Bachelor's thesis Moritz Braun


Sensitivity Analysis of Building Archetypes

Behaviour of parameter influence Copyright: EBC Behaviour of parameter influence for variations of the basic parameter “year of construction” for office buildings

The identification of the dynamic demand behaviour of buildings is used to understand relations between buildings and their energy supply. The presented study examines the heat demand of non-residential buildings. In order to assess the dynamic demand behaviour of buildings, generating thermal simulation models is necessary. A sustainable energy supply system requires a focus on clusters of buildings rather than single buildings. Archetype buildings have proven to be appropriate means to describe large building stocks by representing buildings of similar properties by reference of a single archetype building. Statistical data is used to complete missing information about the topological and physical building structure.

A sensitivity analysis is used in this study to examine those building models derived from archetype buildings. The focus lies on determining model inputs that have a significant influence on the output. In particular, models of office buildings and institute buildings defined by the software tool TEASER are examined. The main difference between both models lies in the composition of usage zones within the building model. The institute building is based on an office building with an additional laboratory usage zone.

Combining accuracy and efficiency, the Morris method is chosen as a suitable method of sensitivity analysis. For each input, lower and upper bounds have to be determined which have a high influence on the calculated sensitivity indexes. In a preliminary screening, eight most sensitive inputs are detected for office buildings and ten for institute buildings, respectively. Those inputs are then examined more extensively by interchanging different sets of basic parameters. These sets are made up from varying values of the year of construction, the net leased area, the number of floors and the height of floors. Hereby, influences by basic parameters on the inputs can be determined.

As a result of this study, the exponent of the equation used to calculate the outer wall area and the set temperature are detected as most sensitive inputs with regards to the heat demand. Regarding institute buildings, the zone area factor related to the laboratory usage zone turns out to be most sensitive. Zone area factors indicate the proportionate area of the corresponding usage zone. The previously mentioned exponent used to calculate the outer wall area shows to be highly influential for the heat demand of institute buildings, too. Moreover, improvement potential for parameter implementation in TEASER can be derived, especially regarding the used building age groups.