3D Visualization of thermal resistance and condensation problems using infrared thermography for building energy diagnostics
© Ham and Golparvar-Fard; licensee Springer. 2014
Received: 26 August 2014
Accepted: 27 October 2014
Published: 25 November 2014
Building deteriorations instigated by material degradations or moisture intrusions are the primary causes for energy inefficiency in many existing buildings. For choosing appropriate retrofits, it is important to carefully diagnose and analyze building areas in need of improvements. In addition to reliable sensing and analysis of as-is energy performance, an intuitive recording and visualization of energy diagnostic outcomes are also critical to effectively illustrate the as-is building conditions to homeowners during retrofit decision-making processes.
Toward this goal, this paper presents a thermography-based method to visualize the actual thermal resistance and condensation problems in 3D while taking static occlusions into account. First, several overlapping digital and thermal images are collected from the building areas under inspection. Using a computer vision method– consisting of image-based 3D point cloud and mesh modeling algorithmsactual 3D spatio-thermal actual 3D spatio-thermal models are generated where surface temperature can be queried at the level of 3D points. Based on the resulting 3D spatio-thermal models and by measuring the reflected and dew point temperatures, the actual R-values of building assemblies are calculated, and the condensation issues are analyzed. Taking static occlusions into account, (1) the distribution of the actual thermal resistance over each building assembly, (2) the detected building areas with condensation problems, and (3) the corresponding geometrical and thermal characteristics are jointly visualized within a 3D environment.
To validate the method and investigate the perceived benefits, four experiments have been conducted in existing buildings. Surveys are also conducted by professional energy auditors. The proposed method provides 3D visual representation of the actual thermal resistance distributions and building areas associated with condensation issues at the level of 3D points across geometrical forms while taking static occlusions into account.
The experimental results and the feedback received from the professionals show the promise of the proposed method in facilitating systematic post-examination of building deteriorations and support retrofit decision-makings. Ultimately, converting surface temperature data obtained from an IR camera into 3D visualization of energy performance metrics and possible condensation problems enables practitioners to better understand the as-is building conditions.
The emerging energy crisis of the building sector and the legislative measures on improving energy efficiency are steering the construction industry towards adopting new design concepts and construction methods that can decrease the overall energy loads. For example, in order to lower the energy cost required for space conditioning, building practitioners are embracing the concept of net-zero passive houses at the design phase. The use of building materials with higher thermal resistance is recommended for better insulation at the construction phase.
The problem of energy inefficiency is however not limited to the design and construction of new buildings. According to a recent report by the U.S. Department of Energy (U.S. DOE ), around 35% of input energy in existing buildings in the U.S. is still being wasted. A primary source of these inefficiencies is thermal performance problems in the building envelope, which can cause an excess in energy consumption for space heating and cooling purposes. Roth et al. () reported that such faulty behaviors can account for up to 10% of the total energy consumption in existing buildings during their operational phase. Despite their significance, identifying and assessing building performance problems and improving the energy efficiency under such conditions is not trivial.
Today, building practitioners widely use Infrared (IR) thermography to identify sources of energy loss in building envelopes. The IR thermography –which detects and measures heat variations from surfaces in a non-destructive manner– is particularly known for its usefulness for identifying thermal defects. Nonetheless, current thermographic inspection processes have several inefficiencies. For example, thermal images captured using consumer-level IR thermal cameras typically have low spatial resolutions (160 × 120 or 320 × 240 pixel) and small fields-of-view compared to their digital counterparts (>1 megapixel). As a result, sensing thermal performance data for buildings, even a single room, requires large numbers of thermal images (e.g., ~300 to 400 thermal images for around 4m × 5m room based on experiments). Considering significantly large collections of 2D thermal images that need to be manually analyzed for the purpose of assessing thermal conditions, their direct application for practical building diagnostics can be time-consuming and labor-intensive. In order to improve the process of sensing thermal performance using raw 2D thermal images, several studies (Borrmann et al. ; [Ham and Golparvar-Fard 2013]; Lagüela et al. ; Wang et al. ) have recently focused on developing methods for 3D thermal modeling of building environments by leveraging digital photogrammetry or laser scanning, and promising results have been reported.
