Risk group determination characterizes photobiological safety in LED lighting (MAGAZINE)

Expanding on a prior series of articles focused on the photobiological safety of LED-based lighting, LESLIE LYONS further details risk group classifications and methodologies that come into play with extended sources such as LED arrays.

Although LEDs bring many benefits to the world of general illumination, solid-state lighting (SSL) manufacturers still need to ensure that products present no photobiological hazards to humans.

In the December 2013 edition of LEDs Magazine, an overview was given of a new approach to the evaluation of the photobiological safety of light sources intended for lighting applications. In this article, we will take a wider view of the IEC TR 62778 document and discuss in depth the challenge of the correct treatment of extended sources such as LED arrays and linear LED light engines.

Interested in more articles & announcements on standards & testing?

IEC TR 62778, "Application of IEC 62471 for the assessment of blue light hazard to light sources and luminaires," provides guidance on the evaluation of the retinal blue-light hazard of sources of light intended for lighting applications. This assessment is based on determining whether or not a source presents, at a distance of 200 mm, a retinal blue-light hazard in excess of risk group one, as defined by the horizontal photobiological safety standard, IEC 62471. Currently presented as a Technical Report (TR), it is implemented through new editions of normative IEC lamp and luminaire safety standards. At the time of writing, the majority of these updates have been adopted in Europe by CENELEC as EN standards, and harmonized to the EU low voltage directive, 2006/95/EC.

While IEC TR 62778 was first published in 2012, the move away from implementing IEC 62471 for lighting applications has been slow, but appears to be gaining momentum. The complex procedure of evaluating sources to IEC 62471 will be replaced by a rather simpler approach but with significant complications presented in the treatment of extended sources. This latter case is, to a certain extent, addressed by IEC TR 62778 Edition 2: 2014.

A significant motivation in the writing of this TR document was the reduction of the measurement burden for luminaire manufacturers. This is achieved in two manners: first, by providing conditions under which the risk group (RG) classification of a primary light source may be transferred to a luminaire; and second, in presenting a choice of assessment methodologies, two of which are based on commonly available data. This TR should therefore be considered from two distinct points of view: that of the primary light source (Table 1) and that of the luminaire (Table 2).

One TR, three assessment techniques

Three techniques are proposed for the assessment of the blue light hazard, an overview of which is presented, in order of required inputs as defined in Table 3.

Method A: A manufacturer can report a table of illuminance values, as a function of CCT (≤8000K), below which RG1 will result. Consulting the table for a source of known CCT, one can adopt the reported illuminance value as E_thr (threshold illuminance). This value may be reported in the datasheet of a primary light source and converted to d_thr (distance threshold) for a luminaire. Where the latter process yields d_thr ≤200 mm, below the assessment distance, then RG1 should be reported. This method includes a safety factor of two and cannot produce a transferable risk group classification.

Method B: A manufacturer can again report a table of luminance values, as a function of CCT (≤8000K), below which RG1 will result. In addition to knowledge of the CCT of the source, a measurement of luminance (cd/m²) is required. The field of view (FOV) of measurement employed in determining luminance should not extend beyond the luminous area of the source, as discussed in the following section. Where the measured luminance of a primary light source is below that reported in the table, RG1 unlimited applies while for luminaires RG1 applies. This method includes a safety factor of two. Where the measured luminance exceeds the tabulated values, one should consider methods A or C.

Method C: The implementation of this technique was discussed in detail in the December 2013 article. Representing the direct spectroradiometric assessment of blue light hazard, this route will in all cases yield the most accurate result.

A fourth option

In the future, a fourth method will be proposed, based on the calculation of the luminance in an 11-mrad FOV from luminance distribution data of a source. A spectral measurement of the same source allows determination of K_B,v, a parameter defined in the TR as the ratio of the blue-light hazard quantity to the photometric quantity. The combination of these two will result in a blue-light radiance value to be compared with the RG1 limit.

Since the assessment method is not reported alongside the result, the impact of transferring data from a primary light source, obtained through methods A or B, to a luminaire, should be considered. The safety factor of two can be quite a high price to pay for simplicity.

