Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
Energy demand for cooling of buildings is projected to drastically increase in the next decades. Urban areas in tropical Asia are expected to contribute a major share to this growing demand. These areas are characterized by a high urban density and a diverse functional mix. On the one hand, a high urban density implies a high energy demand intensity. On the other hand, these characteristics present great opportunities to implement efficient energy technologies at building and district scales with the goal of reducing energy consumption. The goal of this thesis was to develop a method for decision support to design new cooling supply systems. These systems combine the opportunities of two innovations. One innovation is the transformation of building HVAC systems from centralized all-air systems to low-exergy building cooling systems. This type of advanced systems, is referred to as hybrid cooling systems, decouple the sensible and latent cooling processes. They are particularly efficient in hot and humid climates. The second innovation is the use of optimization methods for the energy supply system design at the district scale. New cooling supply systems need to be modeled and optimized to effectively harness the efficiency potential of hybrid cooling systems. The main challenge lies in the complex design decisions spanning across building HVAC systems, centralized and decentralized cooling supply systems, and the need for a fully flexible network design. This flexible network design should consider the whole spectrum from fully centralized to fully decentralized system configurations. The here proposed design and decision support framework consists of two steps. The design step combines three novelties in an optimization model. First, hybrid cooling system models were developed. The cooling energy demand of buildings was defined with temperature and enthalpy profiles, which decoupled the need for latent and sensible cooling supply. Second, a fully flexible network design algorithm, covering the whole spectrum from fully centralized to fully decentralized configurations, was developed. And third, the optimization model is able to freely select operating temperatures of cooling supply and hybrid cooling technologies. The model is then able to generate numerous feasible design solutions. In the decision support step of the framework, analysis methods were used to identify a limited numbers of promising designs. The analysis method considered different performance indicators and uncertainties in the market conditions. In the end, this framework provided three robust and energy efficient design recommendations for a case study in Singapore. These solutions comprised two types of decentralized systems with different network temperatures and one type of system, in which all buildings are disconnected. The design solutions were able to achieve up to 35% savings in operational expenses while reducing the capital expenses by 20% compared to the conventional cooling systems. Finally, it can be concluded that integrating hybrid cooling systems into a district energy system can be a cost-efficient contribution to the decarbonization of urban energy systems in the tropics. The step-wise approach provides reasoning of why certain design solutions perform better than others. These insights can assist building owners, urban planners, and district cooling system operators to understand the advantages and limitations of each design.
, ,
,