These methods for 3D thermal modeling provide a detailed representation of the actual temperature deviations on building surfaces, which is very effective for energy diagnostics. Nevertheless, the problem of characterizing and visualizing performance metrics and thermal problems from these models for post-inspection is still unexplored. The analysis of abnormal thermal regions – either from 2D thermal images or 3D thermal models – is still primarily based on the interpretation of the measured surface temperatures such as hot or cold spots. Yet, the surface temperatures alone do not directly represent the energy performance problems. Due to a lack of a benchmark, the interpretation of energy performance problems is still primarily based on the auditors' experiences and knowledge, which can cause the outcomes to be subjective or inconsistent. Considering a limited number of well-trained professionals, it is also likely that auditors' knowledge and experience would adversely impact the quality of the inspection.
Another important issue – which is often not considered– is that the deterioration rate of building assemblies typically vary even over small surface areas. For instance, when it comes to building envelopes, various regions across the surface may suffer from different levels of degradation. As such, their conditions would change differently over time. Therefore, a method for building diagnostics should assess and visualize the as-is building conditions at point-level across geometrical forms of the building elements so that defects can be precisely detected and localized.
The presence of static occlusions in existing buildings has also raised concerns about the applicability of thermography-based methods for building diagnostics. Ideally, for accurate measurement of surface thermal performance, any objects around the building interior surfaces need to be removed prior to thermographic inspections. However, in practice, for most in-use buildings, it is not trivial to take away all objects that block the line of sight due to the immobility of objects or space limitations. Theoretically, there is no way to measure the thermal performance for the building areas that are occluded by objects, except for encouraging practitioners to remove as many objects as possible before thermographic inspection. Thus, this issue remains as an open challenge.
To address these challenges, this paper presents a new interactive method for 3D visualization of the actual R-value of building assemblies and condensation problems. The proposed method builds upon the authors' prior computer vision research on 3D spatio-thermal modeling using large numbers of unordered thermal imagery collected by a consumer-level IR camera [(Ham and Golparvar-Fard 2013)]. Beyond synthesizing raw 2D surface temperature data, the method extracts and visualizes building energy performance indicators from the collected thermal imagery. It facilitates the process of identifying potential performance problems by allowing auditors to spend less time on analyzing large amounts of thermal images, and instead focus on investigating the causes of the problems and analyzing various retrofit alternatives.
The primary technical contributions of this paper are two-fold: (1) a method for 3D visualization of energy performance metrics and moisture issues at the level of 3D points across geometrical forms of building elements; and (2) a method for minimizing the adverse impact of static occlusions on thermographic inspection in real-world building environments. In the following sections, previous works on measuring R-values of building materials using thermal imagery and analyzing condensation problems are reviewed, and their challenges in applying to building condition assessment are discussed. Next, the research objective, underlying environmental assumptions, and the proposed visualization method are presented in detail. Finally, experimental results on four building assemblies in two case studies and the responses from domain expert evaluations are discussed.
Thermal resistance (R-value) of building assemblies
Analyzing heat transfer conditions through building envelopes is now a required step in rating the energy performance levels of existing buildings (Fokaides and Kalogirou ). This is because during the operational phase, the thermal resistance of building materials gradually decreases. Thus, the actual R-value of the building assemblies with thermal defects is typically lower than the notional value declared by their manufactures. Declining R-values of the building materials results in unnecessary heat transfer through building façades. This in turn increases the operational frequency of heating and cooling systems for space conditioning purposes. Hence, beyond using building materials with high R-value during the design and construction, it is important to explore how the as-is thermal resistance of building assemblies lowers over time.
