Speech intelligibility: sound systems for emergency purposes

Sound Systems for Emergency Purposes (SSEP) are an integral part of a building’s life safety system. They are an essential means of communicating directions to a building’s occupants during a fire situation so that safe and efficient evacuation may take place. SSEP in larger or more complex buildings often rely on spoken directions as part of the evacuation procedure, where it is of critical importance that the message being presented is readily intelligible to all occupants. Needless to say, if speech intelligibility is poor, occupant safety is at risk.

The National Construction Code (NCC) stipulates criteria for the SSEP speech components. Surprisingly, although these legislated requirements exist, it is not uncommon in certain Australian States and Territories for appropriate speech intelligibility testing to be completely overlooked at the time of commissioning and sign-off of new buildings. This leads to the potential situation where a significant number of existing buildings could have less than ideal or even unsafe SSEP.

This article discusses the various aspects of speech intelligibility in SSEP, the importance of objective intelligibility assessments, legislated NCC minimum requirements, compliance verification methods, and factors that adversely affect intelligibility. Finally, the Building Surveyor’s role in certifying the NCC speech intelligibility requirements is discussed.

Speech intelligibility depends on a number of factors in the built environment, including the quality of the SSEP components, loudspeaker positioning, the amount of sound reverberation in the space, and the background noise level. In addition to this, speech intelligibility is also highly subjective and varies from person to person. One’s ability to understand a spoken message increases with familiarity to the message content, but typically decreases with advanced age, as well as if one’s native language is not the same as that of the message. Due to this subjective variability, it is vitally important that intelligibility is assessed objectively rather than subjectively.

The NCC stipulates objective speech intelligibility criteria, referencing Australian Standard 1670.4-2004, which specifies that a minimum Speech Transmission Index (STI) score of 0.5 must be achieved from the SSEP speech components. STI scores range from 0 to 1, with a higher values representing greater intelligibility:

In essence, the NCC requires the minimum intelligibility to be at least fair or better for the average listener. For the not-so-average listener, e.g. where English is not the native language or if the listener is of advanced age, the above scale actually shifts so that achieving an STI score of 0.5 could actually equate to poor or even bad intelligibility, depending on the individual listener. This highlights the fact that the NCC minimum requirement is not particularly onerous, and there may be situations where achieving a score of greater than STI 0.5 would be more appropriate. In an ideal world, the target would be to achieve excellent speech intelligibility in all situations, however, in practice and in certain building spaces this can be quite difficult to achieve.

For SSEP which include spoken directions, the risk associated with having inadequate speech intelligibility is that the spoken warning message may be misunderstood or misinterpreted. Occupants may not receive the message to evacuate at a certain time or to stay put in their current location, or may even do the opposite to that which is intended by the SSEP operator.

Buildings spaces such as car parks, railway stations, stadiums, atriums and roading tunnels are more prone to having poor speech intelligibility and not meeting the STI 0.5 criterion, unless due consideration is given during the design, construction or maintenance process. As mentioned earlier, the factors which adversely affect the STI score include:

• High background noise levels
• Insufficient sound absorption resulting in excessive reverberation (very echoey spaces)
• Inadequate SSEP system components, e.g. underpowered amplifiers, low quality loudspeakers, poor quality message recordings, etc.
• Large distances between the loudspeakers and occupant ear level
• Low directivity of the loudspeakers (loudspeakers that spray sound in all directions rather than directing it towards the specific location of the occupants)

All of these factors are interrelated, meaning that a low STI score can only be improved so far if any one of these components is addressed in isolation. If a space were found to be severely non-compliant, it is likely that most of the above factors would need to be improved to some extent in order to bring the STI up to a compliant level.

As with many aspects of the NCC, however, alternative solutions developed in conjunction with a Fire Safety Engineer may also be utilised to relax or bypass the minimum STI requirement in a safe manner. This approach could be used where it would otherwise be cost prohibitive or impossible to achieve the STI criterion. Alternative solutions are most likely to involve some type of visual indicator; examples of this may be to include evacuation signs in car parks or using scoreboard warnings in sports stadiums, which illuminate when the SSEP is triggered, in addition to the fire alarm system.

In terms of evaluating the STI scores in a building, this can be estimated during the design process, as well as being measured in-situ in completed buildings. During design, the prediction process usually involves the creation of a detailed 3D computer model, and can be quite labour intensive. In-situ testing on the other hand, is a somewhat more simple and straightforward process, provided the correct test equipment is available. The STIPA (STI for Public Address systems) method is most commonly used, which involves playing a specific test tone through the SSEP and recording the resultant sound on a special sound level meter that is loaded with the STIPA signal processing software. With the correct equipment, a single test location can be completed in approximately 30 seconds, making it generally possible to test an entire building in less than a couple of hours. The NCC requires that a building’s SSEP speech intelligibility is verified through in-situ testing.

Upon completion of a new building or major building renovation, the Building Surveyor certifies that the SSEP is compliant with the NCC. In New South Wales and the Australian Capital Territory it appears that Building Surveyors typically request to see evidence that the STI testing has been completed satisfactorily, in accordance with the relevant test standards. Broadly speaking however, in the other Australian States and Territories, it is more common for the Building Surveyor to accept a purely subjective evaluation of the speech intelligibility which is typically highly non-scientific, consisting of a person listening to a repeated phrase played over the SSEP and making a judgement call on whether they can hear the words clearly or not. Particularly due to familiarity with the test phrase, amongst other subjective issues, it is the author’s opinion that the speech intelligibility is likely to be overestimated in most cases using this method. This creates the situation where a potentially large number of existing buildings have non-compliant and unsafe SSEP.

In conclusion, many SSEP make use of spoken directions to assist in evacuation, and it is of vital importance that these directions are intelligible for the safety of all occupants. For this reason, the NCC stipulates a minimum speech intelligibility criterion, which is readily assessed through objective in-situ testing. The thoroughness of such compliance testing varies greatly across the different Australian States and Territories, depending on the amount of lenience offered by the Building Surveyor at the time of building certification.

Developers and building owners should be aware that their SSEP – if it contains spoken directions – requires careful design, construction and maintenance to ensure that speech intelligibility is good enough to avoid unnecessary risk to the building’s occupants in an emergency situation.

-

The author Tim Beresford is the manager of the Melbourne Acoustics team at Norman Disney and Young. During his 9 years at NDY, he has tackled many challenging projects and is regarded within the organisation as the technical expert in HVAC design, spatial acoustics and vibration analysis. His passion for acoustics extends beyond his working life, through his love of performing, writing and recording music.