Although the current revision of Approved Document E has been in effect since July 2003, the long lead times on construction projects mean that its effects are only just beginning to be felt. One area that didn't receive much publicity at the time was Part E3 - the control of reverberation in the common parts of buildings containing flats or rooms for residential purposes.
This article examines the control of reverberation by comparing the measures taken in the design of housing, where reverberation is considered a problem, with the choices made in auditorium design, where reverberation can be a positive quality in the acoustic performance of the space.
Regulation E3 states: 'The common internal parts of buildings which contain flats or rooms for residential purposes shall be designed and constructed in such a way as to prevent more reverberation around the common parts than is reasonable.' The main question, of course, is what is 'reasonable', a proposition we shall come back to later on.
First, some definitions:'Reverberation' - the persistence of sound in a space after a sound source has stopped (i. e. as sound reaching the listener after multiple reflections, having taken a longer path and, therefore, arriving later than the direct sound).
'Reverberation time' - the time, in seconds, taken for the sound to decay by 60dB after a sound source has been stopped.
Reverberant sound can be beneficial. Reflected sound reaching the listener within 35 milliseconds of the direct sound will reinforce the original sound and make it seem louder.
These reflections are used positively in theatres and conference rooms to make speech more audible at the back of the auditorium.
However, reflections arriving later will blur the sound. Remember the old-fashioned railway station announcements that were made from one very loud speaker suspended from the terminus roof? The first part of the announcement, reflected off the roof and walls, coincided with the later part of the announcement, arriving directly at the listener's ear, making the whole thing unintelligible.
So, for clarity of speech, reverberation times need to be short. In contrast, music needs longer reverberation to achieve a fullness of sound. This presents a problem for acousticians designing multipurpose spaces intended to accommodate both speech and music. This problem can be solved electronically or by having finishes that can be adjusted to suit requirements.
Coming back to Part E3, why is reverberation considered a problem in the common areas of housing? This is perhaps a question we should put to the ODPM, who introduced this section of the regulations. We can surmise that there are two concerns:
a) the sound is slower to decay, so any disturbance to residents is extended; or b) the hard surfaces allow reflected sound to reach all parts of the space causing disturbance to all residents on a particular stairwell or common space.
One of the main problems is airborne sound penetrating dwellings via the gaps around ill-fitting entrance doors and acoustic weak points such as letter boxes. The regulators have spotted this weakness and paragraph 2.26 of the revised regulations requires entrance doors to provide a 29dB Rw sound reduction. But even with this better acoustic performance of doors, the present regulations say we still need to control reverberation in the common parts as well.
Be reasonable Returning to the earlier question, how does one demonstrate to the building control officer that the reverberation is 'reasonable'? This is not done by testing, as in other parts of the regulations, but by either a 'rule of thumb' (method A) or 'calculation' (method B). We are going to look at both of these in more detail, by reference to a study of one of Levitt Bernstein Associates' current projects.
Chalkhill Estate in Brent, north-west London, is very close to Wembley Park Station and has magnificent views of the new Wembley Stadium. The residents and client, the Metropolitan Housing Trust, chose to refurbish the lower blocks and to demolish the high-rise, replacing it with family and single-person accommodation in more traditional street patterns.
The work has taken place in phases over several years and the final phases are now under construction. While the buildings in each phase are very similar, they have had to accommodate changes in regulations - such as Part L and Part E. In addition to the two- and three-storey family houses, there are a number of small blocks of flats, four storeys high, with two dwellings per floor, opening directly off the landing.
Regulation E3 requires us to consider the acoustics of these shared spaces and starts with some definitions:
l a 'corridor' or 'hallway' is a space where the longest dimension is more than three times the shortest (long and thin);
l an 'entrance hall' is a space where the longest dimension is less than three times the shortest (short and fat);
l a 'stairwell' contains stairs.
These definitions are important as the requirements vary for each of these spaces, as we see in the following paragraphs:
'7.10 For entrance halls, corridors or hallways: cover an area equal to or greater than the floor area, with a Class C absorber or better. It will normally be convenient to cover the ceiling area with the additional absorption.' This seems simple enough. Measure the floor area and cover the ceiling (assuming it is the same area) with a Class C absorbing material. But the challenge to designers is to make their ceilings at least equal in area to the floor.
