This blog was written by Tanya Barham, CEO at Community Energy Labs
Why Building Decarbonization Matters
Buildings account for 30-40 percent of the world’s energy consumption and carbon emissions. Schools are the largest energy consumers in the public sector, with HVAC accounting for 46 percent of their total energy costs. Additionally, energy is generally the second most pricey budget item for schools after staffing costs. The big question is: how can we heat and cool our schools (and other buildings) while using less energy, emitting less carbon, and spending less money?
Enter: Model Predictive Control
Model predictive control (MPC) has emerged as a powerful tool to optimize building performance. MPC is a model-driven control technique that uses mathematical models to predict the behavior of a building over time. This allows the system to make intelligent decisions about how to adjust the building’s heating, cooling, and ventilation systems in real time. MPC can then dictate the decisions of a smart control AI in order to adjust and optimize a building’s systems to maximize efficiency and comfort.
Compared to other forms of building control, MPC offers several advantages. First, it is more accurate and efficient than traditional control methods when it comes to delivering comfort and energy savings. Second, it can reduce the workload of building operators by automating routine tasks. It can also reduce maintenance and operating costs by optimizing building system performance. Therefore, MPC can reduce energy consumption and carbon emissions, contributing to a more sustainable future.
Sounds great! What’s the catch?
MPC solutions come in two models. Tailored “white-box” models require substantial expertise and funding to install. Cheaper “black-box” models require massive amounts of data and time just to teach the algorithm physics. This means customers with lean budgets and small staffs are faced with two options: go broke purchasing expensive, tailored solutions, or continue spending staff hours they can’t afford to lose wrestling with spreadsheets and constantly fiddling with thermostats if and when they get around to the task.
There is a distinct and short list of climate tech companies looking to solve this problem. Through a recent pilot funded by the U.S. Department of Energy, Community Energy Labs (CEL) proved its ability to implement MPC with drastically reduced cost, time, and expertise for implementation. By leveraging alternative collection methods, CEL’s platform successfully reduced data collection time by 17-38 percent (5-32 hours) compared to standard practice, depending on the building size. CEL’s model also shows the potential to reduce customer onboarding costs by 50 percent without impacting accuracy.
Preliminary results promise to eliminate manual calibration and model tuning, further reducing the amount of expert time necessary to make CEL’s technology work for community buildings. A pilot at Sonora Elementary School demonstrated the efficacy of the model; it reduced 24 percent of total peak power and 30 percent of HVAC peak power and it shifted 16 percent of the cooling load from an on-peak price period to a low price period. This can save the school thousands in utility bills while still keeping students and faculty comfortable. The peer-reviewed article published about this pilot can be found here.
Critical technologies like these are what it will take to make a theoretical solution like MPC actionable and useful for most organizations with tight budgets and limited staff. Project Drawdown estimates that the adoption of building automation systems like these could reduce up to 11 gigatons of CO2 equivalent by 2050.