Recently, several research groups have focused on the problem of non-destructive measurement of the actual heat transfer conditions of building assemblies using thermal images. Madding's work (Madding ) is one of the earliest studies that propose a method for analyzing thermal images to estimate the actual heat transfer conditions for building envelopes. Based on the environmental assumption of a steady-state heat transfer condition of building environments, this work measured the R-value of drywall assemblies using the indoor surface temperature data obtained from thermal images. The measurements were experimentally conducted in controlled lab environments, and then they were compared with the theoretical expected properties. Likewise, based on the similar environmental assumptions, methods proposed in (Albatici and Tonelli ; Dall'O' et al. ; Fokaides and Kalogirou ) measured the overall heat transfer coefficient of the building assemblies. For validation, several experiments were conducted for a few standard wall assemblies. In these studies, the difference between the measured values using thermal images and the notional values was reported to be in the range of ~10% (Madding ), 10-20% (Fokaides and Kalogirou ), and around 15% for cavity walls (Dall'O' et al. ).
For quantifying the as-is thermal resistance of building assemblies, the current non-destructive method regulated by the international standard is to position heat flux meters (HFMs) –a transducer that produce an electrical signal proportional to the heat rate– on the representative building areas and directly measure the heat flow (International Organization for Standardization (ISO) ). However, considering possible variations of the thermal conditions even over small surface areas, it is difficult to ensure that the thermal resistance measured by using a few HFMs will actually represent the entirety of an existing building envelope (Albatici and Tonelli ; Dall'O' et al. ). Interestingly, studies such as (Albatici and Tonelli ) and (Dall'O' et al. ) report 40-80% and 40-175% error between the measurements of the overall heat transfer coefficient using HFMs and their theoretical expected values. Although these observations are primarily based on limited experiments, when one considers the significance of these errors, the applicability of thermal imagery for quantifying the actual heat transfer condition of building assemblies becomes more attractive.
Despite such promising benefits of using thermography, the direct application of 2D thermal imagery as used in prior works is still challenging. In prior studies, a single or at most a few temperature data were extracted from the designated areas captured in 2D thermal imagery, and then a single measurement of the actual heat transfer condition was performed for each building assembly. Hence, these methods assumed that the obtained thermal resistance is representative of all spots on the inspected building assemblies. However, because of different deterioration rates, the actual heat transfer conditions may vary over small scales across geometrical forms. Thus, single measurements will not represent the dynamic variations in the actual R-values that are typically caused by non-uniform deteriorations in the building materials. More importantly, interactive 3D visualization of these dynamic variations in the context of the building under inspection will provide an opportunity for practitioners to better understand the as-is heat transfer conditions.
Condensation problems in building environments
In building environments, water vapor condensation happens wherever exposed building surfaces are at a lower temperature than the dew point. Typically, it is not easy to detect small amounts of the condensed water on building surfaces with the naked eye. Failure in detecting the onset of such condensation issues causes the condensed moisture to accumulate over time. Then, molds and dust mites may build up, which ultimately has potential for causing allergic reactions among the building occupants. In addition, the resulting wet surface may cause flaking paints and peeling wallpapers. This is when the building occupants are more likely to detect the problems with their naked eyes. In building environments, windowsills or thermal bridges around corners of the walls typically have low temperatures. Thus, energy auditors need to proactively analyze these areas whether the surface temperature has reached to the dew point.
Several studies have focused on analyzing the condensation phenomenon on building façades and the risk associated with the occurrence. Bellia and Minichiello analyzed the thermal and moisture performance of building assemblies, and focused on surface condensation issues of building façades and interstitial condensations within multi-layer walls (Bellia and Minichiello ). Aelenei and Henriques investigated the external environmental conditions that cause condensation problems in building environments (Aelenei and Henriques ). They concluded that the thermal convection effect and the amount of moisture in the air are the most influential factors in forming condensation problems on building façades. These prior works primarily examined the sources that cause condensation and identified ways that can reduce the risk associated with these problems. Nonetheless, devising energy auditing methods that can easily detect condensation problems in building environments is still an open research area. Today, by comparing the surface and dew point temperature, thermal cameras enable 2D detection of those building areas that suffer from condensation issues (Hoff ). These detections are typically done on a single image displayed on the screen of the thermal camera. Because these images are not geo-tagged, it is often not easy to figure out at a later stage which building areas are associated with the detected condensation problems. Moreover, these 2D assessments do not provide any information about the surface area of the detected condensation problems. Hence, it is difficult at a later stage to quantitatively estimate the significance of their impact.