'7.11 For stairwells or a stair enclosure:
calculate the combined area of the stair treads, the upper surface of the intermediate landings, the upper surface of the landings (excluding ground floor) and the ceiling area on the top floor.' This is not quite so simple! Measure the floor (but not all of the floor) and some of the ceiling.
'7.11 For stairwells or a stair enclosure, calculate the combined area? Either, cover at least an area equal to this calculated area with a Class D absorber, or cover an area equal to at least 50 per cent of this calculated area with a Class C absorber or better. The absorptive material should be equally distributed between all floor levels. It will normally be convenient to cover the underside of intermediate landings, the underside of the other landings, and the ceiling area on the top floor.' Using this method in our example, the total area is 45.70m 2 and we would need to provide an equivalent area of Type D-rated material. Alternatively, we could provide 50 per cent or 22.85m 2 of Class C-rated material. Our ceiling is only 13.76m 2, and even if we add in the underside of the landings, we still don't have enough area to cover with a Class D material. Remember: the absorptive material should be equally distributed between all floor levels. We are reluctant to start covering the walls, as these are to be fair-faced brickwork, or to apply acoustic material to the precast-concrete staircase soffits, though this would be an option.
The final choice was to select a Type C material for the ceiling and underside of the landings. But choosing the correct material proved difficult.
The various material classes are defined in BS EN ISO 11654:1997, Acoustics - Sound Absorbers for Use in Buildings - Rating of Sound Absorption. Classes range from A to E, with 'A' being the best at absorbing sound.
One thing to note: the absorption characteristics are measured over a range of frequencies, and products must come wholly within the profile to achieve the stated rating.
Information underload When we came to specify products for these spaces, we found that data on acoustic performance are given for products that are being marketed as 'acoustic', but even these rarely state the classification under BS EN ISO 11654. Currently, there is insufficient information from suppliers about their products.
Even when you think you have found a supplier that is up to speed because its website says it supplies Class C and D absorbers and it has produced state-of-theart technical literature, it often still doesn't state the class!
We considered a standard mineral-fibre tile in a suspended-ceiling grid, but felt these could easily be removed/damaged, and accessible suspended ceilings in these locations can be used as a storage place for illegal drugs.
British Gypsum's Gyptone product was the eventual choice for Chalkhill. The works are still under construction and the finishes to the common parts have yet to be installed.
But beware when using perforated products such as these, as their acoustic performance depends on the depth of the void behind and whether you have chosen to fill that with a product such as mineral wool.
So, what are the drawbacks related to method A? These include:
l insufficient ceiling/soffit areas to apply absorbing material;
l suppliers not giving 'class' ratings for their products/materials;
l good absorbing materials are usually soft and, therefore, not durable, which is a problem in heavily trafficked corridors and stairs, particularly in social housing;
l good absorbing materials often not having Class 0 Spread of Flame classification.
Not so simple That was the 'simple' method. With the rather more complex calculation - method B - it may be handy to have a copy of Part E to hand:
l Step 1 - calculate the surface area related to each absorptive material (S1, S2, S3, S4? m 2) for the f loor, walls, doors, windows and ceiling. This is to establish how much absorption is provided by the proposed finishes so this can be compared to the theoretical requirement.
l Step 2 - obtain values of absorption coefficients (a) for each type of surface - for example, the carpet, painted concrete block walls and the timber doors - and for each of the five frequencies: 250Hz, 500Hz, 1,000Hz, 2,000Hz & 4,000Hz. It is very important to note here that you are doing the calculation five times over for each material. In the worked example in the regulations, the values are taken from table 7.1 (below). This is a very limited list of materials.
