Quick answer
UV dose (also called fluence) is the total germicidal UV energy delivered to a target, and it is the single number that determines whether a microorganism is inactivated. It is calculated as dose = irradiance × exposure time, expressed in millijoules per square centimetre (mJ/cm²): UV dose (mJ/cm²) = UV irradiance (mW/cm²) × time (s).
The dose required is pathogen-specific, but most drinking-water regulators converge on a design fluence of about 40 mJ/cm² to assure at least 4-log (99.99 %) inactivation of pathogenic microorganisms (IUVA UV FAQs). Getting that number right depends on three things that are easy to underestimate: measuring the right quantity (irradiance vs. fluence rate), placing and calibrating the sensor correctly, and — for a real reactor — proving the delivered dose by biodosimetry rather than by a single reading. The deep sections below cover each.
1. Dose, irradiance and fluence rate — three quantities, not one
1.1 The dose equation
For a static exposure (a fixed sample under a fixed lamp), UV dose is simply the product of intensity and time:
Dose (mJ/cm²) = Irradiance (mW/cm²) × Time (s)
Irradiance is the radiant power per unit area reaching a surface. It is reported in mW/cm² or in W/m²; the conversion is 1 mW/cm² = 10 W/m² (Ledrise UV fluence reference). Doubling either irradiance or time doubles the dose — the two are interchangeable for inactivation of most organisms, within the limits of dose-rate effects that fall outside this article.
1.2 Irradiance vs. fluence rate (spherical irradiance)
The subtle and frequently mishandled distinction is between irradiance and fluence rate:
- Irradiance is defined on a flat, directional surface. It counts photons according to the cosine of their incidence angle — a photon arriving edge-on contributes almost nothing. This is what a radiometer with a flat window measures.
- Fluence rate (also called spherical irradiance) is defined on the surface of a small sphere. It has no direction and no angle dependency: photons arriving from all directions are counted equally (USHIO — Irradiance vs. Fluence Rate, Holger Claus, 2023).
A microorganism suspended in air or water is a three-dimensional object exposed to photons from every direction, so fluence rate is the physically correct quantity for dose calculations (USHIO application note). The angle dependence of a flat surface is itself a foundational topic — see Lambert's cosine law.
The practical catch: there is no commercial fluence-rate meter on the market (USHIO application note). Practitioners measure irradiance with a radiometer and then apply geometric correction factors to estimate fluence rate.
| Quantity | Geometry | Angle dependence | Measured by | Right for |
|---|---|---|---|---|
| Irradiance | Flat surface | Cosine-weighted | Flat-window radiometer | Surface exposure, sensor readings |
| Fluence rate | Sphere (4π) | None — all angles equal | Spherical actinometer; no commercial meter | Dose to suspended cells/aerosols |
2. Measuring instruments
2.1 Radiometers and UV sensors
The photoelectric radiometer is the most common UV measurement tool: a probe with a small flat window acts as the photon receiver, so the result is irradiance (USHIO application note). A real germicidal sensor never has a spectral responsivity perfectly matched to the germicidal action spectrum, which is one of the main sources of measurement uncertainty (NIST — Calibration and Characterization of UV Sensors for Water Disinfection, Larason & Ohno, 2006). Why the spectral weighting matters is covered in Wavelengths and action spectra.
2.2 Chemical actinometry
A chemical actinometer is a light-sensitive solution whose measurable chemical change is proportional to absorbed photons. The iodide–iodate actinometer is uniquely suited to germicidal UV: it is optically opaque at 254 nm and its triiodide quantum yield is essentially constant across the germicidal band, Φ(I₃⁻) = 0.92 ± 0.02 between 205 and 245 nm (NIST — Quantum Yield of the Iodide–Iodate Chemical Actinometer). The ferrioxalate actinometer is also widely used but is wavelength-dependent in this region: Φ(FeII) ≈ 1.48 below 240 nm and ≈ 1.25 above 270 nm (Goldstein & Rabani — The ferrioxalate and iodide–iodate actinometers in the UV region).
Placed inside a small spherical quartz vessel, the iodide–iodate actinometer becomes a spherical actinometer — it absorbs photons from all directions and therefore measures fluence rate directly, which a flat radiometer cannot (USHIO application note).
2.3 Biodosimetry
Biodosimetry uses a calibrated test organism as the dosimeter. The organism's measured log-inactivation is compared against its known dose–response curve to back-calculate the dose actually delivered. It is the reference method for validating full-scale reactors (Section 4) and is also used alongside actinometry and radiometry in research to cross-check polychromatic fluence measurements (Linden et al. — Polychromatic UV Fluence Measurement Using Chemical Actinometry, Biodosimetry, and Mathematical Techniques).
3. Bench-scale dose determination — the collimated beam
For laboratory dose–response work, the collimated beam apparatus is the standard tool. The widely adopted Bolton–Linden protocol (2003) standardises both the apparatus and the calculation (Bolton & Linden — Protocol for the Determination of Fluence (UV Dose)).
A bare radiometer reading at the sample surface is not the average fluence the cells receive. For a monochromatic low-pressure lamp, four correction factors are applied (Bolton & Linden protocol):
| Correction factor | Accounts for |
|---|---|
| Petri factor | Non-uniformity of irradiance across the sample dish surface |
| Reflection factor | Reflection at the air–water interface (refractive-index change) |
| Divergence factor | Beam spreading between the sensor plane and mid-sample depth |
| Absorption factor | The vertical irradiance gradient as UV is absorbed through the sample depth |
The Petri factor quantifies how uniform the beam is over the dish. Reported values are typically >0.99 for low-pressure mercury exposures and 0.90–0.95 for UV-LED exposures, and best practice is that it should not fall below 0.90 (Bolton & Linden protocol). The LED gap exists because LED arrays are harder to collimate uniformly than a single tubular lamp.