Overview of the proposed method
By using digital and thermal images, the proposed method first generates the 3D spatio-thermal mesh models which contain the surface temperature readings at the level of 3D vertices. Then, with environmental measurements, actual thermal resistances are calculated, and building surface areas suffering from condensation problems are detected at the level of 3D vertices in the mesh. Finally, taking possible static occlusions into account, the distribution of the actual R-value, the detected building areas with condensation problems, and the corresponding geometrical and thermal characteristics are interactively visualized within a common 3D environment. In the following sections, the underlying environmental assumptions, experimental setups, and each step of the proposed method are presented in detail.
Environmental assumptions and experimental setups
The proposed method is based on the following environmental assumptions: (1) the indoor building environment is assumed to have a quasi-steady-state condition of heat transfer during the thermographic inspection process. This is consistent with all recent works on non-destructive measurement of thermal properties (Albatici and Tonelli ; Dall'O' et al. ; Fokaides and Kalogirou ; Ham and Golparvar-Fard ; Madding ). For achieving a quasi-steady-state condition of heat transfer, the building envelopes were not exposed to direct solar radiations and wind loadings as the inspections were conducted before sunrise on a non-windy and cloudy day. This minimizes the influence of the convective heat loss and the surface temperature rise on the exterior of building envelopes. As such, the possible increase in building surface temperatures caused by the release of the absorbed solar energy throughout a day was minimized; 2) it was further assumed that the heat transfer between the indoor surfaces and the thermal camera lens is due to thermal convection and radiation. This assumption is consistent with prior works (Fokaides and Kalogirou ; Ham and Golparvar-Fard ; Madding ). In addition, the building interior spaces were heated for a few hours prior to thermographic inspections, which allowed to clearly capture the as-is heat transfer conditions through building façades during the winter season.
Technical specification of the thermal camera used for data collection
Camera technical items
Built-in digital camera resolution
2048 × 1536 pixels
320 × 240 pixels
2°C or ±2% of reading
Reconstructing 3D building thermal performance using 2D thermography
Here, different from previous works [(Ham and Golparvar-Fard 2013)] and (Golparvar-Fard and Ham ), and to better deal with static occlusions in building environments, a threshold is used to identify those thermal points that have a certain distance from the geometrical baseline mesh. It is noted that the thermal points may belong to objects on or around the wall, not the building assembly itself. Thus, when searching for the nearest thermal points to each vertex in the baseline mesh, a geometrical threshold based on the registration error of the point cloud and the baseline mesh is applied. At a later stage, the points from non-relevant objects are discarded and are not used for calculations. The details will be discussed in Section (Considering static occlusions in 3D visualization of the as-is building conditions).
Measuring actual R-values at the level of 3D points
Within the resulting 3D spatio-thermal mesh models, Tinside, wall can be queried from each 3D vertex. Here, different from prior works, this study focuses on measuring actual thermal resistance of building assemblies while taking possible static occlusions into account as discussed in the previous section.
Detecting building condensation problems in 3D
The indoor temperature as well as the relative humidity are measured with the use of a thermo-hygrometer as part of the thermographic inspection. In this work, the building surface temperature (Tinside, wall) can be queried at the level of 3D vertices from the generated 3D spatio-thermal mesh models. Thus, the building surface areas that are likely to suffer from condensation problems are directly detected by comparing Tinside, wall and T D at the level of 3D points. Here, the building surface areas relevant to condensation issues can be calculated from the mesh model. For each triangle face consisting of the three thermal points that their thermal values are below T D , the 3D coordinates of the vertices are queried, and then the area is calculated by using the cross product of two corresponding vectors that form the triangle face.
Considering static occlusions in 3D visualization of the as-is building conditions
Where, s, T, and R indicate the scaling factor, translational offset, and rotation of 3D spatio-thermal models respectively. The resulting registration error (e) is then used as a threshold for detecting thermal points that are likely to represent static occlusions. Among all Tinside, wall that can be queried from each 3D vertices within the 3D spatio-thermal mesh, those thermal points belonging to non-relevant objects (i.e., above the threshold (e)) are removed as outliers and are not considered in the calculation of the thermal resistance of building assemblies and detecting condensation problems. Once the relevant-to-building geometry thermal points are queried from 3D spatio-thermal mesh models, the actual R-values of the building assemblies under inspection are calculated by using Equation (4), and the building areas associated with condensation issues are detected by using Equation (5).
