Hydroxyl radicals react with almost every organic pollutant at near diffusion-controlled rates, which makes advanced oxidation look universally applicable. It is not. Efficiency is decided by what else in the water is competing for those radicals — and in wastewater, the competition wins by orders of magnitude.
Rate constants are the easy part
Degradation by hydroxyl radical is second order in radical and pollutant:
−d[C]/dt = k•OH,C · [•OH] · [C]
| Pollutant | k•OH (M−1 s−1) |
|---|---|
| Carbamazepine | 8.8 × 109 |
| Atrazine | 3.0 × 109 |
| 1,4-Dioxane | 2.8 × 109 |
| Acetone | 1.1 × 109 |
These cluster within a single order of magnitude near the diffusion limit, which is precisely why the rate constant is rarely the interesting variable. Selecting an AOP on the basis that your pollutant has a high rate constant is close to meaningless — almost all of them do.
The steady-state radical concentration is what varies
In a flow reactor, radical concentration settles where generation balances consumption:
[•OH]ss = Rgen / ∑(ki[Si])Si = each scavenging species present
For UV/H₂O₂, generation depends on photon flux and peroxide absorbance:
Rgen = 2Φ · I0 · [1 − 10−ε[H2O2]l] / lΦ ≈ 0.5 mol/Einstein; ε = 18.6 M−1 cm−1 at 254 nm; l = path length
The factor of two reflects two radicals per photon absorbed. Note that peroxide is both the radical source and, at high concentration, a scavenger — there is an optimum dose, not a monotonic benefit.
Scavenging: the number that decides viability
Every constituent of the water competes for radicals. Summing the products of rate constant and concentration gives the total scavenging rate.
| Scavenger | k•OH | Typical concentration | Contribution |
|---|---|---|---|
| Dissolved organic matter | ≈2.5 × 104 per mg C/L | 10 mg C/L | ≈2.5 × 105 s−1 |
| Nitrite | 5.0 × 109 | 0.02 mM | ≈1.0 × 105 s−1 |
| Bicarbonate | 8.5 × 106 | 3.3 mM (200 mg/L) | ≈2.8 × 104 s−1 |
| Carbonate | 3.9 × 108 | pH dependent | Rises sharply above pH 9 |
| Total for a typical UK effluent | ≈4 × 105 s−1 | ||
The consequence for radical concentration is severe. Against a generation rate of the order of 10−7 M/s, a scavenging rate of 4 × 105 s−1 gives a steady-state radical concentration in the region of 10−13 M — two to three orders below what the same reactor achieves in clean water.
Reactor sizing and energy
With a steady-state radical concentration established, the reaction is pseudo-first-order in pollutant:
τ90 = ln(10) / (k•OH,C · [•OH]ss)Residence time for 90% removal — one log order
Electrical energy per order
The industry metric is the electrical energy required to remove one log order from a cubic metre:
EE/O = P / (Q · log(C0/Ct))P = lamp power (kW); Q = flow (m³/h); result in kWh per m³ per order
EE/O = 10 / (50 × 1) = 0.20 kWh per m³ per order
Published values for wastewater AOP typically fall in the range 0.5–5 kWh/m³ per order, with low-carbon groundwater at the bottom of that band and high-DOM effluent at the top. A figure in the hundreds indicates a units error — most often watts confused with kilowatts, or a stray factor of 1,000 — not a real process.
Design consequences
Everything above points one way: treat the water matrix as the primary design input, not the pollutant.
- Measure DOC, alkalinity and nitrite before sizing. These set the scavenging rate and therefore the reactor. A design transferred from another site on pollutant identity alone will be wrong.
- Consider pre-treatment instead of a larger reactor. Because DOM dominates scavenging, removing organic carbon upstream — coagulation, or granular carbon — raises radical availability for every subsequent cubic metre. Halving DOC roughly doubles the radical concentration, which is usually cheaper than doubling lamp power.
- Optimise peroxide dose, do not maximise it. Peroxide generates radicals and also scavenges them.
- Watch pH. Carbonate scavenges roughly fifty times faster than bicarbonate, so alkalinity that is benign at pH 7.5 becomes a serious sink above pH 9.
- Never design from clean-water kinetics. The rate constants transfer; the radical concentration does not.
Related: advanced oxidation processes, UV system design, ozonation by-products.
Frequently asked questions
What is a realistic EE/O for wastewater advanced oxidation?
Broadly 0.5–5 kWh per cubic metre per log order, with low-carbon groundwater at the bottom of that range and high-DOM secondary effluent at the top. A 10 kW system treating 50 m³/h at one log order works out at 0.2 kWh/m³ per order. Figures in the tens or hundreds almost always indicate a units error rather than a real process.
Why does the same AOP work on groundwater but not on effluent?
Scavenging. Dissolved organic matter consumes hydroxyl radicals faster than anything else present, contributing roughly two thirds of total scavenging at 10 mg C/L. That drops the steady-state radical concentration by two to three orders of magnitude relative to clean water, and residence time rises in proportion.
Is it better to increase lamp power or reduce DOC?
Usually reduce DOC. Radical concentration is inversely proportional to the scavenging rate, and organic matter dominates that sum, so halving DOC roughly doubles radical availability for every cubic metre treated. Coagulation or carbon pre-treatment is generally cheaper than the equivalent increase in installed lamp power.
Does more hydrogen peroxide always give faster oxidation?
No. Peroxide is both the radical source and, at elevated concentration, a radical scavenger in its own right. There is an optimum dose for a given photon flux and water matrix; beyond it, additional peroxide reduces the net radical concentration.