How to Read a UV Product Datasheet

Quick answer

A UV product datasheet packs three different "power" ideas that buyers routinely confuse: electrical power (watts the device draws from the wall), irradiance (UV power landing on a unit of surface, in W/m² or mW/cm²), and dose / fluence (the irradiance multiplied by exposure time, in mJ/cm² or J/m²). Only dose tells you whether a pathogen is inactivated or a coating is cured — and dose is the number datasheets are most likely to not state, because it depends on how you operate the device.

The single most useful conversion to memorise: 1 mW/cm² = 10 W/m², and likewise 1 mJ/cm² = 10 J/m² (IUVA UV FAQs). A worked example from the same source: a fluence of 20 mJ/cm² is reached in 20 seconds at 1 mW/cm² (10 W/m²), but takes 200 seconds at 0.1 mW/cm² (1 W/m²). Same dose, very different irradiance and time. A datasheet that advertises a big "watt" number tells you almost nothing on its own.

This guide walks through the quantities, the wavelength specs, the lifetime claims, the hidden measurement conditions, and the common ways a figure can mislead a buyer.

1. The three "powers" — and why mixing them up costs money

1.1 Electrical power (W)

This is what the device consumes. It is an absolute number, independent of wavelength and time (Phoseon, Understanding UV Output). A "150 W UV lamp" describes the electrical draw — not the UV delivered. Two lamps with identical wattage can differ widely in how much of that energy leaves as useful UV-C, because wall-plug-to-UV efficiency varies by lamp technology. Never treat electrical watts as a performance figure.

1.2 Irradiance (W/m² or mW/cm²)

Irradiance is the radiant power per unit area arriving at a surface. It depends on the source power, the distance, and the dispersion of the beam (Phoseon). The crucial property: irradiance is a here-and-now value. It says how hard the UV is hitting a surface at a given instant, at a given distance — nothing about cumulative effect.

Without optics, irradiance falls as you move away from the lamp; it is also lower the further the target surface sits from the source (Waveform Lighting). So an irradiance figure on a datasheet is meaningless unless the distance at which it was measured is also stated.

1.3 Dose / fluence (mJ/cm² or J/m²)

Dose — the technically correct term is fluence — is irradiance × exposure time (IUVA UV FAQs). It is the quantity that actually determines outcomes: log-reduction of a microorganism, or the energy delivered to cure an adhesive. North America commonly writes mJ/cm²; most of the rest of the world writes J/m² (IUVA UV FAQs).

Because dose = irradiance × time, a low-irradiance device can still deliver a high dose if it runs long enough — and a high-irradiance device delivers nothing useful in a flash exposure that is too short. A datasheet that quotes a dose must also quote the irradiance and the exposure time it assumed, or the number is a marketing figure, not an engineering one.

Quantity comparison table

Conversions: 1 mW/cm² = 10 W/m²; 1 mJ/cm² = 10 J/m² (IUVA UV FAQs).

2. Reading the wavelength spec

UV outcomes are wavelength-dependent — germicidal action peaks near 254–265 nm, curing chemistry responds at different bands — so the spectral spec deserves close reading. Datasheets typically list several distinct numbers (Crystal IS, How to Interpret UVC LED Data Sheet Values).

• Peak wavelength — the wavelength at which emission is strongest. This is the headline number, but a source emits over a band, not a single line.
• Nominal wavelength — a rounded label for a product family (e.g. a "275 nm LED"). The actual measured peak may sit a few nm away. For LEDs, the datasheet usually gives a binning tolerance for peak wavelength: parts are sorted into wavelength bins, and the bin width is the real spread you should design around (Crystal IS).
• FWHM (full width at half maximum) — the width of the emission curve measured at half its peak height; it is the standard measure of spectral width / bandwidth (Ossila, Spectral Resolution). A narrow FWHM means most energy sits close to the peak; a wide FWHM means a meaningful fraction lands at wavelengths far from peak, which may be less effective for the target application.

Industrial UV-lamp standards for non-destructive testing require the emission analysis to report peak wavelength, FWHM, longest-wavelength-at- half-maximum, and the irradiance in the excitation band — and to do so both at ambient temperature and at the maximum rated operating temperature (Magnaflux, ASTM E2297 & E3022 guide). That last point matters: a wavelength figure measured cold is not the figure you get in service.