- (1): a set of relevant-to-building geometry 3D mesh vertices where encapsulates the 3D location , the color value of each vertex which corresponds to the measured R-value according to the given color spectrum, the color value of each vertex which is a metaphor based on traffic light colors to represent problematic and non-problematic areas using red and green colors respectively, and finally the color value of each vertex associated with a temperature reading averaged from all thermal images that have observed the vertex
- (2): a set of 3D points where encapsulates the 3D location and the color of each point in 3D building geometrical point cloud which is the average from the colors in all digital images that observed the point
- (3): a set of thermal cameras representing the thermal camera projection matrix based on the relative pose with respect to the built-in digital camera
- (4): a set of built-in digital cameras representing the digital camera projection matrix for each camera registered in a reconstructed 3D scene
A set of vertices that form each face
The number of vertices and faces in a mesh
The number of feature points and cameras that observe the points
Mapping relationship between points and cameras that observe the points
Results and discussions
Experimental results of 3D thermal modeling and environmental measurements
3D thermal modeling
# of 2D Thermal Images
# of 3D Thermal Points
# of Vertices in Baseline Mesh Model (wall)
# of Vertices excluding static occlusions in Mesh Model by using Threshold (wall)
Computational Time for 3D Thermal Modeling
Outside air temperature (°C)
Indoor air temperature (°C)
Dew point temperature (°C)
Relative humidity (%)
Calculated metrics in 3D
Error on registration (mm)
Average (µ) of R-values of Wall (m 2 K/W)
Standard Deviation (σ) of R-values of Wall
Range of actual R-values by considering 95% confidence
Average (μ) of R-values of Window (m 2 K/W)
Standard Deviation (σ) of R-values of Window
Range of actual R-values by considering 95% confidence
The timing of an inspection would impact the capability of detecting condensation issues. To continuously monitor changes in the environmental conditions and study their impacts on condensation problems, one can place environmental sensors on building interior surfaces and continuously monitor the relative humidity and also the surface and inside air temperatures. However, in practice, only a few environmental sensors could be placed per element. As such, considering the variations of thermal performance over surface areas, it is difficult to ensure if the surface condition measured by a few sensors is representative of the entirety of the building assemblies. Here, the proposed method has potential to provide a solution for selecting the proper locations of installing a limited number of environmental sensors for detecting condensation problems. As a complementary step, after detecting the condensation problems from the proposed one-time thermography-based method, the environmental sensors can be placed at the detected areas for continuous monitoring. This will allow proper positioning of a limited number of sensors and improves the efficiency of monitoring.
Domain expert evaluation and discussions
To discuss the perceived benefits of the proposed method, face-to-face and phone interviews were conducted with eleven actual domain experts. The subjects had 2 to 10 years of practical experiences in energy auditing of existing buildings. They were first asked about the current best practices and bottlenecks in thermographic inspections for energy diagnostics in existing buildings. All subjects responded that today, thermography captured from existing buildings is considered the most effective tool for recording and communicating the as-is energy performance conditions to building owners. However, thermography is primarily used for qualitative documentations in support of leveraging the Home Energy Rating System (HERS) for reporting purposes at a later stage. Most responders used the word "qualitative" to describe the current process. For example, they noted that "Audits can take a long time, up to several hours for large residential houses. IR imaging is done piece-wise through a building. It takes detailed documentation to know where each image was taken and what it is supposed to be showing. This can be confusing, and is often detrimental to the client's understanding, even with a good service provider explaining discoveries.", "Sometimes it is hard to tell where the thermal image is coming from in the building. Often, our audits last about 2–4 hours. In this time, it is hard to note which image was taken where within the building."
