UV/H2O2 Advanced Oxidation Dose: The Code Gives Ranges, Not a Formula

1. What This Document Is, and What It Is Not

The only written technical document currently available that is dedicated to UV/H2O2 advanced oxidation is the group standard Code of Practice for Urban Water Supply Advanced Treatment by UV/H2O2 Advanced Oxidation, issued for public comment on 10 July 2025 by the Liaoning Urban Water Supply and Drainage Association (draft for comment, number not yet assigned, cover marked T/LNUWA xxx-2025). Three boundaries must be stated before any number is quoted.

• It is a draft for comment, not a formally issued text. Clauses may still change before approval and it cannot be cited as a mandatory provision.
• Scope is drinking water only. Clause 1 limits application to UV/H2O2 advanced oxidation in new or reconstructed municipal water supply and drinking water treatment. Industrial wastewater and water reuse are outside its scope.
• It delegates most equipment performance parameters to GB/T 19837, Ultraviolet (UV) Disinfection Equipment for Urban Water and Wastewater Engineering (clause 5.1.3). GB/T 19837-2019 was issued and took effect on 10 December 2019, fully replacing GB/T 19837-2005, and was confirmed as continuing valid at the review of 4 August 2025.

2. Dose and Dosage: The Document Gives Ranges, Not a Formula

This is the point most often misunderstood. The draft provides no formula that computes UV dose from a target removal rate. It gives design ranges only.

2.1 UV dose

Clause 5.1.10, item 2: the UV dose of the UV AOP reactor should be 400 mJ/cm2 to 800 mJ/cm2. Item 1 of the same clause sets the design velocity of the piping system at 1.2 m/s to 1.6 m/s, and item 3 requires sufficient clearance around the reactor, with side clearance reserved for lamp replacement and maintenance.

2.2 Hydrogen peroxide dosage

Clause 5.2.3, item 3: the H2O2 dosage required to reach one log removal of the target pollutant (that is, 90 percent removal) is typically 3 mg/L to 20 mg/L, and the text explicitly states that the actual dosage must be determined by testing against source water quality. In other words, 20 mg/L is an order-of-magnitude reference ceiling, not a guaranteed design value.

2.3 Engineering implication

Read together, the two clauses mean that "dose calculation" for UV/H2O2 in currently available written documents is an empirical range plus test verification, not a derivable design equation. Presenting 400 to 800 mJ/cm2 as the output of some formula is incorrect.

3. Three Water Quality Gates That Decide Whether the Dose Can Be Delivered

• UV transmittance. Clause 4.2.4 requires that a re-filtration unit preferably be installed and that UVT at the reactor inlet be greater than 80 percent. When UVT falls short, raising lamp power is not an equivalent substitute, and the draft offers no correction method for UVT below 80 percent.
• Bromide. Clause 4.2.1 states that the process applies to removal or control of trace pollutants such as taste and odour compounds and emerging contaminants, and that it should be preferred when bromide in the raw water exceeds 100 ug/L. The substance of this clause is a comparison with ozonation: at high bromide, the ozone route carries bromate risk that UV/H2O2 avoids.
• Process sequencing. Clause 4.2.3 requires that the upstream and downstream position relative to existing and new processes be fixed only after testing and a combined technical and economic comparison. Clause 4.2.2 requires reactor module count, H2O2 tank volume and lamp count to be set from source water quality and quantity.

4. Hydraulic Conditions for Dosing and Mixing

Clause 5.2.3 item 4 and clause 5.1.7 together define mixing: mixing should be done through a pipeline mixer; the straight pipe run from the H2O2 dosing point to the UV reactor should not be less than 5 pipe diameters, and a straight run of not less than 3 pipe diameters should be provided after the reactor; a pipeline mixer should also be installed on the piping upstream of the reactor. These two diameter multiples are hard numbers that can be written into design documents and checked on site.

For the reactor itself, clause 5.1.2 recommends a cross-flow tubular reactor, which may use high-power low-pressure high-intensity UV lamps; clause 5.1.5 recommends an internal wall coating to reduce UV absorption by the wall; clause 5.1.6 recommends a baffle at the inlet to reduce dead and recirculation zones.

5. Residual H2O2 Must Be Quenched, and the Parameters Are Quantified

Clause set 5.3 is the most fully parameterized part of the draft and the part most often omitted in design.

