23 Mars – Thesis defense - Lorena de Carvalho Araujo
10 h30 Amphi C - building A29 (university of Bordeaux - Talence campus)
Identification of the intrinsic energy performance of multi-family housing and tertiary sector buildings from in-situ measurements.
Building energy efficiency is a key factor in reducing CO2 emissions and assuring thermal comfort for inhabitants. EU member states are committed to increasing building energy performance to meet the criteria set by the Energy Performance in Buildings Directive (EPBD). Despite the endorsement of building regulations, the as-built energy performance commonly presents a discrepancy with the predicted one, the so called energy performance gap. To close this gap, it is important to have reliable performance indicators to assure new building quality and to estimate the improvements achieved after renovation works. The application of in-situ methods after construction or retrofitting phases enables the measurement of performance indicators, as the whole heat loss coefficient (HLC) and the transmission heat transfer coefficient (HTC). Different in-situ methods for estimating building energy performance are nowadays available with various protocols, mathematical principles and domains of applicability. Among them, the methods relying on fast duration protocol have been mainly conceived for applying in single-family houses. However, multi-family housings and tertiary sector building account for an important part of the building stock, presenting them a relevant potential for energy savings. The current work studies the applicability of a short duration test for identifying the HTC and HLC in large buildings and how to improve the method protocol. The ISABELE method was chosen to be applied to large building typologies. This method was initially conceived to identify the envelope thermal performance of vacant single-family houses. The first challenge encountered for achieving the adaptation objective is related to building dimensions, which hinders the protocol logistics. For facing this problem, two main approaches were considered. The first consists of applying the protocol to the whole building, using the local heating system. This approach is however limited to the conditions of the local system and impacts globally the building normal usage. The second approach is based on the protocol application to parts of the building, where samples of the building envelope have their thermal performance verified. In this case, the main difficulty is related to the heat flow passing through the shared walls. These walls are typically less insulated than the exterior walls, which facilitates the heat flow during an in-situ test and can potentially behave as a noise in the HTC and HLC indicators. Another challenge in this approach concerns the meaning of the final indicator, which is related to just part of the building envelope. Both approaches present their advantages and drawbacks, this is why they have been further investigated to verify their potentials and limits. The investigation work was based on virtual simulation with the use of Pléiades + Comfie for improving the method protocol and studying its limits. Later, both approaches were applied to real buildings to enhance the comprehension of their feasibility in-situ. This work proposes a framework for assessing reliable results of HLC and HTC in large building typologies, based on the ISABELE method. The conclusions of this study are a product of the choices made to face the encountered challenges of adapting this method. As there is a wide space of possibilities to be tested concerning the methods, the building characteristics and weather conditions, many of them were not further studied and constitute part of the outlooks. Nevertheless, similar reasoning could be applied to the adaptation of other methods to large building typologies. The process of adapting a method for the intrinsic building energy performance identification out of its original limits is therefore the main contribution of the present work to the field.