The survey results show the potential of the proposed 3D visualization method for assessing the as-is building conditions as an intuitive communication tool. Since the main purpose of the proposed method is to provide guidance for making efficient and effective energy retrofit decisions by diagnosing and visualizing the as-is building conditions, the proposed method only needs to be conducted once per regular building diagnostic process (e.g., annual building inspection) and does not need to be performed more frequently. Hence, it does not add significant burden to the building practitioners, and allows the proposed method to be repeatable in parallel with thermographic inspection. The proposed method can inspect very small surfaces with potential performance problems (in fact, at the vertex-level of a mesh). This level of granularity, for example, would be helpful when installing blown-in loose-fill insulation is considered as one of the retrofit alternatives to improve the thermal resistance of the defected areas. However, if the detected areas are too small, retrofitting such problems is not likely to be practically feasible due to technical challenges and long payback period. Thus, practical retrofit decisions need to be taken up to a certain minimum size of retroffitable surface. The proposed method chooses not to dictate the minimum size of the retrofittable building area. Rather, by presenting the measurements at the point-level, the proposed method lets the building practitioners set their desired value prior to assessing different energy efficiency retrofit alternatives.
Difficulty in in-situ measurement of as-is thermal properties for building assemblies: Few participants had concerns about the difficulty in accurate measurement of thermal properties in building environments. Here, following prior works (Albatici and Tonelli ; Dall'O' et al. ; Fokaides and Kalogirou ; Madding ), the deviation between the measurements and the notional values was examined. It is noted that wall assembles are not the best choice for such experiments since wall assemblies typically have multiple layers and are likely to suffer from non-uniform and time-variant degradations. Hence, in this case, using the notional R-values as the ground truth for the as-is thermal resistances may not be accurate. In contrary to walls, since windows are mostly standardized and their thermal properties are more likely to be consistent over time, their actual thermal resistance is less likely to deviate from the notional values derived from the as-built documents. Thus, window components were selected for experiments. It is not a trivial task to accurately estimate the thermal properties of the old building assemblies such as the case studies of this paper. Here, the estimation is based on (ColoradoENERGY ; The Engineering Toolbox ) which provide recommendations for such R-values and (European Standard (EN) ). As a result for single-glazed clear windows, the averaged deviations were 12.81% (= (0.301-0.260)/0.301) for case #1 and 10.37% (= (0.301-0.274)/0.301) for case #2. Following the environmental assumptions presented in (Albatici and Tonelli ; Dall'O' et al. ; Fokaides and Kalogirou ; Madding ), the findings in the experiments are consistent with what was reported in those literatures (Section Thermal resistance (R-value) of building assemblies). Here, systematic errors on measuring the input variables for Equation (5) need to be considered. If the thermographic inspection is conducted in outdoor environments, the measurement accuracy would be affected by various external factors such as wind, sun radiation, and shadows. Since the temperatures are measured from inside surfaces and crumpled aluminum foils located in indoor environments, the accuracy of measuring those input variables is typically the same as it is claimed in the technical specification of the thermal camera. As can be seen in the Table 1, the measurement accuracy of the consumer-level thermal camera used is 2°C or ±2% of reading. Here, another prevalent challenge for accurate measurement in thermal resistances in a non-destructive manner is that the measurement needs to be conducted under steady-state conditions as discussed in Section (Environmental Assumptions and Experimental Setups). Despite significant efforts to form a steady-state heat transfer condition during thermographic inspection (e.g., the timing of data collection and experimental setups prior to image data collection as described in Section(Environmental Assumptions and Experimental Setups), in practice, it is not trivial to maintain a perfect steady-state condition during thermogaphic inspection. More research needs to be conducted on how a steady state condition of heat transfer can be continued during thermographic inspection and how the measurement errors caused by non-steady-state conditions can be accounted for and minimized for more accurate analysis;
