10.             Recommendations

The following recommendations are separated into those related to the development of codes and standards and general design practice and those recommendations for future research. 

10.1.         Recommendations for Codes, Standards and General Design Practice

The Technical Committee on Notification Appliances should consider adding text, either in the code or in the annex, describing how much of the vertical and horizontal surfaces in an aisle need to be illuminated to provide sufficient indirect signaling to occupants.  While it would be desirable to base the requirements or suggestions on actual test data, it may be reasonable to use simple geometry and human ergonomics in drafting the language.  A Task Group should be formed to study the issues and make formal recommendation to the committee on how to proceed. 

It is not recommended that the code require strobes in every aisle.  While that would improve system performance, the tests showed that it was not necessary if the stock height, strobe height and aisle spacing allowed strobes in one aisle to penetrate and provide indirect signaling on the surfaces of stock in adjacent aisles.  In addition, it must be recognized that the normal use of these stores may result in a rack/shelving reconfiguration after the strobe system has been designed and installed.  The goal for flexible-use spaces should be a visible system that works for different rack layouts given the height of the strobes, allowed height of stock and the spacing of strobes and aisles.  Nevertheless, wherever possible, designers should place strobes directly over aisles or in a manner that ensures penetration into adjacent aisles. 

The committee should also consider adding text permitting the use of corridor rules in the aisles of stores and warehouses.  A Task Group should study the issues involved in the use of ceiling mounted strobes in corridors at mounting heights higher than the nominal 80 to 96 inches used for wall mounting.  The tests for this project showed that direct signaling did take place from strobes located farther down an aisle.  However, those strobes were at a higher intensity than the 15 cd eff. currently allowed for corridor signaling [i]. 

Clarification is needed on the calculations required to meet the intent of NFPA 72 section 7.5.4.3.  As noted in Annex 13, there are several different methods being used by the industry to calculate the size space covered by a strobe.  For ceiling strobes, should the coverage area be a square or a circle?  Is it necessary to include Lambert’s Cosine Law correction for the illumination of surfaces? 

Following the Danvers test, one participant (a member of the NFPA 72 Technical Committee on Fundamentals) sent a letter asking about the possibility of flashing some or all of the building’s lighting fixtures [ii].  That concept was considered and rejected by various NFPA 72 committees in the 1996 and 1999 code cycles.  However, testing conducted by Underwriters Laboratories, both in the laboratory and in the field, showed that flashing some building lights to produce as little as a 5 to 15 % change in ambient lighting was sufficient to alert occupants [iii].  Issues cited in the rejections by various code committees included how occupants would know what the flashing of building lights meant, emergency/secondary power issues, and the effects of long term testing on light fixtures and ballasts.  It may be possible that the use of building lighting systems in conjunction with strobes will result in more reliable and complete visible communication to the occupants.  The building lighting may be a supplemental system in addition to code designed strobe systems, or it may be an integral part of the signaling strategy. 

In addition to the new text being added to the next edition of the code, NFPA should add additional commentary to the handbook edition of the code and to NFPA’s Fire Protection Handbook.  That commentary should elaborate on the concepts of direct and indirect signaling and other factors discussed in the Test Results and Analysis and Discussion sections of this report. 

Designs should begin with strobes located directly over aisles.  This should include all main aisles and peripheral aisles that not likely to be altered during future store layout changes.  Where the heights of the strobes and the stock are such that strobes can penetrate into multiple aisles, it is not necessary to provide strobes over every aisle. 

10.2.         Recommendations for Future Research

The original proposal for this project envisioned a second phase to address visible signaling in other large spaces such as malls and atria.  One goal of Phase 2 is to test the performance design method in more challenging visual environments.  A second goal is to gather sufficient data to permit drafting of code text either permitting or limiting the performance based approach as an acceptable method of occupant notification for different scenarios.  It is recommended that Phase 2 also include additional tests in warehouses and stores.  Additional testing could help to better understand how much of the vertical and horizontal surfaces in an aisle need to be illuminated in order to provide sufficient indirect signaling to occupants. 

A greater number of deaf and hearing impaired participants would be beneficial for future testing.  In addition to “solicited” test participants, canvassing the regular users of the space using a well designed questionnaire would also be beneficial.  This would be particularly true in spaces such as airports, malls and atria where the number of survey respondents could be quite high, adding to the statistical validity of the data.  Future tests should be scheduled for different times of the day to attract a wider range and a larger number of test participants. 

Participants felt that strobe effects were marginalized in brightly lit areas.  These areas had much higher ambient light conditions than was used in the testing that led to NFPA 72.  Future testing might include specific tests or survey methods to determine reliable threshold levels for signal-to-noise ratios. 

One of the test locations had strobes installed on smooth, suspended ceilings (Danvers).  Another had an open plan ceiling, but with very little clutter and no large obstructions (Kissimmee).  The third location had an open plan that left all of the building structural members and utilities exposed (Reading).  The strobes at that location were generally below the level of most structural members and service utilities except air handling ductwork.  There were several locations where ductwork blocked strobes.  Future testing should include a mix of ceiling configurations to try and determine if ceiling clutter might also be a factor in a strobe system’s ability to provide direct signaling.