the ~0.1 ppmv occupational limit.

Design, materials and operational considerations

Ozone is aggressive to most elastomers and many metals, so wetted and gas-side materials are restricted to 316L stainless steel, PTFE, PVDF, FEP, Viton and concrete; ordinary rubbers, PVC and mild steel degrade rapidly. Contactors are usually deep (5–7 m) counter-current baffled tanks or sealed columns, both to raise the mass-transfer driving pressure and to lengthen the T10 contact time that CT credit depends on.

Because ozone decays, the dissolved residual is measured along the contactor (often at several cells) and the generator is trimmed by feedback to hold the target residual as flow, temperature and demand vary. Feed-gas dewpoint, cooling-water temperature and cell power all move the achievable output, so a robust design includes an oxygen or dried-air margin, standby generator capacity, and ambient ozone monitors with automatic shutdown. Whole-life running cost is dominated by electricity and, for oxygen plants, the oxygen supply — both of which scale directly with the applied dose, reinforcing the case for demand-based dosing rather than a fixed set-point.

How to size an ozonation system

  1. Define the treatment objective. Decide whether the duty is disinfection (CT-driven), oxidation of a named contaminant, or both, and set the target log-removal or effluent quality.
  2. Fix the transferred dose. Select the transferred ozone dose from bench ozone-demand tests and published dose ranges for the duty (e.g. 1-5 mg/L), allowing for the water's ozone demand and decay.
  3. Compute the ozone mass and generator size. Applied mass = Q x dose / transfer efficiency; generator power = applied mass / specific yield. Choose air or oxygen feed and add output margin.
  4. Size the contactor for CT. Set residual and T10 to meet the required CT, apply the baffling factor to get true HRT, and compute contact volume from the design flow.
  5. Check bromate and by-products. For bromide-bearing water, verify predicted bromate stays under 10 ug/L and add pH or ammonia control if needed.
  6. Add off-gas destruction and monitoring. Specify a thermal or catalytic ozone destructor plus ambient and dissolved-ozone monitoring and interlocks.

Frequently asked questions

Is ozonation better than chlorination?

Ozone is a stronger, faster oxidant than chlorine and, crucially, inactivates Cryptosporidium, which chlorine cannot practically achieve. It also avoids chlorinated disinfection by-products. However, ozone leaves no lasting residual for distribution, is generated on site at higher capital cost, and can form bromate. Many plants use ozone for primary treatment and a small chlorine dose for a residual.

Why must ozone be generated on site?

Ozone is thermodynamically unstable and decomposes back to oxygen within minutes to hours, faster when warm, so it cannot be stored, bottled or transported. Every ozonation plant makes ozone continuously from air or oxygen by corona discharge and injects it immediately, sizing the generator to the instantaneous applied-dose demand rather than to any inventory.

What is the difference between applied and transferred ozone dose?

Applied dose is the ozone mass fed to the contactor per unit water flow. Transferred dose is the fraction that actually dissolves, equal to applied dose times the transfer efficiency (usually 85-95%). The untransferred remainder leaves in the off-gas. Disinfection CT and oxidation are driven by the transferred (and ultimately absorbed) dose, not the applied figure.

How does pH affect ozonation?

Hydroxide ions initiate ozone decomposition into hydroxyl radicals, so higher pH speeds the shift from selective molecular-ozone chemistry to fast, non-selective radical oxidation. Alkaline conditions favour micropollutant destruction but shorten the ozone residual and raise bromate formation. Lower pH stabilises molecular ozone, helping build a disinfection residual and suppressing bromate.

What is bromate and why does it matter?

Bromate is an oxidation by-product formed when ozone reacts with bromide naturally present in some source waters. It is a suspected carcinogen regulated at 10 ug/L in the UK, EU and WHO guidelines. Because formation increases with ozone dose, contact time and pH, designers control it by avoiding over-dosing, depressing pH, or dosing trace ammonia to interrupt the reaction pathway.

What ozone dose is typical in water treatment?

Transferred doses generally fall between about 0.5 and 5 mg/L. Disinfection is set by the required CT (residual times contact time) rather than a fixed dose; taste, odour and colour removal often need 1-5 mg/L; iron and manganese oxidation need under 1 mg/L. Bench ozone-demand testing on the actual water is the reliable basis for the design dose.

Sources & further reading