In-Situ Determination of Buildings’ Thermo- Physical Characteristics

Method Development, Experimentation, and Computation

Abstract

Ever since the introduction of energy conversion systems in the built environment, buildings have become responsible for a considerable share of global energy consumption. Many countries have therefore aimed to invest on buildings’ energy efficiency plans to reduce the depletion rate of the fossil resources and the CO2 emissions associated with them. In this context, accurate determination of building’s thermo-physical characteristics is a necessity in the processes which lead to execution of energy conservation strategies in existing buildings. These characteristics are the essential inputs for buildings’ thermal modelling, quality control, energy audits, and energy labelling, the results of which are determinant for energy renovation decisions and policies. In practice, the values of these parameters are not always available because the current determination methods are time-and-effort-expensive, and consequently rarely used. In accordance with the large deviations observed between the in-lab and in-situ thermal behaviour of building components, a special attention is laid on in-situ methods. This thesis aims at developing and testing different in-situ determination methods and approaches at different levels. Theories, simulations, and experiments, are combined for determination of a number of buildings’ most important thermo-physical characteristics.

Transmission losses through the façades are known to be responsible for a significant portion of heat loss in buildings and consequently are investigated in all standard energy calculation methods. Thus, the major part of the thesis is dedicated to the thermal behaviour of exterior walls. The exact construction of existing walls is generally unknown. Consequently, the estimation of their thermal resistance, thermal conductivity, and volumetric heat capacity can be erroneous. Later, the attention is upscaled to the building level where rather than local characteristics, global characteristics are determined.

At the first stage, the walls’ in-situ determination of thermal resistance has been examined. Despite the advantages of the existing standard method, “ISO 9869 Average Method” for measuring this parameter, two problems have been pointed out: long duration and imprecision. Accordingly, this phase describes and demonstrates how the simplest modifications to this standard method can improve it in terms of solving these problems. Heat transfer simulations and experiments in a variety of wall typologies have been applied to show the effect of using an additional heat flux sensor, facing the first one, installed on the opposite side of the wall. Three estimations of thermal resistance based on either indoor or outdoor heat fluxes, and the average of the two values are then defined. It is shown that one of these values satisfies the convergence criteria earlier than the other two, leading to a quicker insitu determination of thermal resistance with a higher precision.

To further shorten the measurement period, in the second phase, a new transient in-situ method, Excitation Pulse Method, EPM, is developed and examined experimentally on three walls. The method is inspired by the theory of thermal response factors. In EPM, a triangular surface temperature excitation is applied at one side of the wall and the heat flux responses at both sides are measured and converted into the wall’s corresponding response factors which then leads to the wall’s thermal resistance. To validate, the results are compared to the ones obtained following the ISO 9869. The good agreement of the results confirms the possibility of measuring the Rc-value within a couple of hours. Applying this method, the overestimation of around 400% between the actual and estimated values (in practice, often based on the construction year) of thermal transmittance was resolved. Thus, EPM is believed to significantly improve the required time and accuracy in determination of the thermal behavior of walls with unknown constructions. Experimental and practical details regarding the design and construction of the method’s prototype as well as its application range are demonstrated subsequently. EPM has been patented in the Dutch patent office (Patent No. 2014467) and can be applied on in-lab and in-situ circumstances.

Following the success in the proof of principle, in the third phase, detailed conditions for correct application of EPM in heavy and multi-layered walls are further studied. Heat transfer theories, simulations, and experiments are combined to evaluate the method’s performance for different types of walls. A specific attention is devoted to the relationship between the walls’ thermal response time and the response factors’ time interval, affecting the accuracy of Rc-value determination. Additionally, other hidden information in the response factors of the walls such as the possible construction are revealed. It is moreover demonstrated that in addition to the thermal resistance, the two main thermo-physical properties of a wall, the thermal conductivity and the volumetric heat capacity, as well as the wall’s thickness can be determined using inverse modelling of the Response Factors. The accuracy and precision of the method is tested in many different ways, fortifying the confidence for future application of this method.

In the last phase, the advancement of smart metering and monitoring systems in buildings are considered. Such smart technologies have led to utilization of the data from, for instance, home automation systems. This data acquisition is referred to as “on-board-monitoring” category of measurements, which removes the hassle, cost, and intrusion associated with locally-conducted experiments. The problem is then observed from a perspective wider than the component level. This time, the thermo-physical characteristics are studied for a whole building rather than just the walls. It is presumed that the current and future houses and their HVAC installations are by default, equipped with basic sensors, providing on-board monitored data. Therefore, the expected available data is measured and used as input parameters. A case study of an occupied apartment, in which air temperatures, humidity, and CO2 concentrations, gas consumption, and meteorological data have been measured for one year is investigated. Global characteristics such as the heat loss coefficient and thermal capacitance are estimated through inverse modelling of a 1st order circuit analogous to the thermal model of the building, and fed by the measurement data. In addition, using construction information, winter daily air change rates leading to ventilation and infiltration heat losses are estimated from the results of the inverse modelling. These results can be used to tailor the energy efficient use of the building.

In summary, the in-situ determination of walls’ thermal resistance is conducted by two methods in this thesis. The first one calls for longer measurement methods (minimum three days), but includes a straight-forward, well-known procedure. This method is highly suitable for high temperature gradients across the wall. The second method, EPM, requires more complicated instrumentation, but in return, in addition to rapid (couple of hours) determination of the Rc-value, it provides the walls’ response factors which are required for a dynamic thermal building simulation. In addition, using the results of this method, the thermal conductivity and volumetric heat capacity can be determined. EPM is most suitable for light-to-medium weighted walls and for homogeneous walls of known thickness. Stable heat flux profiles at the surfaces of the wall increase the accuracy of the method, especially when the temperature gradients across the wall are lower. Finally, as a less intrusive approach, the data from the HVAC installations’ existing sensors can be used. Global characteristics including the heat loss coefficient and the global capacitance can be then determined for a whole building, followed by ventilation and infiltration losses. Despite the low accuracy, this process is more suitable when the smart meter data is available and measurements at component level are not desired.

By introducing and testing new experimental and computational methods and approaches for reliable determination of buildings’ local and global thermo-physical characteristics, this thesis pays a significant contribution to the accuracy of the energy-related predictions and operations, especially within the built environment.