Buyer takeaway: treat the nominal wavelength as a label, the peak as the design centre, and the FWHM (or binning spread) as the real uncertainty you must engineer for.

3. Reading the lifetime claim

"Lifetime" on a UV datasheet is rarely the moment the device stops emitting. It is almost always a point at which output has decayed to a defined fraction of the initial value.

3.1 LED lifetime: L70, L50 and the testing chain

For LED-based products the dominant convention is borrowed from general LED lighting. L70 means the source still produces 70 % of its initial output at the rated hours; L50 means 50 %; L90 means 90 % (Fireflier, Lighting Basics: L70, TM-21, LM-80). Two products both quoting "30,000 hours" are not equal: an L50 of 30,000 h decays faster than an L70 of 30,000 h (Fireflier). Always read the L-number, not just the hours.

Behind those numbers sits a two-standard chain:

• IES LM-80 is the measurement standard. Labs measure output at regular intervals (typically every 1,000 h) over a test period of roughly 6,000–10,000 h (Light Laboratory, TM-21 Testing).
• IES TM-21 is the projection method that extrapolates LM-80 data to a lifetime estimate. Critically, TM-21 only allows projection to six times the actual test duration: a 6,000 h LM-80 test supports a claim up to ~36,000 h; a 10,000 h test supports up to ~60,000 h (Focal Point, Understanding TM-21).

So a datasheet line such as "L70 (6K) >36,000 h" tells you the test was only 6,000 h long and the rest is extrapolation. A bare "60,000 hours" with no L-number and no LM-80/TM-21 reference is an unverifiable claim. Note also that LM-80/TM-21 were written for visible-light lumen maintenance; applying them to UV output is an industry adaptation, and a careful datasheet will state what it actually measured.

3.2 Lamp lifetime: rated hours and end-of-life output

For discharge lamps (low-pressure mercury / amalgam UV-C), the datasheet usually states a rated life in hours plus an end-of-life output percentage. ASHRAE's UV systems chapter notes that UV-C lamps at the end of useful life can emit anywhere from 50 % to 85 % or more of the output measured after the initial 100 h burn-in (ASHRAE Handbook, Ch. 17 Ultraviolet Lamp Systems). Higher-grade lamp construction reduces depreciation (Light Sources, Low Pressure Amalgam Lamps).

The engineering consequence is decisive: a UV-C system should be sized for output at end of life, not for the as-new figure (ASHRAE Handbook, Ch. 17). If a datasheet quotes a glowing irradiance number but is silent on end-of-life output, you cannot tell whether the system still meets dose at the moment a replacement is actually due.

4. The conditions every number hides behind

Every performance figure on a UV datasheet is true only under the conditions it was measured. The honest datasheets state those conditions; the misleading ones omit them.

• Distance. Irradiance falls with distance from the source (Waveform Lighting). An irradiance figure without a stated measurement distance — and the measurement geometry — is not usable. Standardised lamp-output methods such as BS ISO 15727 exist precisely to fix the geometry; the Keitz formula used there takes measured irradiance, distance, lamp arc length and subtended angle as explicit inputs (Test Labs, BS ISO 15727).
• Temperature. Output is temperature-sensitive. For UV-C LEDs, output power is very sensitive to junction temperature, which depends on ambient temperature and on the thermal design around the part; a typical datasheet states a thermal derating figure (one Crystal IS example specifies 0.5 % output loss per Kelvin) and defines the measurement drive current (Crystal IS, How to Interpret UVC LED Data Sheet Values). For discharge lamps, actual irradiance varies with lamp type, ballast, temperature, ageing and quartz soiling (UV Resources, Lamp Power Tool).
• New vs aged. Datasheet figures are usually taken on a fresh part after a short burn-in. Real-world output is lower because of the depreciation discussed in section 3. Ask whether a quoted number is initial or end-of-life.
• Drive current (LEDs). The number of photons emitted scales with the drive current — and so does the heat generated, which then degrades output via junction temperature (Crystal IS, Introduction to UVC LEDs). A radiant-flux figure is only comparable between products if the drive current is the same.

If two datasheets do not state the same distance, temperature, age and (for LEDs) drive current, their numbers cannot be compared directly.

