Development and Assessment of Simplified Building Representations in the Context of an Urban Energy Model

Abstract

For buildings in arid climate area with significant cooling loads, predictions of long term building cooling load is of critical importance. However, numerous challenges and limitations are incorporated in the development of building energy models. The aim of the present study is to evaluate the accuracy that can be achieved using a simplified building energy model in comparison to the one that can be achieved using a detailed building energy model in the framework of an urban energy model. The urban energy model under development attempts to reproduce the energy consumption of the city of Abu Dhabi, including buildings and microclimatic phenomena. Due to climatic specificities of Abu Dhabi, the urban energy consumption is dominated by the air-conditioning load, itself exacerbated by the heat island effect. Developing such a model based on detailed representations of all the buildings is unpractical. The study will assess the loss of accuracy incurred by the implementation of different types of simplified building representations: RC analog, time series (ARMAX), neural network, and a single-zone EnergyPlus equivalent. The goal of the simplification is to reduce the amount of time and effort invested in the modeling task while maintaining a sufficient and valid representation of the actual urban system behavior. The reduced order models will be calibrated against a detailed building energy model using optimization techniques and a sensitivity analysis in order to determine the most influential parameters on the energy consumption. In practice, the simplified modeling is achieved through the tuning of a low-order model so that it is close enough to the detailed model. The urban energy model will be implemented using a coupled scheme incorporating a building energy simulator which in our case is Energy plus (EP) along with a dynamic lumped-parameter urban canyon model simulated on MATLAB. The main use of the coupled scheme is to validate the accuracy that can be achieved by an urban canopy model and its ability to accurately determine the temperature elevation inside the urban canyon and the energy consumption due to the heat island effect. The results of this study reflected a very satisfying performance of the RC model relatively to the other developed simplified models. Furthermore, the study shows that the use of simplified energy models, in the context of an urban energy mode, is worthwhile and results in limited loss of accuracy, when compared to detailed models.

Date of Award	May 2015
Original language	American English
Supervisor	Afshin Afshari (Supervisor)