ºÃÉ«ÏÈÉúTV

It is impossible to successfully identify and tackle failures in design, construction and operation if the various tools used to predict energy use in completed buildings produce different results. But energy prediction software packages can come up with very different heating and cooling requirements in a building but produce very similar carbon dioxide emissions. How does this occur? Is one software package better than another at predicting the energy use of a building and why do they provide different answers?

I am not intending to rate the different software for accuracy. Software packages are often designed to do different tasks:

• Compliance with ºÃÉ«ÏÈÉúTV Regulations
• Calculate the design loads to select heating and cooling systems (Boilers and Chillers)
• Simulate annual energy consumption of a building

Software packages have different data entry requirements and once entered use different calculation engines to produce the energy use prediction.

We also need to be realistic about the complexity of modern software and allow for the fact that most software will have coding errors and some bugs. Just take a look at the bug list from various software vendors

On the first point different packages use varying time steps to break down climatic data into an average figure. Some packages break this down into monthly steps whereas others break it down into daily or even hourly steps. The smaller the time step used the greater the potential for accuracy.

Most software packages require buildings to be broken down into zones depending on the use of that area and where it is located in the building. The size of the zones depends on the software package – some break this down into small elements where others do it by floor or even the whole building. Again the smaller the zones the more accurate the final prediction is likely to be.

We also need to be realistic about the complexity of modern software and allow for the fact that most software will have coding errors and some bugs. Just take a look at the bug list from various software vendors (not just in the construction sector)!

On a technical level using different calculation procedures will produce different results. The CIBSE Admittance technique is better suited to air temperature predictions where these natural and flowing in nature and the ASHRAE response factor procedure can deal better with sudden step gains in internal air temperature.

Sophisticated simulation models can analyse 3D conduction of heat and include how this varies depending where on the building this heat transfer takes place. Computational Fluid Dynamics (CFD) programs can produce sophisticated models of air flow. A simplified model will assume the temperature is the same throughout a space whether this is at your feet or at the ceiling. This doesn’t take account of reality where temperatures will vary depending on the temperature of adjacent rooms or at the building perimeter which can have a significant influence on localised temperatures within a space.

If you are lucky these variations can cancel each other out and give a relatively accurate result. However in some case these variations can add up to create a very inaccurate prediction of energy performance.

In conclusion we have a choice – do we want detailed thermal calculation software that requires far more user input for more accurate results or are we happy with a simplified and quick approach that can provide rough estimates?

Ant Wilson is European leader for advanced design, applied research and sustainability at Aecom