[Zurück]

N. Ghiassi, F. Tahmasebi, A. Mahdavi:
"Harnessing buildings´ operational diversity in a computational framework for high-resolution urban energy modeling";
Building Simulation, 2017 (2017).

(only in english language available) To achieve computational efficiency, efforts toward developing urban-scale energy modeling
applications frequently rely on various domain simplifications. For instance, heat transfer phenomena
are captured using reduced order models. As a consequence, specific aspects pertaining to the
temporal dynamics of energy load patterns and their dependency on transient phenomena (e.g.,
weather conditions, inhabitants´ presence and actions) cannot be realistically represented. To
address this circumstance, we have conceived, implemented, and documented a two-step urban
energy modeling approach that combines cluster analysis and sampling techniques, full dynamic
numeric simulation capability, and stochastic methods. The paper describes the suggested urban
energy modeling approach and the embedded cluster analysis supported sampling methodology.
More particularly we focus on the aspects of this approach that explicitly involve the representation
of inhabitants in urban-scale energy modeling. In this regard, the potential to recover lost dynamic
diversity (e.g., in computation of temporal load patterns) due to the deployed reductive sampling
is explored. Parametric runs based on stochastic variations of underlying building use profiles
facilitate the generation of highly realistic load patterns despite the small number of buildings
selected to represent the simulation domain. We illustrate the utility of the proposed urban energy
modeling approach to address queries concerning the energy efficiency potential of behaviorally
effective instruments. The feasibility of the envisioned scenarios concerning inhabitants and their
behavior (high-resolution temporal load prediction, assessment of behavioral variation) is presented
in detail via specific instances of district-level energy modeling for the city of Vienna, Austria.

To achieve computational efficiency, efforts toward developing urban-scale energy modeling
applications frequently rely on various domain simplifications. For instance, heat transfer phenomena
are captured using reduced order models. As a consequence, specific aspects pertaining to the
temporal dynamics of energy load patterns and their dependency on transient phenomena (e.g.,
weather conditions, inhabitants´ presence and actions) cannot be realistically represented. To
address this circumstance, we have conceived, implemented, and documented a two-step urban
energy modeling approach that combines cluster analysis and sampling techniques, full dynamic
numeric simulation capability, and stochastic methods. The paper describes the suggested urban
energy modeling approach and the embedded cluster analysis supported sampling methodology.
More particularly we focus on the aspects of this approach that explicitly involve the representation
of inhabitants in urban-scale energy modeling. In this regard, the potential to recover lost dynamic
diversity (e.g., in computation of temporal load patterns) due to the deployed reductive sampling
is explored. Parametric runs based on stochastic variations of underlying building use profiles
facilitate the generation of highly realistic load patterns despite the small number of buildings
selected to represent the simulation domain. We illustrate the utility of the proposed urban energy
modeling approach to address queries concerning the energy efficiency potential of behaviorally
effective instruments. The feasibility of the envisioned scenarios concerning inhabitants and their
behavior (high-resolution temporal load prediction, assessment of behavioral variation) is presented
in detail via specific instances of district-level energy modeling for the city of Vienna, Austria.

urban energy modeling, GIS, bottom-up engineering model, building stock, occupant behavior, building performance simulation

http://dx.doi.org/10.1007/s12273-017-0356-1