Rapid 3D Energy Modeling for Retrofit Analysis of Existing Buildings Using Thermal and Digital Imagery

A 1.139Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE

How can you detect and spatially locate thermal defects and air leakage, and have a rapid assessment of the current energy performance of your buildings in 3D? A new 3D spatio-thermal modeling technique takes a collection of inexpensive and often existing thermal and digital 2D imagery, analyses them for similarities, and then displays them in a reconstructed 3D space. The resulting 3D models enable inspectors and energy auditors to move through virtual models of buildings, look at thermal and digital imagery from any angle, perform geometrical and temperature measurements in 3D, and see where these images were taken in relation to other images or to the underlying geometry of the building scene.

The Need for Rapid Energy Performance Modeling of Existing Buildings
With increasing efforts to improve building energy efficiency, the necessity of modeling present conditions of building energy performances is on the rise to establish proactive building retrofit strategies for potential energy savings. Retrofitting existing buildings is important since they continuously undergo degradation over their service life. Timely measurement and assessment of building energy performance helps owners and facility managers to identify potential areas for better retrofit, and meet environmental and economic goals. Despite the increasing attention to energy performance assessment of existing buildings, the current process of energy data collection and modeling is challenging:

1. Recent energy data collection practices provide valuable insight on monitoring appliance-level energy-use or collecting environmental data using sensor networks, nonetheless these methods do not adequately measure building system degradation due to materials aging. Moreover, installation of the sensor networks for energy data collection can be expensive and time-consuming to set up for many existing buildings;

2. Current modeling practices using existing energy performance simulation tools (e.g., EnergyPlus, Ecotect, and eQuest) are time-consuming and labor-intensive due to the need for manual 3D building modeling, model parameters entry, and model calibration. As a result, the process of constructing energy models often requires weeks to months, and therefore often tend to be restricted to only high-profile and high-budget budget projects.

3D Spatio-Thermal Modeling
Thermography is defined as detecting and measuring heat variations emitted by an object and transforming them into visible images. In terms of building energy performance modeling, thermography is a robust tool in recording, analyzing, and reporting actual energy performance of existing buildings. Thermal images from buildings are directly influenced by the building energy performance such as energy transfer through building elements (e.g., thermal bridges) or space heating and cooling energy related to HVAC systems (Fig.1). Images labeled with `b’ in Fig. 1 show the thermal characteristics of the same areas captured in `c’.

At present, thermographic inspections are primarily based on taking 2D thermal images from interior and exterior building elements. However, since these 2D images can only show what is in their field of view and are not geo-tagged in relation to other images or to an underlying 3D model, they are not sufficient to provide a comprehensive representation of thermal distribution and energy performance in a given space. There is a need for a more detailed interpretation and visualization of thermal performance in a given building space under investigation, put thermal imagery into the building context, and geo-register visual thermal performance data in 3D.

Using digital and thermal imagery of existing residential and commercial buildings, research at Virginia Tech has resulted in a new 3D energy performance modeling approach that visualize how temperature is spatially distributed in a given 3D space. To develop the system for 3D spatio-thermal modeling, Prof. Mani Golparvar-Fard, Assistant Professor of Civil Engineering at Virginia Tech worked with his PhD student, Youngjib Ham.

In the proposed method, the digital images are capturing the as-is condition of existing buildings, and the thermal imagery are capturing the actual energy performance. Since the proposed method automatically generates 3D point cloud models of both geometrical condition and thermal performances of existing buildings with a collection of unordered digital and thermal images, it has the benefit of streamlining the entire energy modeling process. The resulting 3D point cloud models are then jointly registered within a single virtual 3D environment. The outcomes are 3D spatio-thermal models which simultaneously visualize 3D thermal distribution and geometrical conditions in a given space at the time digital and thermal images are captured.

Fig. 2, 3, and 4 show initial experimental results using 429 thermal and digital images taken with FLIR E60 camera. The experiments were performed in an office room of an existing instructional facility on campus of Virginia Tech. Fig. 2 shows the dense 3D reconstructed point cloud models of building (Fig. 2a) and thermal performance (Fig. 2b). The locations and orientations of each camera registered in a virtual 3D environment are automatically calculated without help from GPS, Ultra Wideband, or RFID technologies. Rather this information is automatically calculated from the visual content of collected imagery. Once a camera is visited in the virtual environment (Fig. 3a and d), the camera frustum in 3D scene (Fig. 3b and e) is automatically texture-mapped with a full resolution of the image that was used to capture it (Fig. 3c and f).

Fig. 4a shows a 3D building point cloud model which is superimposed with a 3D thermal point cloud model within the same site coordinate system. With the availability of full-resolution imagery in the reconstructed 3D scene (Fig. 4b), the user can interactively acquire critical information related to an actual representation of thermal and geometrical conditions of both interior and exterior building environments.

The resulting 3D spatio-thermal models could be used as a robust tool for building retrofit analysis. It is intended to automatically provide inexpensive and frequent energy performance data, quantify the space conditioning of existing buildings, enable improved predictions of building energy performance on the basis of identifying deviations between actual and expected thermal performance data, create much-needed feedback loop with design predictions, and finally identify potential building or systemlevel candidates for building retrofit purposes. Benchmarking the expected energy performance and monitoring the deviations can support proactive building diagnostics through quick analysis of energy problems, and further supports decision-making as to how energy loss in existing buildings can be minimized. These areas are currently being explored as part of our ongoing research.

Youngjib Ham is currently a PhD student and Vecellio pre-doctoral fellow at RAAMAC research lab in the Department of Civil and Environmental Engineering at Virginia Tech. His research is primarily focused on rapid 3D energy modeling and retrofit analysis of existing residential and commercial buildings using thermal and digital imagery.

Mani Golparvar-Fard is currently Assistant Professor of Civil Engineering at Virginia Tech. He is also director of the Real-time and Automated Monitoring and Control (RAAMAC) research lab (www.raamac.cee. vt.edu) wherein research focuses on creating and developing computer vision and augmented reality systems that can automatically track building and construction performance metrics using site digital photos, video streams as well as building information models. As an entrepreneur, he is launching a new startup company (Vision Construction Monitoring LCC) to offer his research results to the building and construction industry.

A 1.139Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE

About the Author

Guest Contributors

Guest Contributors ... Articles from Experts in the LiDAR community
Contact Contributor Article List Below