FIG. 1. A white phosphor-converted LED chip is composed of a single, small blue LED emitter behind a larger phosphor layer. The spectral radiance measurement in an 11-mrad FOV, centered on the blue chip, will yield different values of CCT and KB,v to the total spectral radiant-flux measurement of the same source.

Finally, given that CCT and K_B,v are typically computed from a total spectral radiant-flux measurement of the source, one should consider whether or not the resulting spectral distribution be representative of that obtained in the blue light radiance measurement. In some instances, such as that highlighted in Fig. 1, the spectral distribution obtained in the radiance measurement would have a significantly higher blue peak and therefore different values of CCT and K_B,v than that obtained in the total flux measurement, encompassing emission from the entire source. The safety factor of methods A and B ought to cover this eventuality but the determination of K_B,v does not.

Conditions for the transfer of risk group classification

A prerequisite for the transfer of a blue-light hazard classification, relying upon the law of conservation of radiance, is that the measurement be performed over a circular measurement FOV that does not extend beyond the luminous area of the source. This measurement is sometimes called "true" radiance to distinguish from the measurement of "physiological" radiance considered in the photobiological safety assessment. Fig. 2 depicts the two measurement conditions.

FIG. 2. Conditions differ for measurement of "true" (upper) and "physiological" (lower) radiance.

According to Method C, where a source has a diameter <2.2 mm, spectral irradiance should be measured to report only E_thr, whereas for sources with a diameter >2.2 mm, spectral radiance should be measured, for which RG0 unlimited, RG1 unlimited, or E_thr can result. This need comes from the fact that at 200 mm, the required 11-mrad FOV encompasses a circular area of 2.2-mm diameter. To avoid confusion in determining whether a measurement of radiance or irradiance should be performed, it would be more accurate to consider if the source fully extends beyond a circle of 2.2-mm diameter or not. If a primary light source under-fills this area then a risk group classification cannot be transferred. The FOV could in this case be reduced to ensure the measurement of true radiance, but this may give rise to an overestimation of the hazard.

Taking this idea further, in measuring a primary light source, and adopting the result as a measure of true radiance, one makes assumptions about the radiance uniformity across the 11-mrad FOV. Such assumptions may not be valid if the radiance profile included regions significantly higher than the average. This case may be considered in the future.

As a final note on the transfer of classifications, since primary light-source manufacturers have hitherto reported risk group classifications of their products to IEC 62471, the transfer of these classifications to luminaires may still be sought. In consideration of the blue light hazard, classifications can be transferred only if a measurement of true radiance of the primary light source was made. One should therefore consider the diameter of the circular area evaluated by each IEC 62471 risk group - RG0 (20-mm), RG1 (11-mm), and RG2 (0.34-mm diameter). If, in the luminaire, more than one emitter falls within the circular diameter corresponding to the reported classification, transfer is not permitted.

Determination of d_thr

Where a luminaire exceeds the limits of RG1 at 200 mm, one should determine at which distance from the source the blue light RG1 limit radiance is found. For ease in evaluating this parameter, the blue light radiance in 11-mrad FOV is considered through the irradiance geometry: The threshold illuminance, E_thr, at the RG1 blue-light limit irradiance can then be computed. This has the advantage of allowing the use of goniophotometric data (and the inverse square law) or an illuminance meter to determine d_thr, the distance at which E_thr is found.

This does not take into account the fact that the measurement should be evaluated in an 11-mrad FOV. If the luminaire under test subtends >11 mrad at d_thr, it follows that emission from the source outside the 11-mrad FOV contributed to the measurement of illuminance used to find the location of E_thr, leading to an overestimate of d_thr. In certain circumstances this may not affect applications, but it can impact the product marketing since one will tend to select those sources having a shorter d_thr, with the perception of being safer.

FIG. 3. The sequence depicts an example of the variation of blue light radiance in an 11-mrad FOV with distance. The average radiance over the FOV falls below the RG1 limit only where the source falls entirely within the FOV.

