English

Energy models are used in the residential sector to determine the baseline energy consumption and to predict the future energy demand. They represent a useful tool for evaluating energy saving measures and the effects of different CO2 emission reduction strategies. However, the results of the modeling are subject to multiple sources of uncertainty which are mainly related to the input parameters. In this work, the uncertainty of seven residential energy models is analyzed.

Buildings are the main consumers of electricity across the world. In the electricity system, it is critical to have a realistic forecast of buildings’ demand for adequate power planning and management. However, in the research and studies related to building performance assessment, the focus has been on evaluating energy efficiency of buildings whereas the instantaneous power consumption of systems has been overlooked.

This paper presents a new methodology of machine aided HVAC system optimization with currently available building simulation and optimization frameworks. The method consists of a two part approach using Modelica for accurate but simple system simulation as well as Matlab for automated model configuration, result evaluation and parameter variation.

The paper focuses on control optimization of nonlinear system models where the differential-algebraic equations are formulated in Modelica. Two optimiziation approaches utilizing (i) direct search methods and (ii) derivative-based methods are compared. The approaches are applied to a thermohydraulic low-exergy heating and cooling system model in form of a borehole heat exchanger, a heat pump, a hydraulic distribution system, thermally activated building systems (TABS) and a simplified building model.

In previous studies, the authors combined load shifting and Model-based Predictive Control (MPC) for the optimization of a HVAC system. The precision of the predicted model outputs (primary energy, comfort in temperature and peak heat flow rate) depends on the accuracy of several factors (model parameters, inputs and initial state).

This research aims to define a modelling approach to simulate District Cooling Systems (DCS). A model of the network has been developed using the equation-based object-oriented language Modelica. This model includes a cooling production plant, a distribution network of pipes and 6 substations. This integrated modelling approach allows us to study interactions between substations cooling demand and cooling production plant efficiency. Hourly measurements from Eastern Paris DCS are used as inputs for cooling demand. A simplified model of substations with ideal control has been developed.

A simulation model is developed that allows to investigate the infection risk for Legionella Pneumophila in the design phase of a DHW system and to test the effectiveness of disinfection techniques on an infected system. With the thermodynamic model, the Legionella P. infection risk of the DHW recirculation loop in a case study building is assessed and important components for an optimization study on the trade-off between infection risk and energy efficiency are identified. 

The major goal of this study is to demonstrate the feasibility of the dynamic modeling and simulation of both conventional and direct exchange geothermal heat pump applications particularly with regard to performance evaluations. Therefore, in this research, dynamic models for the simulation of geothermal heat pump systems with the working fluid propane for the application in building-scale energy systems have been developed based on the Moving Boundary approach and the challenges of dynamically evaluating these kind of energy systems have been hereby met.

This paper presents a new approach to calibrate air handling unit models. This approach studies every heat exchanger component separately based on the inverse problem framework, the Preisach model of hysteresis and machine learning techniques. For each component model, the first step is to solve the inverse problem in order to calculate the optimal control signal that generates the output values expected from real data.

In this paper, four different data-driven algorithms including AutoRegressive with eXternal inputs (ARX), State Space (SS), Subspace state space (N4S) and Bayesian Network (BN) are evaluated and compared using a case study of predictions of Air Handler Unit (AHU) thermal energy consumption. Training and testing data are generated from a dynamic Modelica-based AHU model.