In my July 2014 article, I discussed the progression of the speech intelligibility requirements within NFPA 72, National Fire Alarm and Signaling Code. So now that you’re familiar with the requirements, let’s delve into some of the basic considerations to be made when designing an emergency communications system (ECS) that uses voice announcements.
In laying out a system, the end objective is to provide a message that can be understood. The message should be simple, yet covey the instructions that will allow the occupants within a premises to understand the event that may be taking place and take action with the instructions that are being given.
The message may be as simple as, “Attention please, there is a report of a fire in the building! Please leave by the nearest clear exit!” It might also signal that more detailed information is on the way: “Attention please, there is a report of a fire within the building! Please remain in place for further instructions!” These instructions may be prerecorded or announced live. In either case, they to must be clear and understandable.
Keys to Getting the Message Across
I’ve always preferred a prerecorded message, as the tone of the voice can be properly modulated, with just the right level of urgency. Depending on who is providing a live message, there may be a level of excitement in delivering it that obscures the message. Additionally, any person who broadcasts a live message needs to be trained in the systems operation and in how to speak clearly through the system. If that person is speaking like they have a mouth full of mush, then the message may end up being garbled. The same holds true if they are yelling into the system. It is important for an ECS installation that instructions are provided to the users of the system.
But how is one certain that even if you have the most professionally recorded voice, and fully trained users of the system for live voice that the message will be intelligible? For this you need a little understanding on acoustics. Since my space here is limited, I’ll leave out the heavy technical terms … there are many textbooks that address this subject in great deal. For our purposes, let’s address some basics that influence system design.
Within a space such as a room, sound that is produced in the simplest of terms will either be absorbed or reflected. This is dependent on a number of factors:
- Size of the room
- Ceiling height
- Floor coverings
- Wall coverings
- Ceiling coverings
- Furniture within the room
- Temperature and humidity
- Number of people within a room
NFPA 72 requires that a speech transmission index (STI) of 0.7 be achieved. This is based on a scale of 0.0 to 1.0 in which 0.7 is good. With a 0.7 STI, 67-90% of the syllables, 87-94% of words and 95-96% of sentences spoken will be understood. The prediction of STI is independent of the language spoken. Thus if an STI of 0.7 can be achieved within a space in English, it can also be achieved in French.
There are a number of software packages available through which this data can be entered and a predicted STI can be obtained. These are very complex packages and should only be used with proper training and a complete understanding of acoustics. If I were to design a voice system for a large venue such as an auditorium or covered mall I would consider one of these programs. I would also enlist the services of an acoustical engineer to use the modeling programs.
Page 1 of 2 pages 1
Fire/Life Safety · Columns ·
Emergency Communications Systems ·
Fire Side Chat ·
Mass Notification ·
NFPA 72 ·