Buildings are responsible for approximately 40% of the total annual consumption of energy, including provision of lighting, heating, cooling, and air conditioning. Many countries in OECD Europe have enacted measures to improve energy efficiency in the building sector. For example, the European Union (EU) approved a binding legislation, which aims to meet its ambitious climate and energy targets for 2020.
In this context, there is an increasing interest in the use of building energy models to predict their performance. Particularly, the application of simulation-based optimization methods coupled with building simulations is gaining wider interest. In problems requiring complex models and many objective functions, these approaches may be highly time-consuming and produce ultimately results that are difficult to analyze. This work presents a novel approach for optimizing buildings according to several economic and environmental objectives that is based on combining an objective reduction method with surrogate modelling optimization tools.
The capabilities of our methodology are demonstrated through a case study based on a thermal modelling of a house-like cubicle where the thickness of the insulation walls of the building exterior surfaces are defined as variables and optimized with appropriate algorithms. The objective is to minimize simultaneously the cost and several environmental impacts of the building. The cost indicator includes the electricity cost for heating and cooling the building and the cost of the materials for its construction. The environmental impact accounts for the impact of the electricity and the impact of the materials used. This impact is quantified following the life cycle assessment (LCA) methodology, specifically the Eco-Indicator 99 (EI99). This approach considers the impact caused in all the stages in the life cycle of the product being assessed. In this study, rather than considering a single aggregated environmental metric, which is the standard approach, we asses 11 environmental impacts, 10 midpoint indicators related with human health, ecosystem quality and resources depletion and one aggregated endpoint indicator.
Results show that 9 of the 12 objectives are redundant and can be removed while still preserving the problem structure. Among the 3 objectives finally kept, one corresponds to the economic cost, which is directly included in the objective function, while the other 2 are environmental impacts that are highly correlated with the indicators that are removed. Results show that solving the problem by considering only the economic cost along with the widely used Eco-Indicator 99 (which quantifies the aggregated environmental impact) might change the dominance structure of the problem, thereby losing some Pareto solutions.
In addition, the surrogate model reduces significantly the resolution time compared to the standard approach that runs the simulation software in every iteration of the optimization algorithm.
Results indicate that our approach can provide significant economic and environmental improvements (35 % in economic performance and 21 % and 106% in carcinogenic and ionising radiation impacts) respect to the base case (cubicle without insulation).
The methodology presented is intended to promote energy efficiency in buildings while also considering the optimization of the associated environmental impact. This tool can guide decision-makers towards the adoption of more effective regulations for improving the economic and environmental performance of the building sector.