Sensitivity analysis for investigating the energy performance of a retrofitted kindergarten under different weather scenarios
MetadataShow full item record
Sensitivity analysis is a technique that aims at estimating how the uncertainty in the independent variables of a mathematical model effects a particular dependent variable given a predefined set of assumptions. This method can be used in building performance analysis for effectively assisting the model, useful to support the design and assessment of high-performance buildings. This thesis is and investigation on the use of sensitivity analysis to support building performance analysis, with regards to a high-performance building model. The project is divided in two parts. The first part aims to develop the necessary programming infrastructure for running a global sensitivity analysis, and buildings a general platform to automatically perform it. This part investigates the computational efficiency of the simulation approach, the numerical instabilities and technical issues concerning the global sensitivity analysis. The second part pivots around a case study that is used to test and fine-tune the platform on a high-performance building model compliant with the new Norwegian guidelines for calculation of energy performance of buildings named NS3031 (TEK-10). A key aspect of the investigation is to run the simulation for a series of different weather files to investigate the implications of local climate variations and of future weather projections on the energy performance of the case study through the sensitivity analysis. The outcome of the analysis is presented using sensitivity indices that quantify the individual impact of the input variables on the energy performance of the building model, for each of the climate scenarios. The outcome of the development phase is a general platform for running a complete sensitivity analysis through the whole-building and dynamic energy simulation software IDA-ICE. The main findings of this project are an over-all importance of the energy system side of a building on energy performance of the building, while the building envelope showed a lesser impact than expected that is explicable due to the high requirements foreseen by the new Norwegian TEK-10. Regarding the different weather scenarios, the sensitivity analysis shows that the same set of variables are the most sensitive for the building s energy use in present, future and local climates, even though the magnitude of the effect is different in all the tested scenarios. According to the finding of this project, the new Norwegian standard TEK-10 offers a robust set of design requirements against specification uncertainty under future weather projections and in two different Norwegian climate zones. Furthermore, such TEK-10 requirements appear to have already optimized with respect to robustness the main design variables related to the building envelope and, hence, the design variables that affect the most the energy performance of TEK-10 compliant buildings are those related to the energy systems.