Simon Lannon (Welsh School of Architecture) has provided the following by way of a contextual overview of the visualisation tool in development:
The simulation of the urban environment is a complex process; the EEP methodology simplifies this by using simple standard energy prediction tools, and ways of grouping houses together. The grouping of houses usually follows the type of house e.g. terraced, semi detached or detached, this is reasonable for simple problems, but when trying to predict the energy use of a detached house it could be two ends of a very large scale, from a labourers cottage to a mansion. The best way to group houses in this project is by their size and when they were built. To do this a number of common house types are surveyed, the results of these surveys are clustered together to give groups of houses with similar energy predictions.
The groups of houses are then modelled using the SAP technique to give a baseline or start point. From this baseline these groups of houses are modified to improve the energy efficiency, bolt on solar panels and take into account potential changes in the occupant’s behaviour. In this example it has been assumed the occupants might accept low inside temperatures by wearing more clothes.
The prediction of energy use for each of these modifications is undertaken then applied to each of the houses in the sample area, in this case most of the houses in Neath Port Talbot, South Wales (around 55,000).
Occupant behaviour also has an impact on whether a particular type of modification will take place. This is represented in this project by trigger points for ten year steps from 2020 to 2050, the trigger points represent this occupant behaviour year by year, and are associated with the likelihood of an occupant undertaking energy efficiency measures. The impact of energy efficiency measures is modelled by a series of trigger points that end with all the houses in the area having at least simple energy efficiency and 50% having a more expensive energy efficiency measures such as external insulation cladding.
The process of all modelling has been more complex than initially thought, perhaps with hindsight the sample should have been smaller, but the task has been completed. The task was to model 55,000 houses over 50 years in ten year steps, for 625 different scenarios, as sum total of around 172 million calculations and 9 GB of data.