A case study exploring regulated energy use in domestic buildings using design-of-experiments and multi-objective optimisation

Ralph P Evins, P Pointer, Ravi Vaidyanathan, SC Burgess

Research output: Contribution to journalArticle (Academic Journal)

38 Citations (Scopus)

Abstract

The newly-released Standard Assessment Procedure (SAP) 2009 [1] underpins all energy calculations for Building Regulations compliance and Code for Sustainable Homes ratings for domestic buildings in the UK. A newly-developed three-stage optimisation framework is applied to the outputs of SAP for a case study concerning a 2-bed mid-level flat. Firstly a comprehensive full-factorial Design-of-Experiments analysis is performed to determine the significance of each input to the outputs of SAP (carbon emissions, running costs and overheating risk). This allows many of the inputs to be disregarded as non-significant. Next a multi-objective optimisation algorithm is applied to all significant variables to simultaneously optimise regulated carbon emissions versus capital and running costs, constrained by limits on overheating and roof area. Finally a more detailed multi-objective optimisation using greater precision is conducted on all variables that exhibit complex behaviour, i.e. which do not take a single value for all optimum solutions. Information is obtained concerning parameter significance and optimal parameter settings, which is presented as graphical design guidance using the process of ‘innovisation’. This will assist engineers in achieving high-performing, cost-effective designs
Translated title of the contributionA case study exploring regulated energy use in domestic buildings using design-of-experiments and multi-objective optimisation
Original languageEnglish
Pages (from-to)126 - 136
Number of pages11
JournalBuilding and Environment
Volume54
DOIs
Publication statusPublished - Aug 2012

Fingerprint Dive into the research topics of 'A case study exploring regulated energy use in domestic buildings using design-of-experiments and multi-objective optimisation'. Together they form a unique fingerprint.

  • Cite this