The most obvious way to cut energy costs, lowering consumption, isn’t as easy as shutting some lights off for commercial buildings. Energy costs are up across the board with the exception of renewables, and power peaks keep the meters ticking higher. Service level agreements dictate comfort levels even if floors go largely unoccupied. HVAC systems, typically the biggest energy eater, may be disconnected from the brain of the building, the building management system (BMS).
But machine learning and AI are making use of all the sensors that exist in modern buildings today, automating how and when energy is used based on actual conditions inside buildings today. Conditions that have only gotten more complicated.
Controlling the uncontrollable
The most powerful forces that impact a building’s energy usage are outside of the four walls of the building: humidity, weather, outside temperature, sun exposure, and thermal conditions.
Most building managers and engineers barely have time to leave the building for lunch, let alone monitor what’s happening outside and think about how that’s impacting different floors and sides of the building. Just like sensors have been the gateway for AI inside buildings, they’re doing a similar job in managing the external factors that dictate comfort and lighting on the inside.
One former (relative) constant, that’s now squarely in the variable bucket, is occupancy. The consistent 9-to-5 schedule for everyone with an access badge doesn’t exist anymore. Occupancy is the newest, and perhaps most frustrating, guessing game in property management, and it has big implications for energy usage, particularly heating and cooling. Office buildings are now acting more like supermarket and mall parking lots, designed to accommodate enough people at the busiest shopping times even though they are rarely at full capacity.
From an energy perspective, a real-time connection between occupancy sensors at the floor or suite level, HVAC systems, and the grid is a must-have for building operators to make better, more efficient energy decisions. In terms of the energy loop, the grid itself has a solid understanding of what’s happening from it to the substation level but does not have that same intelligence about what’s happening from the substation to the building. AI is bridging this gap, helping to bring new control to two formerly uncontrollable factors, the price and cleanliness of energy.
The new economics of power peaks
Let’s go back to that parking lot analogy. The grid is designed in a similar way, accommodating the highest demand for energy on the hottest and coldest days, or even parts of the day. Sometimes, there just aren’t any more spots available. That’s what happens when the grid is overtaxed and can’t handle all the calls for energy it’s getting. At these peaks, power is often pulled from the dirtiest sources and is at its most expensive price.
Powering buildings based on actual occupancy can lower these peaks. If occupancy is only at 50 percent on a hot summer Friday and the building knows which floors are nearly or completely empty, the building can automatically control heating and cooling needs accordingly.
For buildings with onsite renewables and energy storage, knowing when the grid is entering these peak power periods allows the buildings to turn to their own energy when it’s most impactful to lower both costs and emissions.
Emerging carbon marketplaces
Emissions that can’t be avoided and surplus renewable energy credits may also find new life in a carbon trading marketplace. Japan is one of a handful of countries testing a public Carbon Credit Market (CCM), which would allow companies operating in the country to buy and sell carbon credits from non-fossil fuel energy production as a part of their larger decarbonization strategies in addition to credits generated from publicly funded projects
This public CCM is in addition to the private sector carbon markets that already exist in Japan.
Carbon pricing, which is already common in Europe and Asia, will be set and adjusted through an algorithm based on good old economics, supply and demand.
John Gilbert, Executive Chairman of Prescriptive Data and creator of AI-powered smart building platform Nantum OS, is quick to point out the potential of these markets. “Carbon isn’t priced in the U.S., but countries like Japan are exploring real trading mechanisms so carbon can be traded within and between portfolios. That would allow real estate companies to trade high intensity with lower intensity carbon to create better average emissions across the board.”
Japan’s public CCM is in beta mode today, one of the many levers that the country is using to reduce carbon emissions by 46 percent in 2030.
Low tech, high impact
While carbon marketplaces may be futuristic, less “exciting” AI doesn’t mean less effective. A building’s equipment, particularly its HVAC and water systems, can make or break energy budgets and decarbonization strategies. According to the U.S. Department of Energy, the use of the most efficient wall, window, and HVAC equipment available today could reduce commercial cooling needs by 78 percent.
Preventive maintenance, such as the regular replacement of key parts like air filters, and equipment life cycles, presents timely ways to improve the building’s resilience and energy efficiencies without ballooning capital budgets.
Laurie Kerr, Climate Advisor at the U.S. Green Buildings Council, explains, “Looking at the end of equipment life is the most cost-effective time to improve the building’s system because you’re just paying for the incremental increase of the new equipment. If the design team can understand the financial cycles of the building to determine when money is going to be available and match that to existing equipment life cycles, plans for decarbonization become more sustainable.”
Pairing this financial modeling with system performance data, equipment history, and energy utilization paves the way for quick wins. And Kerr is quick to point out that any win is a win when it comes to lowering energy consumption and emissions. She shared an important reminder, “Let’s not forget that operational savings from tuning and running the building properly are important in the context of these larger goals. So many aspects of retrofitting are extremely expensive but don’t overlook the inexpensive ones. Carbon is carbon, and it doesn’t care if you paid a lot or a little to reduce it.”