REMI
Ecopilot

Building energy management: The AI advantage

Tuesday, August 2, 2022

There’s a science to building performance. And when it comes to saving energy and providing occupant comfort, the solution lies in recognizing how thermodynamics and weather data can be used by artificial intelligence (AI) for optimal climate control.

“A building’s energy usage is directly influenced by its thermal mass and weather conditions,” explains Tobias Janes, technical director, with EcoPilot Canada. “All of those factors can be measured and assessed in real-time; but more importantly, that data can be used by building automation systems (BAS) to respond to exterior conditions more efficiently and cost-effectively.”

Understanding thermal mass and its role in the energy-saving formula is step one. In simplest terms, thermal mass describes the building’s ability to absorb, store, and release heat generated from internal (e.g., people, equipment, lighting, etc.) and external (e.g., outdoor temperature, humidity, solar radiation, etc.) sources. A building’s thermal mass can fluctuate based on the above-mentioned factors, creating more and less demand for HVAC equipment.

Weather conditions also play a role in determining how hard HVAC equipment has to work to maintain ideal temperatures. Since 2019, Crombie REIT’s Scotia Square complex has reduced HVAC energy savings by 19%, HVAC costs by 17%, and recued CO2e by 1,215 metric tonnes of CO2. Every 2-minutes, Ecopilot® calculates the building balance point temperature and provides a new offset to the BAS based on ever-changing building variables. The result is energy savings while maintaining occupant comfort.

As such, Janes explains, part of Ecopilot®’s solution involves using weather-related data points (e.g., humidity, wind speeds, temperature, solar radiation, etc.) in combination with thermal mass calculations to inform more efficient HVAC
operations: “With Ecopilot®, for example, we can use AI to assess a building’s thermal mass in real-time and use that data alongside weather-related data and forecasts to recommission the HVAC system every two minutes.”

“The result is that these systems start running far more efficiently, use less energy, and reduce their carbon emissions – all while keeping people a lot more comfortable,” he continues.

Enter: Model Predictive Control 

If “automatically re-commissioning HVAC systems based on real-time thermal mass and weather data” is too much of a mouthful, try “model predictive control,” or MPC. As Ecopilot® explains: “An MPC system anticipates the energy needs of a building based on data it gathers, which it then uses to optimize the building’s thermal behavior. As a result, Ecopilot® reacts with changes to heating, cooling, and ventilation, by foreseeing changes in the weather before they occur.”

Benefits of an MPC system include:

  • Automated performance enhancements: Weather-related data can be used by a BAS system to manage HVAC equipment more efficiently. For example, Ecopilot®’s AI-driven solution works with a BAS to constantly assess
    current and future weather conditions and adjust HVAC operations accordingly every two minutes. This automatic process is ongoing and requires zero operator interaction.
  • Easier and more proactive HVAC control: Interpreting weather data enables greater flexibility when programming the building’s HVAC on days the weather won’t interfere. This provides the benefit of recognizing high and low operating performance periods so the building’s system can adjust accordingly.
  • Enhanced responses: By incorporating weather data into a property’s existing HVAC control, the systems can be further programmed to automate energy-saving actions. This makes a crucial difference in improving a commercial building’s energy consumption.“Dealing with weather conditions will always be a substantial challenge for property management,” says Janes. “However, our AI solution helps buildings adapt automatically so that they’re responding to those conditions in the most energy-efficient manner.”
  • Return on investments: Environmental data can be used to show building owners the benefits of installing energy-efficient items (e.g., photovoltaics or wind turbines) and predict their return on investment.
  • Optimizing equipment: Weather data can be used to make investment decisions, such as determining when to use a smaller chiller at full capacity versus a larger chiller at half capacity in order to save on energy and maintenance.

AI in action

Ecopilot®’s AI solution, designed to work alongside any BAS, proactively prepares a building for its energy requirements, using only the energy required to maintain consistent tenant comfort. The AI prevents the control system from overreacting.

AI-driven HVAC controls may seem like a solution for the future, but consider the success that Ecopilot® has already demonstrated in buildings around the globe.

Ecopilot

In the graph above, for example, Ecopilot® was called into a Toronto building to help it gain more control over its internal heating and cooling functions amidst ever-fluctuating weather.

As the case study notes, “In addition to the thermodynamics properties of the building, which remain static, Ecopilot® also uses the collection of real-time data to assess other fluid variables, like weather, people, and heat generated from solar radiation, or machines, that is also stored in the building and released over time.”

In this Toronto scenario, Ecopilot®’s AI produced HVAC efficiency savings of 45%, without sacrificing building occupant comfort.

Comfort in data
The science is complex, but the results are clear: measuring and assessing a building’s thermal mass and weather-related influences in real-time with AI results in smarter, more adaptable, and more efficient climate control strategies.

For more, check out a series of Ecopilot® case studies.

Ecopilot

Leave a Reply

Your email address will not be published. Required fields are marked *

In our efforts to deter spam comments, please type in the missing part of this simple calculation: *Time limit exceeded. Please complete the captcha once again.