BSRIA visits the Heelis building, the new headquarters of The National Trust. How well is it working, is it energy efficient, and what do the occupants think of it?
Like many organisations with a strong environmental conscience, the National Trust was keen to ensure that its new central office building trod upon the earth as lightly as possible. In Trust language, sustainability translated into low energy consumption, low running costs, and an outstanding place to work. The Trust also wanted open-plan offices to encourage good communication between departments, formerly in different buildings.
The Trust also desired a brownfield site. A suitable location was found on a plot of land among former railway engineering sheds in Swindon. The trapezoidal site was a challenge for architect Feilden Clegg Bradley and environmental consultant Max Fordham who designed the shell and core.
The footprint of the Heelis building closely follows the boundaries of the site. The building's design - from its pitched roofs to the use of blue engineering brick - gives an affectionate nod to the nineteenth century sheds.
The building was needed as the Trust wanted to centralise staff from six sites. People came from 1970s smoked-glass office blocks and converted stately homes. They therefore arrived with varying expectations of the new building.
The Heelis building was developed by Kier Ventures as a pre-let for the National Trust.
In the analysis that follows, it is important to recognise that Keir Ventures used the RIBA Stage D report as the basis of costing and financing the project. Although the design team developed the engineering concepts to get the most effective packages, they were not aware that the developer had already set the budget. Ultimately, this affected the choice of some engineering systems, which are not as tightly specified as they could have been.
The design team settled on a two-storey deep-plan building on a north-south axis, with the longest facade angled due south. The construction is conventional, being of a steel frame on a concrete base, with a pitched roof, in-filled with exposed, 80 mm thick, pre-cast concrete planks to provide thermal mass.
The 30 degree pitched roofs on the south side provided a suitable orientation for 1554 photovoltaic panels, while the north-facing slopes provided a location for northlights. These are located between prominent motorised extract ventilators (called snouts).
The envelope is a mixture of aluminium curtain walling to the south, with smaller windows set into brick walls of the remaining elevations. Two courtyards break up the deep-plan nature of the building and to enable cross-ventilation. Lightwells through the first floor mezzanines bring daylight to the deep plan areas of the ground floor.
The Heelis building is mostly naturally-ventilated, with fresh air supplied through motorised windows and panels, and exhausted through the high-level snouts through a mixture of wind-assistance and stack effect. (Five snouts were designed to take fans, which would provide two air changes per hour on hot, still days. These have not been fitted.) Mechanical ventilation is provided for a canteen serving staff and public areas. Two conventional downflow refrigeration units cool the single communications room.
While the facade and snout vents are open and closed by a Trend building management system (BMS) on space temperature, occupants of perimeter desks can override the automatic settings for 60 minutes by use of robust switches by the windows. The snouts serving the meeting rooms can also be manually operated.
The BMS is programmed with a night-cooling algorithm. This serves to discharge the heat built up during the day and cool the concrete roof and first-floor slab.
In winter, a mechanical ventilation system with heat recovery is employed to avoid the wasteful loss of warm air. The objective is to reduce winter heat loss by about 70 percent. The air is supplied at low velocity under the raised floor and supplied into the occupied zones via trench heating slots.
Electric lighting is predominantly of T5 fluorescents in batten fittings, controlled by occupancy detectors. Compact fluorescent lamps used in corridor downlights and other fittings. High intensity discharge lighting is used in the public cafe and atrium.
Space heating is by finned-coil trench heating around the perimeter. This, and the ahu heater batteries, are supplied by a 153 kW output gas-fired condensing boiler to handle the base load and two atmospheric 265 kW output boilers to handle peak loads. The perimeter circuit is on a conventional 82C flow, 71C return.
The design team set high energy-efficiency targets, hoping to achieve around 75 kWh/m2/y for gas and electricity (35 kWh/m2 for gas and 85 kWh/m2 for electricity, including the server room and catering kitchen. As electric lighting can account for 30 percent of the energy use of a naturally ventilated building, the designers aimed to achieve a daylight factor of five percent in the office areas.
