A reduced data dynamic energy model of the UK houses

2018-06-06T11:13:30Z (GMT) by Ali Badiei
This thesis describes the development of a Reduced Data Dynamic Energy Model (RdDEM) for simulating the energy performance of UK houses. The vast quantity of Energy Performance Certificate (EPC) data stored at the national scale provides an unprecedented data source for energy modelling. The majority of domestic energy models developed for the UK houses in recent years, including the Standard Assessment Procedure (SAP) model used for generating EPCs, employ BREDEM (Building Research Establishment Domestic Energy Model) based steady state calculation engines. These models fail to represent the transient behaviours that occur between building envelope and systems with external weather conditions and occupants. Consequently, there is an ongoing debate over the suitability of such models for policy making decisions; which has raised the interest in dynamic energy models to overcome these shortcomings. The RdDEM eliminates the main drawback associated with dynamic energy modelling, namely the large amount of required input data compared to steady-state models, by enhancing a reduced set of data which was originally collected for EPCs. A number of new inferences and methodological enhancements were tested and implemented in the RdDEM using a sample of semi-detached houses. In this way, SAP equivalent input data could be converted automatically for use in dynamic energy modelling software, EnergyPlus. Simulations of indoor air temperatures and space heating energy demand from the RdDEM were compared to those from SAP for 83 semi-detached houses. The comparison was also carried out with more detailed models, on a sub-set of the modelled dwellings. Finally, the predicted energy savings that resulted from energy efficiency improvements of the dwellings were compared and estimated potential for saving energy from the RdDEM was quantified. ii The results show that it is technically feasible to develop dynamic energy models of these houses using equivalent inputs. In the majority of cases, the RdDEM predicted lower indoor air temperatures than SAP, and consequently the energy demands were lower. The RdDEM predicted annual space heating demand to be lower than SAP in 72% of the houses, however the difference was less than 10% in 94% of the houses. The RdDEM predicted slightly higher (< 2%) energy saving potentials compared to SAP when the same set of energy saving measures were implemented in both models. The development of these new methods for automatically creating SAP equivalent inputs from reduced data but for use in a dynamic energy model offers new opportunities for inter-model comparisons as well as a dynamic alternative to the SAP when variations in energy demand and indoor air temperatures are required.