Elstree Film Studios by Andy Munro MIOA.

Introduction In early 1997 Munro Associates was approached with an invitation to bid for the design of new sound stages at Elstree Film Studios, which had been purchased in 1996 by Hertsmere Borough Council in order to ensure its future as a major British movie making facility. Considering that the studios had hosted such major productions as Star Wars, Indiana Jones and, more recently, Saving Private Ryan, it was not wishful thinking to propose that additional shooting and recording capability should be designed to the highest standards and even improve upon existing performance criteria. The client's requirements were clearly aimed at a flexible, expandable complex which could offer exceptionally high sets which would be both fully and continuously ventilated, and quiet enough to record, even under a megawatt of lighting; an oxymoron by existing standards. Design Brief for Stages 5 & 6 Stage size and orientation It has been determined that the minimum required area for each stage was 1500 m2 and that two stages were required, with the option of two more in the foreseeable future. The stages should be sound proofed to a degree which would attenuate normal external traffic and site noise and therefore enable sound recording in all normal day to day circumstances. The most useful dimensions are almost but not quite square, giving the widest shooting angle in each plane. The stages should be in line and directly interconnecting for future operational flexibility although this should not compromise acoustic separation between the stages as they would normally be used independently of each other. Height to grid A clear working height of 15 m was required for set construction although some lighting would be rigged below this datum. The structural truss design and spacing would determine the lighting grid level and the two were to form an integrated structure in order to reduce steel requirements and to avoid unnecessary redundancy and duplication. Cat walks and safe access must also be provided to all areas at grid level together with stairs and fire exists. Due regard must be taken of CDM requirements. Acoustic properties The initial target reverberation time for each stage was set at one second. This was based on the volume and mean free path data available and the realistically achievable absorption coefficients for walls and soffits. This assumed no useful absorption at floor level although this would often be a factor in creating a localised acoustic environment. The background noise level on the floor of each studio should be NC20; it was necessary that all the elements of the design were optimised for this noise level if it was to be achieve as any weakness in the building envelope or in the building services could have contributed to a failure to meet the noise criteria. Ventilation Research had shown that few stages are provided with satisfactory ventilation and that normal filming operations are often compromised by high temperatures on the set and the need to open loading doors and switch on noisy extract fans. It was decided to design a ventilation system which used the massive under floor slab area to provide a constant flow of air at low noise level and at outside air temperature. Velocity control would be used to regulate cooling load and comfort factors and independent back ground heating was to be provided. Heat extract was to be at roof level, through attenuators mounted in the side wall elevations. Design calculations were to be provided at the earliest possible stage in view of the structural implications of the floor and roof design. Power It has been verified that each stage may require up to a megawatt of electrical power in extreme circumstances and so a feeder of 1000 amps per phase per studio should be considered the minimum provision and 1500 amps may be advisable. Various voltages may be required on the set and an accessible distribution system should be provided for each studio. Internal power distribution is normally set for each production and cable trunking and safe grid level trays should be provided at regular points along each wall and truss line. Services Apart from specialised services for the studios each stage would require water and normal electrical services for cleaning, maintenance and working. Flood lighting should provide a good working light level of 500 Lux. The ancillary accommodation will require all the usual services with provision for at least 100 staff. Lighting As camera lighting is set for each production it is necessary to provide only power and a grid for fixing each lighting unit. Power supply and ancillary lighting had already been detailed be others, including a completely new sub station and high voltage transformer. Ancillary accommodation A number of dressing rooms and reception area were to be designed with due regard for acoustic isolation from the stage, especially lifts and stairs. Access and escape Each stage was required to offer direct access to heavy goods vehicles and large pre-fabricated scenery. It was decided to make each main door five metres high. As large number of people were to be anticipated for audience shows and exhibitions it was necessary to provide rapid egress for up to one thousand, including adequate disabled facilities. Preliminary Works Munro Associates was appointed in August 1997, to undertake a full acoustic design and Anthony Evans & Partners were contracted as Architects. A preliminary design scheme had already been submitted and this was based on a double skin shell with a woodwool slab inner envelope which would double as both cavity wall leaf and internal acoustic absorption. This decision was made after a detailed study of the cost effectiveness of various schemes, as well as consideration of the long-term effects of scenery building, water cannons and minor explosions! Site Survey A detailed appraisal of noise generated at the proposed site was carried out during normal working hours. Care was taken to ensure flights to a nearby airfield were passing within the corridor most likely to cause interference although heavy goods traffic and engine revving proved to be the loudest local source. The noise levels were compared to the requirements to meet NR20 and then checked against the sound reduction index of a single, 140mm homogeneous concrete block wall. The result is shown in Table 1. The performance of a reinforced woodwool slab roof was also compared to the isolation requirements and it was found that two, 150mm reinforced slabs, each screeded on one side only, and laid with the maximum cavity depth between them (see detail) could only achieve the results shown in Table 1. As the wall performance would be increased by the inner woodwool shell it was clear that the roof would be the limiting factor. The final performance of the building shell would be improved by the following factors which made is safe to adopt the proposed envelope; Extra steel cladding to roof and walls Distance from roof to stage floor High sound absorption in the worst case 260 and 500 Hz bands. In addition to the outer wall isolation, the inner woodwool acoustic shell was designed to give an additional cavity wall advantage by virtue of a