Continuous measurement to form time-series of thermal resistances: Besides the measurement accuracy issues, a few responders described `4D visualization' as a desired feature, e.g. "Maybe a 4D visualization done with the presence of all the objects that exist in the building.", "A time series IR webcam of a building over the course of a year." In prior work on measuring actual U-values (Fokaides and Kalogirou ), the results were validated based on different sources of information: instantaneous surface temperature data samples of 10 measurements obtained from a thermal camera and continuous time-series heat flux data collected from HFMs over a period of 168 hours. According to the experimental results, it is shown that the outcomes of the two methods were very close to each other in terms of their accuracies. However, if surface temperature data can be collected continuously, the actual heat resistances can be modeled and explored in 4D (3D + time) as opposed to static information. Visualizing dynamic energy performance metrics in form of a 3D animation can significantly improve the understanding of time-varying heat transfer phenomena. This also enables to provide an opportunity to analyze the impacts of building envelope retrofit decisions. For example, energy auditors can conduct a before/after analysis on thermal resistances when additional insulation layers are being installed. Such analyses would support implementation of periodic building maintenance programs. More research on 4D visualization of energy performance metrics need to be conducted; and finally
Minimizing false negative detection for static occlusions: One survey participant noted that "Accounting for furniture and other obstructions seems like a challenging task". The proposed method of setting a geometrical distance threshold has potential to ignore those building areas with static occlusions. Nevertheless, as discussed in the previous section, false negatives can still occur due to incomplete segmentations. In the experiments, those happen around thin objects on walls such as frames (e.g., Figures 8 and 9(e)) or tiny objects around windows (Figure 10(b) and (c)). This is because the distance between thermal points representing such objects in 3D point clouds and the nearest vertex in a geometrical baseline mesh is most likely to be within the threshold based on the registration error. As a result, the thermal resistance of the occluded building assemblies was not accurately calculated. More research needs to be conducted to better segment non-relevant objects from the reconstructed 3D scenes.
Effective building diagnostic tools should represent abnormal building conditions in an intuitive format to allow energy auditors better communicate the as-is building conditions to the homeowners. Toward this goal, this paper presents a method for visualizing the as-is building condition assessment at the level of 3D points. The outcome of the analyses are visualized within a 3D environment, and static occlusions are accounted for. In this 3D environment, practitioners can navigate and interactively query information about the building geometry, surface temperature data, thermal resistances, and analyze condensation problems. The proposed method was tested in a residential building and an instructional facility, and the following domain expert evaluation was conducted. By converting surface temperature data sensed from a consumer-level single thermal camera into 3D visualization of the actual R-value distributions and the location information of possible condensation problems, building practitioners can better understand the as-is building conditions. Future works include addressing the following challenges in (1) forming and maintaining a steady state condition of heat transfer; (2) studying 4D visualization of energy performance metrics based on time-series of building surface temperatures; (3) minimizing false negative segmentations for static occlusions; and (4) Varying the level of visualization for both global and local inspection of as-is energy performance based on different Levels of Detail (LOD). These are currently being explored as part of ongoing research.
- Aelenei D, Henriques FMA: Analysis of the condensation risk on exterior surface of building envelopes. Energy and Buildings 2008, 40(10):1866–1871. 10.1016/j.enbuild.2008.04.003View ArticleGoogle Scholar
- Albatici R, Tonelli AM: Infrared thermovision technique for the assessment of thermal transmittance value of opaque building elements on site. Energy and Buildings 2010, 42(11):2177–2183. 10.1016/j.enbuild.2010.07.010View ArticleGoogle Scholar