• 5.3.1: a quenching unit to destroy residual H2O2 shall be installed downstream of the process.
• 5.3.2: a new granular activated carbon filter should be used; where new construction is impossible, an existing sand filter may be converted to a carbon-sand filter.
• 5.3.3: where the carbon filter follows sand filtration and no further turbidity removal follows, downflow (gravity down-flow) configuration is preferred.
• 5.3.4: coal-based granular activated carbon shall be used, with particle size and size distribution complying with CJ/T 345; 8 mesh x 30 mesh or a test-determined specification may be selected.
• 5.3.5: empty bed contact time should be 6 min to 20 min, empty bed filtration velocity 8 m/h to 20 m/h, and carbon bed depth 1.0 m to 2.5 m.
• 5.3.7: H2O2 shall not be detectable in the effluent of the quenching unit.

Clause 5.4 further requires the control centre to modulate lamp output power and H2O2 dosage in real time against source water quality, flow and effluent targets, and to provide accumulated lamp operating hours, over-temperature protection, low water level protection, cleaning fault alarms and lamp fault alarms.

6. Verifiable Equipment and O&M Criteria

• 5.1.1: the reactor shall be equipped with a UV intensity sensor and an on-line mechanical cleaning system; an on-line UVT analyser may be added.
• 5.1.9: outdoor electrical equipment shall have a degree of protection not lower than IP65; indoor equipment not lower than IP54 (per GB/T 4208).
• 6.1.2: the reactor shall be filled with water and preheated before commissioning.
• 6.1.5: lamps shall be replaced when service life is reached or when UV intensity falls below 85 percent of the initial value.
• 6.1.6: quartz sleeves shall be replaced when transmittance after manual cleaning falls below 80 percent of the initial value, or when cracked.
• 6.1.11: for shutdown maintenance, switch off lamps first, then slowly close the inlet valve, then shut down the reaction unit, the H2O2 dosing system and the main power supply, and only work on the equipment after lamps have cooled and the unit has fully stopped.
• 6.2.2 to 6.2.5: the H2O2 storage area shall carry hazardous chemical safety notice cards; anti-static work clothing, protective goggles, gloves and boots shall be provided; fire protection shall comply with fire regulations with adequate fire separation.

7. How Dose Is Actually Determined: Appendix A Is a Measurement Method

Appendix A of the draft (informative) defines a performance test procedure that is, in substance, back-calculating dose from measurement.

1. D.3: adjust flow to target; the target flow error at the start and end of sampling shall not exceed 5 percent.
2. D.5.1: run a steady-state test using a UVT modifier (coffee, lignosulfonate or humic acid); sample at intervals of one hydraulic retention time; the time from zero until the UVT ratio between inlet and outlet approaches 1 is the system stabilisation time.
3. D.5.2: with lamps off, run inlet and outlet mixing tests; the difference in average UV254 or target pollutant concentration between two sampling points shall be under 5 percent, and the error of three samples at the same point shall be under 5 percent.
4. D.6.1: sampling shall take place 10 min after the system reaches steady state; D.6.3: samples shall be stored at about 4 degrees Celsius; outlet samples must not be exposed to UV light.
5. D.7: gives the log removal L = lg(C0/C) and removal rate expressions, used to evaluate performance at a given oxidant dosage, UV dose and water quality.

In other words, "dose calculation" is implemented in engineering practice as follows: size the plant for 400 to 800 mJ/cm2 and 3 to 20 mg/L, then measure the log removal actually achieved for that water using Appendix A, and adjust. This is logically distinct from the dose-versus-inactivation curve used in disinfection and must not be conflated with it.

8. Gaps to Close Before Writing Design Documents

• How to handle UVT below 80 percent is not specified - it must be resolved by pretreatment testing.
• The draft normatively references GB 5749, Standards for Drinking Water Quality, but its body gives no control limits or conversion method for by-products such as bromate or formaldehyde; GB 5749 must be applied in full and the relevant parameters listed separately.
• The draft does not apply to wastewater; industrial wastewater and water reuse projects may only treat the above ranges as reference, not as acceptance criteria.
• The draft provides no energy consumption indicator; power demand and unit water cost must come from pilot testing or comparable operating plants.

Sources: clauses 4.2.1, 4.2.3, 4.2.4, 5.1.2, 5.1.7, 5.1.9, 5.1.10, 5.2.3, 5.3.1 to 5.3.7, 6.1.2, 6.1.5, 6.1.6 and 6.2, plus Appendix A, of the draft Code of Practice for Urban Water Supply Advanced Treatment by UV/H2O2 Advanced Oxidation; standard information page of GB/T 19837-2019.