screed on the inside leaf. The exposed outer surface (facing the studio) was to be left unsealed to give as much absorption as possible. The properties of the chosen slab, which was 50 mm thick with 150 mm edge reinforcement were as in Table 2. Floor design The design of the floor was originally intended to create a series of large tunnels, through which air could be fed to a series of large tunnels, through which air could be fed to a series of grills around the perimeter. Each section was to have a span of three metres at which the floor would have a natural frequency of 40 Hz and a static deflection 0.05 mm for a live point load of 3 kN. This was checked in order to avoid camera shake during live action such as a moving cart. Unfortunately it was discovered that the site was contaminated with asbestos and it was deemed necessary to build a solid floor over the site which involved the least possible soil removal. Vibration measurements showed that a fully floated floor was not required and indeed would have been prohibitively expensive. The floor was to be covered with two layers of marine plywood in order to facilitate the fixing of scenery without damaging the concrete screed. It was estimated that this would give approximately 150 additional Sabines of absorption across the frequency spectrum. Ventilation It was decided to design the ventilation system to work above ground level with two large perforated socks at the corners of the north wall. This would run almost the full height of the studio and be of two metre diameter, giving a total supply area of 100 m2. Variable velocity fans were specified to offer some control over working temperature and noise levels. Exterior ducting of 18 metres to each sock, plus both primary and secondary silencers would give the required noised reduction. Extract was to be at roof level with a return fee to the air handling unit for winter heating. The system was designed to remove all the hot air under full lighting loads, and to give more conventional re-circulatory ventilation when rigging and working without lights. It has been noted that conventional sound stage design relied on massive, noisy extractor fans and open doors which severely compromised the acoustic performance of the building. There was also evidence of extremely high temperatures building up during unventilated periods with fires occurring on occasions! Final Acoustic Design Although the manufacturer's data was well tried and tested, several factors were different in both application and design of the system. In particular the roof slabs were to be fixed with the troughs facing each other and only one inner surface and the top (outer) surface sealed with screed. The stiffness and hence low frequency isolation performance was dependent on both the steel truss design and the fixing of the slabs and this could not be modelled without considerable analysis which was judged to be unnecessary, given the reasonable safety margin for isolation. The walls slabs were more conventionally coated, with the screed on the cavity side but they were independently mounted in a steel frame which was to fix to the outer block wall (see detail drawings in Figures 1 and 2). This was done to maximise de-coupling and hence isolation but the effect on low frequency absorption was difficult to model. It was decided that the performance would be better than in a rigid, fixed wall. Although air absorption would be significant it was ignored, given that only high frequencies would be affected. Fig 1. Examples of construction details. Fig 2. Construction details continued. Initial test results Table 3 shows a comparison of initial test data and the original design calculations. A corrected value for the acoustic absorption is indicated which allows an appraisal of the real performance of the woodwool shell. In most respects there is good correlation between prediction and performance with the following exceptions; The low frequency reverberation time is considerably less than the simple prediction and this was inevitably a consequence of using a frame construction with large panels capable of random vibration and resonance, as well as a certain percentage of leakage into the cavity wall. There is a general lack of data available on the low frequency (less than 100 Hz) performance of large panel walls and the fact that the averaged absorption coefficient was almost doubled to 0.33 is a very useful property where noise control and modal damping is a design goal. Fortunately the stages required as little reverberation as possible and so the result was deemed a success. The reverberation time at 1 kHz was higher than predicted and this was estimated to be the result of the large amount of exposed steel channel along the edge of each panel plus the large amount of steel in the roof space, for walkways and lighting. As the reverberation was less than predicted at 500Hz it was agreed that the design data was changed by the type of construction and mounting of the panels, which did not conform exactly to the manufacturer's original test certificates. Overall the preliminary results were acceptable ad the high frequency absorption was much as predicted with evidence of additional air absorption above 4 kHz. Final Test and Commissioning File ELSRT1 in Figure 3 shows the reverberant decay plots (Schroder) for 1/1 octave bands from 63Hz to 500Hz. The lowest curve, at 500Hz, was measured at just under 1s while the other curves are all of the same gradient and show no evidence of strong echoes or secondary decay slopes. The second Schroder plot (Figure 4) shows the decay slopes from 1 kHz to 4 kHz. Again the absence of strong discontinuities in the gradient indicates good diffusion and distribution of reflected energy. This was borne out by more traditional testing with hand claps and dropping of heavy objects. Sound Insulation At the time of writing some work was yet to be finalised to door seals and ventilation system were not fully installed but early indications were that the building did meet its background noise requirement of NR 20 in the presence of normal site traffic. Plane and truck noise was inaudible and the noise spectrum of the ventilation followed the NR20 curve when measured at a distance of 10m. Summary The new stages are a major addition to the facilities at Elstree and they are the highest in Europe with the lowest noise floor in normal operation of any in the UK, of this size. The use of woodwool as a combined barrier and acoustic tuning material is not new but to use it exclusively was a challenge dictated by budget and structural constraints. The final design proved to be both practical and cost effective. The performance has been proved to meet the client's demands and I look forward to seeing the Elstree Studios credit on many future, British made movies. Acknowledgements Particular acknowledgement and appreciation is due to the following, for design team contribution; Anthony Evans, Architect (overall project); Clive Glover, Architect for Munro Associates; Bob Wiles NTP Quantity Surveyors; Alan Philpott, FCR M & E Consultants; Gary Redman DFP Structural Engineering. Andy Munro MIOA is at Munro Associates, Unit 21 Riverside Workshops, 28 Park Street, SE1 9EQ