- Bellia L, Minichiello F: A simple evaluator of building envelope moisture condensation according to an European Standard. Building and Environment 2003, 38(3):457–468. 10.1016/S0360-1323(02)00060-4View ArticleGoogle Scholar
- Borrmann D, Elseberg J, Nüchter A: Thermal 3D Mapping of Building Façades. In Intelligent Autonomous Systems 12. Edited by: Lee S, Cho H, Yoon K-J, Lee J. Springer Berlin Heidelberg, Berlin, Germany; 2013:173–182. 10.1007/978-3-642-33926-4_16View ArticleGoogle Scholar
- ColoradoENERGY. (2013). R-Value Table: Insulation Values For Selected Materials. http://www.coloradoenergy.org/procorner/stuff/r-values.htm.Google Scholar
- Dall'O' G, Sarto L, Panza A: Infrared screening of residential buildings for energy audit purposes: results of a field test. Energies 2013, 6(8):3859–3878. 10.3390/en6083859View ArticleGoogle Scholar
- European Standard (EN). (1997). Glass in Building - Determination of Thermal Transmittance (U-value) - Calculation Method (EN 673).Google Scholar
- FLIR system. (2010). Thermographhic Measurement Techniques (FLIR Reporter Professional, Professional Edition, pp. 133–137).Google Scholar
- Fokaides PA, Kalogirou SA: Application of infrared thermography for the determination of the overall heat transfer coefficient (U-Value) in building envelopes. Applied Energy 2011, 88(12):4358–4365. 10.1016/j.apenergy.2011.05.014View ArticleGoogle Scholar
- Golparvar-Fard M, Ham Y: Automated diagnostics and visualization of potential energy performance problems in existing buildings using EPAR -Energy Performance Augmented Reality-models. Journal of Computing in Civil Engineering 2014, 28(1):17–29. 10.1061/(ASCE)CP.1943-5487.0000311View ArticleGoogle Scholar
- Ham Y, Golparvar-Fard M: An automated vision-based method for rapid 3D energy performance modeling of existing buildings using thermal and digital imagery. Advanced Engineering Informatics 2013, 27(3):395–409. 10.1016/j.aei.2013.03.005View ArticleGoogle Scholar
- Ham Y, Golparvar-Fard M: EPAR: Energy Performance Augmented Reality models for identification of building energy performance deviations between actual measurements and simulation results. Energy and Buildings 2013, 63: 15–28. 10.1016/j.enbuild.2013.02.054View ArticleGoogle Scholar
- Ham, Y, & Golparvar-Fard, M. (2014). Three-dimensional thermography-based method for cost-benefit analysis of energy efficiency building envelope retrofits. Journal of Computing in Civil Engineering., [http://ascelibrary.org/doi/abs/10.1061/%28ASCE%29CP.1943–5487.0000406]Google Scholar
- Hoff R: Water condensation – a Temperature Related Phenomenon. 2011. http://www.moistureview.com/blog-034.Google Scholar
- Horn B: Closed-form solution of absolute orientation using unit quaternions. Journal of the Optical Society 1987, 4(4):629–642. 10.1364/JOSAA.4.000629MathSciNetView ArticleGoogle Scholar
- International Organization for Standardization (ISO). (2007). Building Components and Building Elements-Thermal Resistance and Thermal Transmittance-Calculation Method (ISO 6946).Google Scholar
- Lagüela S, Díaz-Vilariño L, Martínez J, Armesto J: Automatic thermographic and RGB texture of as-built BIM for energy rehabilitation purposes. Automation in Construction 2013, 31: 230–240. 10.1016/j.autcon.2012.12.013View ArticleGoogle Scholar
- Madding R: Finding R-Values of Stud Frame Constructed Houses with IR Thermography. Proc, InfraMation; 2008.Google Scholar
- Mikron Instrument Company. (2012). Table of Emissivity of Various Surface.Google Scholar
- Roth K, Westphalen D, Feng M, Llana P, Quartararo L: The Energy Impact of Commercial Building Controls and Performance Diagnostics: Market Characterization, Energy Impact of Building Faults and Energy Savings Potential. TIAX LCC Technical Report Cambridge, MA; 2005.Google Scholar
- Singham JR: Tables of emissivity of surfaces. International Journal of Heat and Mass Transfer 1962, 5(1–2):67–76. 10.1016/0017-9310(62)90102-3View ArticleGoogle Scholar
- The Engineering Toolbox. (2013). Heat Loss through Building Elements due to Transmission. http://www.engineeringtoolbox.com/heat-loss-transmissiond_748.html.Google Scholar
- U.S. DOE. (2010). 2010 U.S. DOE Buildings Energy Databook. Washington D.C.: U.S. Department of Energy.Google Scholar
- Wang C, Cho Y, Gai M: As-Is 3D Thermal modeling for existing building envelopes using a Hybrid LIDAR system. Journal of Computing in Civil Engineering 2013, 27(6):645–656. 10.1061/(ASCE)CP.1943-5487.0000273View ArticleGoogle Scholar
- Wanielista M, Kersten R, Ron E: Hydrology Water Quantity and Quality Control. John Wiley & Sons, Hoboken, NJ; 1997.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.