You can't have intelligent buildings without intelligence - an understanding of how buildings work How smart is your building?
'This funded-research approach to systematic case studies should be used to look at building performance more generally'
If we can contemplate 1 per cent for art, how about 0.1 per cent for making sure buildings work? This is the scale of investment in post- occupancy evaluation that Rab Bennetts believes could make a significant difference to our understanding of how buildings work. Call them intelligent buildings if you want.
Bennetts was giving the keynote address to a recent conference on intelligent buildings*. This is an area of great innovation, yet the amount of publicly available feedback is small.
Bennetts is particularly concerned about the effects on students and their recycling of heroic buildings. He sees students copying Future Systems' egg with no idea whether their design will work. Or Foster's Commerzbank. Have they noticed, from David Lloyd Jones' recent book, that the hvac energy consumption of the bank is 130kWh/m2/y compared with 70 for Sheppard Robson's Helicon building? This is not to dismiss the bank, but students (and teachers) should have more such facts to help them make better-informed decisions.
Bennetts praised the work of probe, a series of detr-supported post-occupancy evaluations, published in cibse Journal. In the journal these mainly focus on the detailed engineering of energy performance in individual buildings. But wider lessons are being identified (see below). Bennetts argued that this funded-research approach to systematic case studies should be used to look at building performance more generally.
Ideas for shopping
Another of probe's themes is the importance of designing for building manageability in order to maintain designed energy performance over a building's life. Ian Taylor, the energy manager for the Sainsbury supermarket chain, illustrated this in a striking way - though his approach is made cost-effective by the exceptional fact that Sainsbury has 700 largely similar buildings and is finishing another new store every two weeks.
Up to 20 areas are being monitored in a store each day, mainly for temperature. Measuring at 15-minute intervals creates performance profiles fed back to headquarters, where they are compared with profiles of expected performance. The central management team knows without further investigation many of the causes of discrepancy between actual and expected profiles: for example that ovens have been put on too early or the dehumidification regime has been over-ridden. This knowledge has been codified. Automatic e-mails can arrive on a local store manager's terminal within 12-24 hours, pointing out there is a problem, giving probabilities of what the causes are and indicating the fuel-cost implications.
Such a tight loop between performance and management does not necessarily lead to sticking with the tried and tested. Knowing that it can keep a grip on performance gives Sainsbury the confidence to innovate. A future store in Greenwich is being designed to cut energy consumption by 50 per cent. It will include opening windows in the sales area and displacement ventilation, provided from perimeter floor-level outlets rather than from the usual overhead silver spaghetti.
Probing more deeply
The probe team offered detailed lessons, particularly on the people-issues of realising good energy performance. These include:
hvac - Sophisticated systems and controls can be effective, but typically require a sophistication of building management skills only found in larger buildings.
Natural ventilation - In open-plan spaces it is difficult to find a balance between manual and automated control that fits the office politics and the users' understanding of building performance. There were also technical problems with mechanising window opening. Generally, occupant comfort scores tended to be higher in the smaller, simpler, naturally- ventilated buildings.
Lighting - There were often problems with commissioning complex lighting- control systems and matching control strategies to work patterns. The technologies of presence detection and lighting control via the telephone were sometimes problematic.
Automatic controls - These should allow the occupants to boost conditions or extend use hours, with automatic reversion to predefined output levels. Effective user interfaces are needed, including an immediate, perceptible system response.
For Jim Read of Arup Communications, the point of building intelligence is to achieve adaptability in building use. In reviewing the success of current technology to realise this, he pointed to the difficulty of finding any truly integrated buildings - where information on all aspects of building performance and building use can be interrelated. Integration of information can allow the building performance to fit diverse patterns of use - throughout the building and through the day - without waste. There is a risk from making integrated systems increasingly complex, but Read believes this is manageable if the user interfaces are standardised and thus become friendlier to users.
The main technical barrier to integration is a lack of standard protocols which would allow ready data sharing between subsystems. Read also noted that the glass fibre vs copper cabling debate remains unresolved, and that for communication we still need cables.
Most of the conference papers were more focused on the systems than those noted above: case studies of buildings and services, advocacy of particular approaches to controls and cabling, using the intelligence to support elderly and disabled people, facilities management and maintenance, smart skins and future materials. Plus some early thoughts on the urban scale - the wired city and its implications for the organisation of work.
*Intelligent Buildings: Realising the Benefits. Oct 1998. European Intelligent Building Group, tel: 0181 400 6136. £50