The designers have been monitoring the building during the first 18 months of occupation, improving and fine-tuning the building's operation alongside the usual defects liability duties. This kind of professional aftercare has benefited both the client and the design team, serving to both improve the building and educate the designers about what works and what needs more attention. As part of this fine-tuning, the design team appointed Building Use Studies to conduct an occupant satisfaction survey.
It has not yet been possible to obtain a complete and accurate picture of the building's annual energy use. While sub-meters are installed, owing to the budget restrictions the installation is less comprehensive than the designers had intended. The electricity sub-meters are not monitored by the BMS, and on-site record keeping is not yet routine. In spite of repeated requests, Heelis has not yet had a gas bill.
A preliminary analysis of the 2006 data showed (after correcting for use, weather and special items) that gas and electricity consumption in the main office areas was virtually identical to the good practice levels given in Energy Consumption Guide 19 for a naturally-ventilated open-plan type 2 office. Although considerably higher than the design expectation, this is still creditable.
There are two main areas of disappointment in relation to the design
estimates: gas for heating and electricity for lighting.
Some reasons for the higher gas consumption were apparent on the day of the visit. Despite a night of heavy snow and a midday external temperature of 2C, about one-third of the automatic windows on the first floor of the south facade were slightly open. As the day continued, these windows opened further. Meanwhile, in an adjacent meeting room the perimeter heating was not responding to the room temperature sensor and the roof vents were opening up instead, drawing more warm air through the door (left open when the room is free) from the adjacent open offices.
The BMS logs also revealed high running hours of the office heat-recovery ventilation, which is supposed to operate for occupied hours in cold weather only, but had recorded an average of 3100 h/y year, suggesting considerable potential for fine-tuning. While the specific fan power and fan energy consumption are low, excessive fan operation could increase gas consumption significantly, particularly if natural and mechanical systems operate together, as observed. The bms records also suggest that frost protection may be bringing on the fans and boilers, and not just the pumps.
The office space accounts for less than half of the building's CO2 emissions. The two other major users are the communications room (about 40 percent with its air-conditioning), and the catering kitchen (about 15 percent including the associated retail areas). Both of these also offer significant opportunities for improvement.
The communications room has a free cooling system for winter use, but was observed not to be running at full capacity while the chilled water system was also running, contrary to the design intent.
An occupant survey was carried out by Building Use Studies (BUS) in November 2006, as part of an ongoing post-occupancy analysis of the building by the architect. The feedback from the occupants is being used by the building managers and the design team to fine-tune the building.
All staff in the building were handed a self-completion questionnaire. The response rate was 92 percent (a sample size of 242). The survey results are compared against a benchmark of UK buildings that are all green in design intent.
The Heelis building scores well for design, needs, space and lighting. The lowest scores are for perceived control, some of the noise variables and some of the temperature and air variables.
Overall, the Heelis building is within the top 40 percent of the BUS green building dataset.
The building is reported to be reasonably comfortable overall, but can be too cold in winter and too hot in summer. Some people reported that the automated ventilators opened at odd times.
The survey results show that 68 percent of staff believe that their productivity is either not affected or not improved by the building. (The UK benchmark mean for perceived productivity is minus 2.06 percent, so any perceived productivity score above zero can be regarded as a good result.) Overall, the Heelis building is among the best buildings in the class of larger open-plan UK offices that rely on significant natural ventilation.
While ratings of thermal comfort and noise may seem disappointing, it is a challenge to satisfy occupants of larger, naturally ventilated open-plan buildings like the Heelis building. Its survey scores are among the highest achieved for buildings of this type.
The Heelis building is matching good practice, and has the potential to be better.
Over the coming months, the Trust and design team will be working together to fine tune performance and to set up robust routine monitoring and energy management procedures for the future.
One area that has particularly surprised the Trust and the design team is the disproportionately high energy consumption of the communications room and the catering kitchen. These are issues that normally do not get the full attention of design teams, despite warranting careful attention in the drive to minimise overall energy consumption.
Thanks are due to Liz Adams, facilities manager of the Heelis building, Feilden Clegg Bradley and Max Fordham for permission to bring post-occupancy survey data into the public domain, BUS for the occupant satisfaction analysis, and Bill Bordass for the energy use analysis.
For more information contact Roderic Bunn at BSRIA:
Tel: +44 (0) 1344 465600