Rapid Building Energy Modeler: The Key to Fast Energy Efficiency

Rapid Building Energy Modeler: The Key to Fast Energy Efficiency

April 21, 2014

A multi-billion dollar market exists for reducing the energy use of existing buildings, if scientists can only figure out a way to substantially reduce the cost and time required to assess building energy performance, recommend energy performance measures, and identify problems in building operations.

This is the goal of RAPMOD, the Rapid Building Energy Modeler, a collaborative project involving the University of California, Berkeley, the Lawrence Berkeley National Laboratory (Berkeley Lab) and engineers Baumann Consulting. RAPMOD, which was funded by an innovation grant from the Advanced Research Projects Agency-Energy (ARPA-E), is designed to tackle this problem head on.

The technology, worn as a backpack, is designed to scan a building’s interior, using several types of sensors, as its wearer walks through the building. RAPMOD generates a visual map of the building that can be input into energy simulation models and used to develop an understanding of the building’s energy performance, leading to a list of recommendations for improving its efficiency.

RAPMOD is based on technology developed by Avideh Zakhor, Qualcomm Professor of Electrical Engineering at UC Berkeley’s Department of Computer Science and Electrical Engineering, and her students. Zakhor has been developing technology to produce indoor three-dimensional models since 2007 under the sponsorship of Army Research Office (ARO) and Air Force Office of Scientific Research (AFOSR). She developed the first fully automated fast outdoor mapping system in 2005, which was licensed by Google in 2007 to help produce its 3D Google Earth product.

Since then, Zakhor’s research group has been advancing the technology for use in indoor 3D mapping since 2007. In 2012, they teamed with a group of researchers led by Philip Haves, Leader of the Simulation Research Group in the Environmental Energy Technologies Division of Berkeley Lab, and with engineers Bauman Consulting, to adapt the technology to generate energy models of buildings quickly and inexpensively.

“It’s possible to reduce the energy consumption of existing buildings significantly,” says Haves, “through retrofitting – replacing old equipment with more energy-efficient technology – and through ‘retro-commissioning,’ the process of improving the routine operation of buildings by making equipment function properly.”

Prior research suggests there is a potential to reduce whole building consumption in the U.S. by 16 percent through retro-commissioning, which uses ‘low-cost or no-cost’ measures. This maps to an energy-savings potential of $30 billion by the year 2030 and annual greenhouse gas emissions reductions of about 340 million tons of CO2 per year. Retrofitting has significant cost but can result in energy savings of 20 to 50 percent. Energy modeling is required for the detailed analysis needed to achieve deep savings cost-effectively and is also helpful in maximizing and verifying the savings from retro-commissioning.

“The problem,” says Haves, “is that retrofits projects often ‘cream-skim’, saving about 10 percent while ignoring potential for deeper savings. We aim to reduce the cost, and improve the accuracy, of energy modeling to reduce the cost of identifying retrofit measures that will produce deep savings.”

Creating building energy models is expensive and time consuming and requires a lot of skill. Many existing buildings have incomplete, outdated, or no design documentation, requiring specialists to go into the building and laboriously make measurements that they can import into the software required to create the model. The primary goal of the RAPMOD project is to reduce the cost of preparing an energy model for use in retrofit analysis and in model-based retro-commissioning. This same model can also be used in performance monitoring during routine operation to detect equipment faults and other operational problems.

There are also non-energy applications of the technology. It could be used to create maps of building interiors for emergency first responders and Architecture Engineering Construction (AEC) companies could use the system to generate maps of the interior building structure and services (such as HVAC ducts, and gas, power, and water lines) during construction. Such maps would help building managers keep their buildings in good repair and running well during the building’s life. Game designers and real estate industry could also make use of interior mapping in their work.

Realizing that Zakhor’s 3D modeling technology offered a faster way of gathering the data needed for these models, Haves invited Zakhor to explore a collaboration between her lab and Berkeley Lab’s Simulation Research Group. Bauman Consulting was brought in to advise on industry practices and costs and conduct testing and demonstration. A prototype version of the RAPMOD system was shown at ARPA-E’s Technology Innovation Conference in late February 2014, and, a day later, demonstrated for member of Congress at a showing on Capitol Hill.

How It Works

RAPMOD is fitted with several different sensors, including a LiDAR, which measures the distances to building surfaces using a laser, a visible light camera, and an infrared sensor. The camera and LiDAR generate a photorealistic three-dimensional model of the building’s interior as the user walks through hallways, into rooms, and up and down staircases.

The infrared sensor measures the thermal properties of windows and detects thermal defects e.g. in wall insulation or moisture leaks. It also measures the heat coming from lighting systems, other equipment, and building occupants, providing the model with information required to calculate the energy required to heat and cool the buildings.

A major advantage of RAPMOD is that it doesn’t need to be operated by high-cost-energy experts. Technicians will be able do the building walkthrough and measured data will upload automatically for processing and importing into the energy modeling software. All this drives down the cost of producing the model substantially.

One of the major tasks in the research has been to integrate the infrared sensor into the equipment, and to determine how much it can tell users about the thermal characteristics of the building materials—the insulation in the walls, and the windows’ the U-values, a measure of how well they perform at retaining heat to the interior.

A first version of the RAPMOD system that maps building geometry is expected to be ready for field testing in the summer of 2014. A version that measures window properties and characterizes internal heat gains is expected to be ready for field testing and demonstration by the end of 2014.

The research team is now seeking partners among architect, engineering and construction firms, consulting engineering firms, facility managers, energy service companies, and others to help with testing and demonstrating the technology in existing facilities.

Allan Chen