How dose attenuates through depth (the absorption factor) ties directly to Layer thickness and dose scaling.
4. Verifying the dose in a real reactor — RED and validation
A bench collimated-beam dose does not translate directly to a flowing reactor. Water moves through on countless different paths; a fast-moving element near the wall gets far less dose than a slow one near the lamp. The result is a dose distribution, not a single dose.
To handle this, validated UV reactors report a Reduction Equivalent Dose (RED): the dose that, applied uniformly in a collimated-beam test, would produce the same log-inactivation of the test organism as the reactor delivered to the real, polydisperse flow (US EPA — Ultraviolet Disinfection Guidance Manual, 2006).
The RED is established by biodosimetry validation: the reactor is challenged with a test organism across its operating range of flow, UV transmittance and lamp power, and the measured inactivation is converted to RED. Because the RED depends on the organism and the conditions used, a validation factor is applied to account for uncertainties and dose-distribution effects. Under the US EPA UVDGM framework, validation supports guaranteed UV doses between 10 mJ/cm² RED and 120 mJ/cm² RED (US EPA UVDGM, 2006).
In Europe the equivalent role is played by DVGW W294 (Germany) and ÖNORM M5873 (Austria), which set reactor-validation and reference-radiometer requirements for drinking-water UV disinfection (sglux — UV-ÖNORM sensors). For reactor and water-treatment context see Drinking-water system types, and for the regulatory landscape Standards and certifications.
5. Sensor placement, calibration and drift
Once a reactor is validated, its duty UV sensor monitors ongoing dose in operation. Three practical points govern whether that monitoring stays trustworthy:
- Placement. The duty sensor is installed at a defined, reproducible point — typically through a pressurised, water-tight measuring window — so that the geometric relationship between sensor, lamp and water path matches the conditions under which the reactor was validated (sglux — UV-ÖNORM sensors). A sensor reading is only meaningful relative to the validated geometry.
- Calibration. Sensors for low-pressure (254 nm line) and medium-pressure mercury systems are calibrated for irradiance responsivity against a low-pressure mercury lamp under ÖNORM M5873-1/-2 and DVGW W294-3 (sglux — UV-ÖNORM sensors). Calibration of reference radiometers should be checked regularly and at least every 12 months (sglux — DVGW UV calibration).
- Drift. Any photodetector exposed to UV ages: its responsivity slowly changes, so a reading that looks stable can quietly become wrong. This is exactly why periodic re-calibration against a reference radiometer is mandatory rather than optional (sglux — DVGW UV calibration). Known practical problems with matching real sensor spectral responsivity to the required curve — especially for medium-pressure systems — add further uncertainty (NIST — Calibration and Characterization of UV Sensors, 2006).
A common field error is to treat a single radiometer reading as proof of dose. It is not: it is one input. The dose is only assured when sensor placement, calibration currency and the validated dose-distribution model all hold together.
Cross-references
- Wavelengths and action spectra — why germicidal sensors must be spectrally weighted.
- Lambert's cosine law — the angle dependence that separates irradiance from fluence rate.
- Layer thickness and dose scaling — how UV is absorbed through depth (the absorption factor).
- Drinking-water system types — where validated-dose reactors are used.
- Standards and certifications — DVGW, ÖNORM and EPA UVDGM in the wider regulatory picture.
- Far-UV-C 222 nm — a band where fluence-rate vs. irradiance distinctions are especially relevant.
- How to read a UV datasheet — interpreting the irradiance and dose figures manufacturers publish.
- Pathogen-specific dose–response curves (coming) — a dedicated table of required fluence per organism.
- Medium-pressure vs. low-pressure dosimetry (coming) — polychromatic dose weighting in detail.
Sources
- US EPA — Ultraviolet Disinfection Guidance Manual (UVDGM), 2006 — the primary US federal guidance on reactor validation, biodosimetry and Reduction Equivalent Dose.
- Bolton & Linden — Protocol for the Determination of Fluence (UV Dose) in Bench-Scale Collimated Beam Experiments — the standard collimated-beam method and its four correction factors.
- NIST — Quantum Yield of the Iodide–Iodate Chemical Actinometer — defines Φ(I₃⁻) = 0.92 ± 0.02 across the germicidal band.
- Goldstein & Rabani — The ferrioxalate and iodide–iodate actinometers in the UV region — wavelength dependence of actinometer quantum yields.
- NIST — Calibration and Characterization of UV Sensors for Water Disinfection (Larason & Ohno, 2006) — sensor spectral-responsivity and calibration uncertainty.
- USHIO — An Explanation of Irradiance vs. Fluence Rate (Holger Claus, 2023) — manufacturer technical note on the irradiance/fluence-rate distinction and the absence of commercial fluence-rate meters.
- Linden et al. — Polychromatic UV Fluence Measurement Using Chemical Actinometry, Biodosimetry, and Mathematical Techniques — cross-method fluence verification.
- IUVA — UV FAQs — practitioner reference for the ~40 mJ/cm² 4-log design dose.