I have been spending some quality time in a baby hospital over the past few days. The hospital is really world-class, the equipment relatively new and the staff absolutely professional. It is too bad the building performance is not keeping up. The issue has to do with noise and, I believe, is easy to remedy by using a few kilograms of steel.
In a ward with many babies one would expect that noise reduction was a primary criteria of design. As monitor alarms go off they initiate a cacophony of baby cries that cause a great deal of stress to the already overstretched nurses and anxious parents.
Noise is unwanted sound. Sound may be understood as a propagating vibration of air. It involves small air pressure waves that the human ear drums, along with the brain, can sense. These pressure fluctuations are referred to as sound pressure. For instance at 1m distance from a person speaking normally the sound pressure is about one-millionth of the atmospheric pressure.
The sound pressure level (SPL) in a given frequency band is related to the root-mean-square sound pressure, p, in that band:
SPL = 20 log p/pref
Where:
SPL = sound pressure level, dB
p = sound pressure, Pa
pref = reference value of pressure, Pa (commonly 20 μPa and corresponds roughly to the weakest sound that can be heard by humans)
Thus, the threshold of hearing perception corresponds to an SPL of 0 dB. In contrast, an SPL above 130 dB tends to produce pain in the ear. At 100m distance from a jet aircraft taking off one is likely to be exposed to an SPL of about 130 dB. Doubling of the sound pressure results in a 6 dB increase in SPL while halving of the sound pressure results in a 6 dB decrease.
Sound could both originate at a distance, and thus be airborne, or it may originate by banging against the boundary of a room – impact. Moreover people are perturbed both by noise that penetrates the boundaries of a room as well as by the reverberation of noise within a room.
As most recording studios are well aware, in order to reduce noise reverberation in a room, one must use sound absorbing boundaries. All materials can be assigned an absorption coefficient between 0 and 1 that defines the percentage of sound that is absorbed by the material. 0 represents total reflectivity and 1 represents perfect absorption. For instance an open window will exhibit perfect absorption.
| Type of boundary | Noise absorption coefficient |
| Insulated stud wall with 15mm gypsum boards | 0,04 |
| Wood floor | 0,1 |
| Carpet on concrete | 0,3 |
| Suspended mineral fibre acoustic ceiling | 0,6 |
| Suspended glass fibre acoustic ceiling | 0,9 |
Absorption coefficients for common boundaries (sounds in the 500Hz to 1000Hz range)
A large room with mostly hard surfaces, such as a cathedral, will have a long reverberation time (the time it takes to reduce the sound by 60 dB) of 5 seconds while a smaller room with many sound-absorbing surfaces – say a recording studio – will have a short reverberation time of half a second. Classrooms should have 0,7 seconds or less in order to comply with typical standard requirements while concert halls may have reverberation times of 2 seconds when fully occupied.
One must also provide boundary elements that reduce the transmission of sounds into a room. This can be measured by the sound isolation capacity of boundary elements. While most office rooms must have boundaries that provide sound isolation of 40 to 50 dB, hospital rooms have higher requirements in the range of 50 to 60 dB.
We will not delve into how sound isolation requirements are derived for building projects. Suffice it to say here that such requirements depend on the anticipated source of noise outside the room and the room’s intended use. The design may also consider the background steady-state hum that is available within the room to mask transient private conversation.
A distinct source of sound that is hard to isolate is noise from impact. The most common source is associated with people walking. Other common sources include rolling carts, floor impacts in fitness centres and rain. A boundary that is adequate for isolating airborne sound may be found to be inadequate for impact isolation – and in-situ testing may be required on some projects.
A typical insulated steel stud wall with 15mm gypsum board on either side can achieve airborne sound isolation of up to 50 dB. This can be higher than that for typical single and double skin brick walls. Higher insulation is possible by increasing the stud depth, board thickness or insulation. This makes steel stud walls excellent at separating spaces in hospitals where light-weight, quick and reconfigurable walls are a critical acoustic asset.
When it comes to sound isolation between floors a key requirement for acoustic performance seems to be the use of well detailed ceiling systems. Precast plank, cast-in-place and other floors without ceilings typically have poor acoustic isolation performance, both for airborne and impact noise. Intelligently detailed ceiling systems can achieve floor and ceiling sound isolation of 50 to 60 dB for airborne noise and 40 to 60 dB for impact noise.
The solution to the noise pollution in the hospital wards could thus have been as easy as using absorbent walls, floors and ceilings, while also adding stud walls and isolating ceilings around the baby units. In addition to providing privacy for the babies’ families this may, in the long run, result in healthier babies, happy parents and a more productive hospital staff.
It would be wise for architects who work for the department of health and private hospital groups to consider acoustic performance in the design of future hospitals. As for the existing hospitals it may make more sense to carry out retrofit work that can be completed economically by South African Light Steel Frame Building contractors without the need to rebuild new wards.
