Author:

Abstract:

The residential sector accounts for approximately 30% of the energy consumed in developed countries. This demand is currently covered not only by fossil fuels but also renewable energy sources that ensure a reduction in polluting emissions but which are generally distributed, generate intermittently and are difficult to manage. This requires the development of energy policies that reduce global consumption, as well as control and management systems that target the final consumer. In order to deal with this issue a detailed knowledge of the consumers’ behaviour is needed, both at an aggregate level for the management of the system and at an individual level for the development of measures to adapt their consumption. Furthermore, in this novel context, the feasibility of the different available strategies must be studied in addition to the benefits that can be obtained from their implementation and the control measures that can be developed. This PhD Thesis addresses the development of an energy modelling system for the residential sector as a way of predicting the electricity demand in households and establishing demand response strategies, energy policies and control actions that ease the integration process of distributed energy resources accordingly. The selected modelling technique follows the so-called bottom-up methodology, which enables the consumption in the residential sector as the sum of the individual contributions of each device installed in each household to be obtained. In addition, the simulation of these profiles is carried out using stochastic techniques that allow the heterogeneous and unpredictable behaviour of residents to be reproduced with a high temporal resolution. The modelling system has been divided into three main components which include the consumption due to lighting systems, the heating and air conditioning devices demand and the general appliances consumption. This has facilitated a detailed study of different energy saving policies and the assessment of potential demand response strategies, as well as the development of novel energy management techniques. All of these measures together with the modelling system have been implemented in a simulation tool which was also provided with renewable production data, collected in actual installations. Therefore, not only has the consumption been studied on its own, but also the integration of various resources has been assessed. Some of the studied measures are: replacing devices with more efficient technologies in the case of lighting systems, implementing low-level demand response strategies for household appliances, studying the impact on the low-voltage grid of increasing installation rates of certain technologies such as air conditioning systems and developing novel control techniques in the context of a smart community that can improve the hosting capacity of renewable solar production. Finally, the models and strategies studied in this work have been combined with an advanced metering infrastructure under the umbrella of a smart building. In this context, they provided an additional source of information towards the digitalisation of the electrical system where the extensive use of data allows for the implementation of even more advanced control strategies and will undoubtedly lead to future developments under the paradigm of Smart Grids.