Note that application of this technique is currently provided for white light sources only, since for colored light sources the use of photometric detectors may give rise to a significant spectral mismatch error. Although no specific guidance is given, it is reasonable to use goniophotometric or illuminance meter data, provided that a comparative measurement against a spectroradiometer be made to generate a correction factor.

Determination of d_thr of extended sources

In the 2014 edition of IEC TR 62778, guidance is provided in Annex D to address the case where a source subtends >11 mrad at the initial estimate of d_thr, here labeled d_N for clarity. Nominally applicable to arrays of LED sources, the principles of Annex D can be applied to any extended source.

The variation of blue light radiance with distance is not easy to predict since it results from averaging the luminous area with the dark background over the measurement field of view, as demonstrated in Fig. 3. It is therefore strongly dependent on the spacing and the shape of the luminous area of the source with respect to the circular FOV.

FIG. 4. The upper image depicts the source extending beyond an 11-mrad FOV at d_N, demonstrating overestimation of threshold distance. At refined estimate, d₁ (lower), only one emitter falls in an 11-mrad FOV and d₁=d_thr.

Having determined d_N, one should evaluate whether or not the luminous area of the source extends beyond the circular area described by an 11-mrad FOV at d_N, of diameter 0.011∙d_N. Where the source falls entirely with this area, then d_N=d_thr. Otherwise, as is the case in the upper image in Fig. 4, d_N is overly conservative and can be refined using the following process. It should be noted that Steps 3 and 4 of Annex D approach this question from a different perspective, but due to ambiguity in the definitions therein, this simple consideration is recommended.

The distance, d₁, at which the E_thr of a single packaged LED occurs, should then be determined. It is presumed that this will be measured using an illuminance meter lest the luminous intensity of the single emitter be too low for measurement with a (luminaire) goniophotometer. The area described by an 11-mrad FOV at d₁ of diameter 0.011∙d₁ should then be considered. If only one emitter falls within this area as is the case in the lower image in Fig. 4, then d₁=d_thr. Where more than one emitter falls within this area, and where E_thr was taken from the LED emitter datasheet, as opposed to resulting from the measurement of the luminaire, it does not necessarily follow that RG1 be exceeded at this distance. It is in this case recommended to perform a spectral radiance measurement at d₁ in an 11-mrad FOV, if the result is below the RG1 limit, d₁=d_thr. In all other cases the true value of d_thr lies between these extremes, so the default position is to adopt the worst case, d_N=d_thr.

Should one wish to continue the analysis, one could consider the distance, d₂, at which is found the E_thr of, for example, an LED module or light engine used in a luminaire. One would simply repeat the same process described here.

To measure d₁ (or d₂) requires that all other emitters in the luminaire be extinguished or covered, which in many instances is neither easily realizable nor practical. One could compute the required increase in the area over which the 11-mrad FOV must average to reduce the blue light radiance (L_B), to the RG1 limit. Estimating the area (A) of a single emitter, the diameter of the FOV required to reach the RG1 limit, d=√L_B×A/π×10000. If only one LED falls within the circle of diameter d, d_thr=d/0.011. If more than one emitter falls within the area, the number of emitters should be included in the computation and an iterative method applied. Although this technique may present different challenges to that of Annex D, it can be useful for the analysis of some sources, and in other cases as a screening mechanism to determine where effort should lie.

Conclusion

The implementation of IEC TR 62778 and the new approach to the evaluation of the photobiological safety of sources intended for lighting applications will in many instances lead to a simpler assessment. In others, where a refinement of d_thr is sought, additional interpretation will be required, yet interpretation in standardization can be problematic. This highlights the need for a sounder metrological approach to the determination of d_thr. It is expected that now, with the harmonization of the revised editions of the majority of lamp and luminaire standards to the EU low-voltage directive, IEC TR 62778 will be increasingly consulted and applied. It is also anticipated that the issue of extended sources has not yet been laid to rest.

LESLIE LYONS is the technical support manager of Bentham Instruments Limited, UK (bentham.co.uk), and a member of BSI and IEC committees TC 76, Optical Radiation Safety and Laser Equipment, and the Photobiological Safety Panel of IEC TC 34/SC 34A - Lamps.