LED Technology: Important Technical Briefing for NFPA Members

A very important NFPA standard amending motion has been raised to overturn a vote passed in the autumn of 2014 by the NFPA SIG-NAS subcommittee. The SIG-NAS 2014 vote would limit the light pulse duration of notification appliances to 20 milliseconds during a single flash cycle, a factor of 10 decrease from the current allowance of 200 milliseconds according to NFPA 72 2013. The 2014 SIG-NAS decision would effectively disallow new LED based notification appliances to serve the needs of the life-safety market.

Download the report for more information.

2016 Proposed Change to NFPA 72

The amending motion, to be presented this June 2015, proposes to overturn that 2014 SIG-NAS vote on the basis of there being insufficient data to support the cited claims.

The amending motion appears as follows:

NFPA 72 Motion Seq 72-2 - Why the Motion Should Fail

Mircom has created this blog and commissioned independent research and review on the experiments cited on the subject to ensure that NFPA membership has a solid technical background on this issue ahead of the critical vote, on the amending motion, taking place at the annual conference and exhibition in Chicago, June 2015 at the technical meeting.

We believe that there is insufficient data to support restricting the flash pulse width to 20 milliseconds. Such a radical change will impact further innovation in this field and possibly have unintented consequences.

Accordingly we motion to return the NFPA text to its original content. A review of the experiments and data that was cited to support restricting the flash pulse width is available here.

For more information on the exact proposed changes reference CAM motion 72-2 regarding section

We hope you strongly consider the remainder of this blog post and read the full text of the NRC produced study on the issue before casting your vote this June.

LED (Light Emitting Diode) technology is currently used in countless convenience and emergency management applications. Common uses of LED technology for safety applications include: traffic lights, vehicular headlights, internal and external airplane emergency lighting, air traffic control towers, emergency vehicle strobe lighting (police, fire, and ambulance), and much more.

The Current Issue: Pulse Duration

The SIG-NAS subcommittee of NFPA 72 voted in the autumn of 2014 to make a change to its section of the standard. The change reads:*

A The maximum light pulse duration shall be 0.2 second 20 milliseconds with a maximum duty cycle of 40 percent.

The committee justification for this change states that “research has shown that the use of longer pulse widths with lights rated using the concept of candela effective will result in poor detection performance” although an exception was made for corridor situations.

Three studies were cited that formed the basis of this decision. To fully investigate this issue and bring some clarity, Mircom partnered with the National Research Council of Canada (NRC)* to conduct a thorough review of these studies, existing scientific data, and international standards related to the visibility of flashing signals (in the context of emergency lighting).

This review encompassed an assessment of the evidence base for the existing and proposed requirements, and a summary of the interpretations of these data as used in other jurisdictions. You can download the full report here.

Here are some concerns of the 3 Experiments Cited as foundational background for the change

1 & 2) Savage(2011, 2012) experiments

These two unpublished experiments compared the detectability of a xenon strobe flash with flashing light-emitting-diode devices assembled for the purpose of the experiments alone (i.e., these were not UL listed LED-based notification appliances)

Among a number of key concerns NRC presented regarding this research is the following:

  • The Xenon strobe was different in colour from the other experimental conditions
  • The experimenter controlling the movement of the pole was not blind to the experimental condition and the method to indicate detection left considerable room for unconscious error
  • The viewing task caused the stimulus to change in a manner that would never happen in a real installation where the device is in a fixed location. As the pole moved closer to the viewer, the flash became a smaller and more intensely lit area on the screen

These variances and the compromising of the field of view rendered the dependent variable in the experiment meaningless.

Overall, these two experiments provide no support for the hypothesis that pulse duration affects flash detection for the purposes of emergency notification. The experimental method is fundamentally flawed. (NRC-RR-351 / Page ES 3)

3) Bullough, Skinner, & Zhu (2013) experiments

These six experiments by Bullough et al. all had good experimental design and included null trials. It is important to understand the conclusions that Bullough et al. arrived at.

Specifically the (3) study concluded:

  • Neither the conventional effective intensity nor relative instantaneous illuminance increase on their own were useful metrics for predicting the performance of a signal light when viewed indirectly

There were no findings that shorter pulses were more detectable and in most cases the longer pulses had better detectability

  • “in Figure 6a, the 10 ms flash with a 2% illuminance increment yielded only 20% detection, but the 100 ms flash with the same illuminance increment was detected 90% of the time. The 1 ms signals were hardly ever detected.” (Bullough et al. Parameters for Indirect Viewing of Visual Signals Used in Emergency Notification, page 8)

Bullough et al. findings were reconfirmed in the NRC-RR-351 review on pages 12 and 13 in that “neither pulse duration nor peak intensity alone are good predictors of detection probability.”

Key summary observations from the NRC review of the 3 studies cited for lowering the pulse duration

  • The proposed change in pulse duration for visual notification appliances is based on a misapprehension. (NRC-RR-351 / page 15)
  • Parties supporting this change have stated that data exist to show that flash detection improves when the pulse duration is lower. A thorough scientific assessment of the studies cited shows that the statement is not true. (NRC-RR-351 / page 15)
  • The presented data does not support this change. (NRC-RR-351 / page 16)

The conclusion of the NRC report was as follows:

The research that has been offered in support of the proposed change to shorten pulse duration does not support this change. In our (NRC’s) expert opinion, the proposed change in pulse duration is not warranted.

Concluding Remarks

Mircom would like to make clear that we are not inherently for nor opposed to the allowing or disallowing of any technology in the markets we serve.

Our point of view is that we must allow our membership and the scientific method to determine the applicability of any new technology for the life safety market.

Our concern is when there seems to be a decision made either with a lack of transparency for the research opined to support a conclusion or when there is unclear interpretation of that research.

We encourage everyone to conduct their own research and get all the facts before this important vote takes place in June. You can download the full NRC report here.

*Why the NRC?

NRC is uniquely positioned to conduct this work:

  • Internationally recognized track record of successful lighting applications research in both advanced laboratory facilities and field research;
  • World-leading human factors expertise ensuring that lighting and controls options at least maintain, and ideally improve, occupant well-being;
  • Unique facilities for experimental investigations of lighting topics, including the full range of photometric equipment and access to facilities and expertise at the national metrology laboratory (NRC Measurement Science and Standards);
  • Participation in important national and international codes and standards bodies, including: NECB, IESNA, CIE

Download the report for more information.