Smart buildings are only as intelligent as the systems that operate them. Much of the data collected in IoT applications goes nowhere, but when utilized in modern simulation software it can be a game changer
by Dave McKenna, Editor of CREB on July 30, 2020
Real estate owners are spending money fast on smart building technology and IoT devices. Global spending on data-enabled systems exceeds $15 billion annually and is expected to double in the next five years. Yet forty-two percent of smart-building data is never used for any analytical purpose or decision making according to a recent study. It is like asset owners are filling up their car with 15 gallons of gas and letting 6 gallons drain out into the street.
Bractlet co-founder and CEO Alec Manfre has his eye on all those lost gallons of data to help make buildings more energy and capital efficient. Over the past decade, Bractlet has created a sophisticated Building Performance Simulation (BPS) that creates a “digital twin” of a building and can execute an almost infinite series of virtual tests to optimize the building’s operation. The mathematical representation of the building is so accurate that it can predict the result of a planned new control system or a capital investment with 98% accuracy.
Models like this have been around since the days of the Moon landings. As long as there have been digital computers, academics have used them to run the complex thermodynamic calculations to study building dynamics. These methods have been mostly scientific exercises with some limited application in the design of new buildings. According to Manfre, the power of this technology has just not been effectively applied to on-going building operations. “The fact is that BPS tools have never taken full advantage of the resources available,” said Manfre. “Building simulation modeling has not been used to its fullest potential. It’s mostly been used in new construction to rough-size HVAC loads and meet LEED or code requirements. It has never been used as a strategic asset.”
Founded in 2011, the Austin, TX based smart-building and analytics firm utilizes massive volumes of building data captured from a combination of existing building management systems, proprietary sensors, and open-source data to create the digital twin of a building. This virtual representation of the structure reproduces the building mathematically based on its actual finished configuration. This virtual building is then subjected to the rigorous Bractlet modeling algorithms which are capable of highly accurate predictions of how the building will react when investments are made or other changes occur.
Declining IoT Data Cost
The cost of collecting building data has been driven down in recent years by the emergence of inexpensive IoT devices. There has never been more data available for use by smart buildings, and the volume increases every year. “The technology to capture, store and process the volume of data required for the models just didn’t exist ten years ago. The typical simulation requires trillions of calculations. We’ve developed a horizontally scalable infrastructure in the cloud that can handle many simulations at scale,” said Manfre.
It is common for significant capital investments like replacing a chiller to be made with very rough like-as estimates jotted on the back of an envelope. The problem is that buildings are complex, mutually dependent systems, so changing one system will have many second-order effects. “With our simulations we are able to test many different chillers and control strategies and we can know exactly how the building will perform with each combination,” said Manfre.
Unlike the design tools used to simulate building performance based on libraries of benchmark data aggregated from similar buildings, the Bractlet platform uses actual data from the building coupled with proprietary modeling algorithms to create a model that has a 98% accuracy record. “It is very powerful because we can precisely simulate future conditions before you actually spend any money to do it,” said Manfre.
Simulating Capital Investments
In a recent case study, Bractlet collected a half-billion data points for a portfolio and executed a simulation that represented the equivalent of 17,000 building-years to evaluate hundreds of potential capital investments. The projects recommended by the simulation resulted in $2.7 million in annual savings and an estimated increase in asset value of $45 million. “The typical implementation collects twenty million data points per month in a building. We have automated the analysis that a building engineer has traditionally done by hand and is not really equipped for. We let the machines handle all of that for them and free the humans to the more critical tasks.”
The ability to accurately predict the impact of capital improvements is a powerful advantage for the asset owners. “It creates trust with the financial decision makers to have confidence that they can make the best decisions for their buildings,” said Manfre.
The predictive power of the simulations is of particular interest during periods of significant disruption, as in the current pandemic. “It is very easy for us to model the effect of ‘fewer-or-no people’ in the building and configure the systems appropriately,” said Manfre. “We recently detected a suddenly-vacant floor in a large building of one of our customers and were able to proactively alert them to respond.”
Bractlet operates in all the major U.S. metropolitan areas with over 30 million square feet under management. The science of building performance modeling makes sense for assets of just about any site. “The typical installation is in buildings of 50,000 square feet or more. We can either help customer leverage data they are already producing in their building automation system and/or deploy easily installable power meters. Our product tiers give them flexibility to get started at an analysis tier that makes the most sense for them,” said Manfre.
The operational efficiencies of using detailed simulations to optimize operations can be material. “We’ve seen reductions of 10-12% in energy consumption from optimizing existing controls systems,” said Manfre. The overall impact can be double that number when capital efficiencies are accounted for. “We know the science of buildings really well. We can help owners figure out how to be run their buildings. We help the owners discern good ideas from snake oil,” said Manfre.
As the reach of IoT and smart building technology expands, the translation of the data they generate into useful information becomes more relevant. Billions are being spent on the collection of these data and it stands to reason that the analytical tools will evolve with it.