Alterations to the programme

It was created to make life simpler. Yet it has been far from plain sailing for the Part L compliance software SBEM since its launch 12 months ago. Users found it favoured air-conditioned buildings over naturally ventilated ones, that buildings which complied under SBEM wouldn’t necessarily pass with other compliance tools, and that the assumptions used by the software weren’t made available to users.

Terry Dix, MEP Europe business leader with Arup says: “Part L and SBEM mean it is now harder to design a naturally ventilated building because of the reduction in emissions and overheating criteria.” Terry Wood, partner of the Hibec Partnership, which specialises in running SBEM calculations for architects and contractors, agrees: “The quickest and easiest way to pass Part L seems to be to put air-conditioning into a building. It is very tough to pass with a naturally ventilated building.”

Paul Davidson, director of the BRE’s Sustainable Energy Centre, concedes the point. He says: “You have to think harder to make a naturally ventilated building Part L-compliant because it is easier to strip out energy use from a building which has higher levels of building services technology in it. With natural ventilation, you’ve only got heating and lighting to deal with. Heating is already well constrained because of the existing good insulation standards and the new air tightness regulations.” Davidson says BRE is reviewing this issue to see if there is an undue incentive away from natural ventilation.

Kept in the dark

Another problem is that the algorithms and assumptions inherent in the SBEM software were not available to users, as Dix explains. “One issue we have as engineers is that this is a black box. This makes it difficult for us to analyse and find the cause of any outputs that don’t seem to be what we would expect.”

Davidson says BRE is working to resolve this issue, too. “We understand this is a matter of trust, and if the system produces an answer an engineer doesn’t expect, they can’t really explore why that has happened. We are working on a detailed manual which will list the references to algorithms, or explain any new ones we have developed.” In the meantime, the BRE refers engineers to a summary of the method in the new BRE Information Paper (BRE IP2/07).

The fact that SBEM and the commercially available software (such as IES) produce different results is also unsatisfactory. A building may fail Part L on one system but pass it on another. Moreover, there have been a number of versions of SBEM and its software interface iSBEM, which also give different outcomes. Dix says: “Over the year we have had various software versions which give various answers. If a building complies in March, it may not comply in June, and be different again in August.”

It is very tough to pass Part L with a naturally ventilated building. The quickest and easiest way seems to be to use air-conditioning.

Davidson points out that in building energy modelling, it is inevitable that different systems and different versions within systems will give different answers. “BRE has been exploring the variance issue for decades,” he says. “It happens because different software and versions will make different assumptions about the data. This is particularly true for SBEM and the commercial software because SBEM works on a monthly basis and the approved alternatives use hourly simulation.”

Another complication is the notional building, which is the Part L 2002 equivalent building. “However, in the 2002 Regulations there is no actually defined building. So to create a notional building, we have had to specify it based on the rules outlined in Part L 2002. This means that we have had to make assumptions,” he adds. The pressure to produce the software quickly is one reason why these discrepancies have occurred, but Davidson says BRE is working towards agreement on these issues.

He says: “Users should note that if you start the Building Regulations approval process with one version of the software, you can stick to it all the way to completion. Also, in order to clarify things for users, any future versions will be released in April and October, so everyone knows when a new version is due.”

Biased system

SBEM users say the way data is treated seems biased towards certain elements. The system also lacks boundary checkers, which means users have to ensure they enter data correctly. Adam Sime, an executive engineer at Hoare Lea, says lighting is one area in which the system doesn’t reflect practice: “Lighting is heavily weighted in SBEM. There is an assumption that when the sun is up, lights are off. In reality, people often close the blinds to see computer screens so the lights will remain on.” Terry Wood adds: “SBEM is very precise about U-values, but takes a broad-brush approach to mechanical and electrical services, which will have a big impact on the efficiency of a building. The balance isn’t quite right.”

Davidson believes this emphasis on some elements reflects Part L, rather than the software, as well as the fact that we have more information on some elements of a building. He says: “Taking the example of U-values, we do have a lot of information on them. They are considered important within the SBEM system because in some types of construction thermal bridges can be critical, particularly as wall U-values improve. Poor design and detail in these areas can lose a lot of heat.

“On the other hand, it must be remembered that SBEM is not a modelling tool, so its treatment of HVAC is general – it simply says, ‘You are using this type of cooling system, which has an average seasonal efficiency of X.’ The system assumes any system is sized correctly. This is a simplification, and I agree there is more we can do to make the HVAC system more sophisticated. It is one of the improvements the Department for Communities and Local Government has asked us to address.” Davidson also says BRE is developing a “sanity check” on data. This will include a flag on the compliance output document, which will highlight any issues to building control. For example, it might say, ‘This building relies on a U-value of 0.001 – go and check!’”

Various versions give various results, but overall it is an enormous task for any piece of software and we admire what BRE has done.

Where next?

Keeping SBEM up to date is a continuous process. On behalf of CLG, BRE has attempted to respond to comments from industry and de-bug the system over the past year. Most users recognise this has not been an easy process and have been helpful in the improvement process. “Overall, this is an enormous task for any piece of software to do and we are in admiration of what BRE has done,” says Dix.

One of the main points made by users is that training is essential, and that SBEM and iSBEM become easier to handle with experience. Ted Pilbeam, building services manager at Fitzpatrick Contractors, has used the software extensively and overseen training of others in his department. “When I first began to use SBEM, I was not entirely confident in the outputs. However, the more you use this system the better it works. There is an awful lot of input to deal with, and you have to be sure about what you are putting in. I have found that once people have it explained to them, they realise it’s not as bad.” Terry Wood agrees: “Once you are used to it, SBEM does the job. At the end of the day it is still a fairly simple tool to use, once you have done the training.”

Davidson would like to see SBEM develop to deal with more building and building services engineering types and a variety of sustainable technologies. “We would like to be able to do more things with SBEM. At the moment the more low energy and intelligent the designers make a building, the more unlikely it is that you can use SBEM on it. For example, SBEM doesn’t deal with night cooling. We would like to build this in because it is the sort of technology that could become more widespread as we have to deal with increasing temperatures without going straight to air conditioning.”

Currently, SBEM is mainly used for new buildings, but existing buildings will also fall under the new regulations implementing the Energy Performance of Buildings Directive that will be published soon. Developments are under way to ensure that the system can cope with the demands of energy performance certification. “For existing buildings, it could be seen as an unreasonable penalty to expect people to buy the commercial software and training, unless they have a number of buildings to deal with. We would need to improve the range of buildings that SBEM can cover and many of the standards that it is based on have mechanisms for doing this,” says Davidson.

This point about certification of energy use in existing buildings is an important one for the construction industry. While concentrating on proving Part L compliance, designers need to be aware that the buildings they work on will need to continue to demonstrate energy efficiency in operation. It may well be easier to pass Part L 2006 with a fully air-conditioned building, but a long-term view suggests that exploring energy efficient options will be the sensible approach.