Architects are notorious for using familiar materials in unfamiliar locations and are likely to specify materials not on the list. Acoustics, Noise and Buildings by Parkin and Humphreys has a slightly longer list. It would be helpful if the Approved Document pointed us to other sources of acoustic data.
l Step 3 - calculate the absorption area (m 2) related to each absorptive surface (i. e.
for the floor, walls and doors) in octave frequency bands (absorption area = surface area x absorption coefficient). Repeat five times for each material.
l Step 4 - calculate the sum of the absorption areas (m 2) obtained in step 3.
l Step 5a - calculate the volume of the space.
l Step 5b - calculate the total absorption area (AT) required. In the worked example in the regulations, the volume is 30m 3 and, therefore, 0.2 x 30 = 6.0m 2 of absorption area is required.
l Step 6 - calculate the additional absorption area (A) to be provided by the ceiling (m 2). If any values of minimum absorption area are negative - see 2,000 Hz and 4,000Hz - then there is sufficient absorption from the other surfaces to meet the requirement without any additional absorption in this octave band (additional absorption = AT - total absorption area [from step 5]).
l Step 7 - calculate the required absorption coefficient (a) to be provided by the ceiling (required absorption coefficient = additional absorption area / area of ceiling).
l Step 8 - identify a ceiling product from the manufacturer's laboratory measurement data that provides absorption coefficients that exceed the values calculated in step 7.
So, what are the drawbacks related to method B? These include:
l suppliers only giving performance data for products marketed as 'acoustic';
l the limited list of materials in table 7.1 and no other reference given;
l good absorbing materials often not having Class 0 Spread of Flame classification.
To satisfy the requirements and gain approval, the regulations state that a report is to be submitted to the building control officer containing the following information:
l a description of the enclosed space (entrance hall, corridor, stairwell, etc);
l the approach used to satisfy requirement E3: method A or B. With method A, state the absorber class and the area to be covered;
with method B, state the total absorption area of additional absorptive material used to satisfy the requirement;
l plans indicating the assignment of the absorptive material in the enclosed space.
Music to the ears This is all a far cry from the specification process for St Luke's Church, London, where acoustic consultant Kirkegaard Associates of Chicago was employed to assist in the selection and positioning of materials to provide ideal conditions for the varied acoustic requirements of recordings, orchestral rehearsals, recitals, conferences and education events.
The UBS and London Symphony Orchestra (LSO) Music Education Centre was created by restoring and converting the ruins of Hawksmoor and James' Grade I-listed former church. Built in 1732 and repeatedly underpinned and shored up over the course of two centuries, St Luke's Church on London's Old Street finally succumbed to bad soil conditions in 1959, when the roof and interior structure were removed to prevent the loss of the entire building. Only the four exterior walls, the spire, and the crypt with its 1,012 residents remained after the emergency demolition.
Renovations have been completed to convert this landmark building into the primary rehearsal and outreach space for the LSO, with a seating capacity of 350. In addition to its rehearsal function, the building is the new home of the LSO's Discovery Programme, a series of educational programmes for children and adults - involving live and recorded events, distance learning, broadcast and broadband technologies.
The scheme has been designed to exacting acoustic standards, including 65mm of glass in two separated layers on the windows, and complex construction to avoid noise transmission within the building, while preserving natural light for the rehearsal space with minimal impact on the exterior appearance of the building. The musicians particularly like being able to see the outside world, as they are usually tucked away in enclosed auditoria or recording studios.
The design contrasts the new steel structure that supports the roof and balconies with the original walls, which are retained in their raw, unplastered state.
An entirely new structural system of four large 'tree columns' arranged on a single 50m-square structural module has been inserted into the central space. The trees support roof trusses, whose bottom chords will serve as catwalks. The trusses, in turn, support a slate-finished concrete roof slab from which a heavy wood ceiling is suspended resiliently. The hardwood panels are supplied by Soundcheck (Bridgeplex) and provide sound absorption to help control reverberation.
Reverberation, loudness, and reflection control are achieved using an extensive system of retractable banners that allow the acoustic properties of the space to be adjusted to the most advantageous conditions.
Heavy felt banners, supplied by Triple E, line the walls and can be dropped down to stop sound reaching the brickwork, reducing the reverberation time and allowing the hall to be used for conference or education events.
There are also similar banners that drop down from the ceiling in the main body of the hall whose function is to minimise any adverse reflections and again give the users the means to adjust the reverberation time of the space to suit the current activity.
The Jerwood Hall is sufficiently large to permit a full symphony orchestra and chorus to rehearse in comfort and has been well received by its new residents and visitors alike. As a postscript, it is perhaps interesting to note that concert venues are no longer spaces desiccated to the performance of classical music. Their viability is dependent on the more popular amplified music events and, acoustically, that's a whole new article.
Andy Jobling is technical manager at Levitt Bernstein