5. Common ways a UV datasheet can mislead a buyer

The UV market still lacks a universal performance-labelling standard, which leaves room for vague or unverifiable claims. The IUVA explicitly warns that some vendors make vague, false or misleading claims without verifiable specifications or supporting test results, and that the absence of a universal performance standard makes such claims hard to police (UV Solutions Magazine, The Need for Consistent Germicidal UV Labeling). A peer-reviewed evaluation of consumer "UV-C sanitiser" devices found that not all of them actually deliver effective UV-C output (medRxiv, Beware of UVC sanitizers: not all are good).

Watch for these specific patterns:

1. Electrical watts sold as UV performance — see section 1.1. A high wattage is a power bill, not a UV figure.
2. Irradiance with no distance — a big mW/cm² number measured 1 cm from the lamp window collapses at working distance.
3. Peak quoted as if it were average — peak irradiance at the centre of a beam is not the average over a treated area; treated-area dose is what matters.
4. Dose with no assumed irradiance or time — a "delivers 40 mJ/cm²" claim is incomplete without the exposure time and geometry it assumes.
5. Lifetime hours with no L-number or LM-80/TM-21 basis — see section 3.1; an unqualified "60,000 hours" is unverifiable.
6. Initial output presented as service output — no end-of-life figure means you cannot size for a worn lamp (section 3.2).
7. Kill-rate / log-reduction claims with no test conditions — IUVA's guidance is blunt: if something seems too good to be true, it probably is, and buyers should ask manufacturers for verifiable specifications and third-party test results (IUVA, Enhancing the New Normalcy with UV Disinfection).

The buyer's defence: demand that every figure come with its conditions, convert everything to a common unit, and judge the device on dose delivered at working distance, at end of life — not on the biggest headline number.

Cross-references

• UV Dosimetry Fundamentals — the dose / fluence concept in depth.
• Wavelengths and Action Spectra — why the wavelength spec changes the outcome.
• UV-LED Lifetime and Degradation — the physics behind L70 / L50 decay.
• UV Lamp Technology — discharge-lamp types and their depreciation behaviour.
• Lambert's Cosine Law — why angle, not just distance, changes the irradiance you read.
• Layer Thickness and Dose Scaling — translating a surface dose into a real process.
• Choosing a UV System — putting datasheet reading into a full purchase decision.
• LED vs Mercury Decision Guide — how lifetime and output specs differ between the two technologies.
• A buyer's checklist template for requesting datasheet conditions from vendors is (coming).

Sources

• IUVA — UV FAQs — definitions and unit conversions for irradiance vs dose / fluence; the 20 mJ/cm² worked example.
• IUVA — Enhancing the New Normalcy with UV Disinfection — buyer guidance and the "too good to be true" warning.
• Phoseon Technology — Understanding UV Output (Technical Note) — electrical power vs irradiance vs dependence on distance and dispersion.
• Crystal IS — How to Interpret UVC LED Data Sheet Values and Figures — peak / nominal wavelength, binning, junction temperature, drive current.
• Crystal IS — Introduction to UVC LEDs: Fundamental Characteristics — drive current vs photon output and heat.
• Magnaflux — A Guide to NDT UV Lamp Standards (ASTM E2297 & E3022) — spectral parameters required by industrial UV-lamp standards.
• Ossila — Spectral Resolution / FWHM — FWHM definition.
• Fireflier — Lighting Basics: L70, TM-21, LM-80 — L70 / L50 / L90 definitions.
• Light Laboratory — TM-21 Testing — LM-80 measurement intervals and test durations.
• Focal Point — Understanding TM-21 in Relation to Specifying Luminaires — the TM-21 six-times-test-duration projection cap.
• ASHRAE Handbook — Chapter 17, Ultraviolet Lamp Systems — end-of-life output range and sizing for end of life.
• Light Sources — Low Pressure Amalgam Lamps — lamp construction and depreciation.
• Waveform Lighting — How to Determine UV-C Irradiance Requirements — irradiance dependence on distance.
• UV Resources — Lamp Power Tool — factors that make actual irradiance vary (ballast, temperature, ageing, soiling).
• Test Labs — BS ISO 15727: Measuring the Output of a UV-C Lamp — standardised geometry and the Keitz formula inputs.
• UV Solutions Magazine — The Need for Consistent Germicidal UV Labeling — the absence of a universal performance label and vague-claim problem.
• medRxiv — Beware of UVC sanitizers: not all are good — peer-reviewed finding that some consumer UV-C devices underperform.