A city scale physically disaggregated bottom-up energy model

By Miroslava Kavgic

Supervisors:

Dejan Mumovic

Alex Summerfield 

2007-2013

This thesis describes the development of the first domestic energy and carbon model in Serbia which has been used to investigate the technological feasibility of achieving space heating energy consumption and associated CO2 emission reductions within Belgrade’s housing stock by 2030. BElgrade’s Domestic and Energy and carbon Model (BEDEM, in Serbian a medieval city wall) combines external and on-site generated data, the whole building dynamic energy simulation software ‘TRNSYS’, and a generic optimisation program called ‘GenOpt’. 

Whilst this model is primarily demand side orientated, it also considers changes in energy efficiency on the supply side. The BEDEM model has been used to develop five explorative scenarios, namely: a ‘Base Model’, a ‘Demand 1’, a ‘Demand 2’, a ‘Supply’, and a ‘Demand 2 and Supply’ scenario. The overall results suggest that the largest domestic space heating energy reductions could be achieved by combining the energy-efficiency performance upgrade of dwelling fabrics and district heating system seasonal efficiency improvement. Yet, in the shorter-term, the improvement of the district heating system’s seasonal efficiency is the most beneficial measure, whereas dwelling renovation to an ambitious standard at a lower renovation rate provides considerably larger energy and carbon reductions than dwelling renovation to a lower standardat a higher renovation rate. 

While the model is of considerable value as a policy tool, the results of uncertainty analyses revealed that a lack of knowledge of just a few key input parameters generate rather large uncertainty in the model predictions. Therefore, for any recommendations based on model predictions to be of use in policy formation, the models need to be validated against existing data and uncertainties within the model investigated thoroughly and, where possible, quantified.