Every commercial building contains significant thermal storage capacity — it's built into the concrete slabs, masonry walls, and water-based mechanical systems that make up the structure. This thermal mass absorbs heat when the building warms and releases it when the building cools, acting as a natural buffer that moderates indoor temperature swings. Most building management systems treat it as background physics — not as an asset that can be deliberately charged and discharged to shift HVAC demand.
That's a missed opportunity with real dollar value. The same thermal mass that passively moderates temperature can be actively managed: pre-cooled during off-peak hours to absorb morning heat gains without requiring live HVAC staging, then allowed to release that stored cooling during the peak tariff window while mechanical cooling runs at reduced capacity. No capital investment. No battery storage. The storage is already there.
What Thermal Mass Actually Means in a Building Context
Thermal mass is the product of a material's specific heat capacity and its mass — the amount of energy required to raise the temperature of the material by one degree. Concrete has a specific heat capacity of approximately 0.84 kJ/kg·K. A 100,000 sq ft commercial building with 6-inch concrete floor slabs on all levels contains roughly 800,000–1,200,000 kg of concrete, giving it a thermal storage capacity in the range of 670–1,000 MJ/°K — or equivalently, the capacity to store several hundred kWh of thermal energy per degree of temperature change across the slab mass.
In practical terms: if you pre-cool a concrete-heavy commercial building structure by 2°F below normal setpoint overnight, you've pre-loaded roughly 180–300 kWh of cooling capacity into the building's thermal flywheel. That stored cooling will absorb heat gains in the morning — occupant body heat, solar gains, lighting — without requiring the chiller to run at full capacity. The chiller effectively deferred work it would otherwise do during the on-peak window to the overnight off-peak window, at off-peak energy rates and without generating a demand event during peak tariff hours.
Time Constants: How Fast Buildings Heat and Cool
The useful thermal mass isn't just total mass — it's accessible thermal mass, which depends on time constant. A time constant in building thermal modeling is the characteristic time for a building's interior temperature to change in response to a step change in boundary conditions, roughly analogous to the RC time constant in an electrical circuit.
Lightweight steel-frame construction with dropped ceilings and minimal exposed concrete has a short time constant — 1–3 hours. These buildings heat and cool quickly, which means their thermal mass is depleted fast and pre-cooling delivered 6 hours before occupancy provides limited benefit. Heavy masonry construction — poured concrete structure, exposed concrete ceilings, thick exterior walls — has a long time constant of 8–16 hours or more. Pre-cooling delivered the night before is still partially active at 10 AM the next morning.
This distinction matters enormously for forecasting-based demand management. A heavy-mass building with a 12-hour time constant can be meaningfully pre-cooled 8–10 hours before the on-peak window. A light-frame building needs pre-conditioning within 2–3 hours of occupancy to retain benefit. The forecast-based staging schedule has to reflect which type of building it's operating — using the wrong time constant produces either over-early pre-cooling (energy waste) or under-effective pre-cooling (demand charge not avoided).
Chilled Water Systems: Intentional Thermal Storage
Buildings with central chilled water plants have an additional thermal mass asset that's often overlooked: the chilled water loop itself, and more specifically, any thermal storage tanks that may be part of the system.
A standard chilled water loop in a 200,000 sq ft commercial building might contain 15,000–25,000 gallons of chilled water at 42–48°F supply temperature. Lowering the chilled water supply temperature by 4°F during off-peak hours — effectively over-chilling the loop — stores approximately 50–80 kWh of additional thermal capacity in the water volume. When morning peak tariff hours arrive, the chilled water plant can operate at reduced output while the pre-chilled water loop absorbs building heat gains at a lower rate.
Buildings with dedicated chilled water thermal storage tanks (ice storage or stratified chilled water tanks) take this further. A 500-ton-hour ice storage system can defer 500 ton-hours of cooling from on-peak to off-peak hours — displacing the most expensive peak-rate energy entirely. Managing these systems optimally requires exactly the kind of day-ahead thermal demand forecast that tells the operator how much storage to charge overnight based on tomorrow's expected cooling load.
We're not saying every building should install thermal storage tanks — the capital cost requires careful ROI analysis. We're saying every building already has passive thermal mass that's being ignored, and exploiting it costs nothing except the forecasting capability to time it correctly.
The Pre-Cooling Window and Its Limits
Active thermal mass pre-cooling — deliberately running HVAC below normal occupied setpoint to charge the building's thermal flywheel — sounds straightforward but has practical constraints that are important to understand before relying on it as a demand management strategy.
First, occupant comfort limits the pre-cooling depth. Chilling a space to 68°F overnight to pre-load cooling capacity creates a cold building at 7 AM when early occupants arrive. Most commercial settings tolerate 70–71°F minimum setpoint during unoccupied pre-cooling — which limits the thermal storage depth. A 2°F pre-cooling differential is achievable and useful; a 6°F differential that creates a refrigerator-cold workspace at 7:30 AM is not operationally viable.
Second, pre-cooling in high-humidity climates requires careful dew point management. In Minneapolis summers, dropping indoor temperatures below the dew point of infiltrating outdoor air can create condensation on cool surfaces — particularly in buildings with significant envelope leakage. Pre-cooling setpoints need to account for forecast outdoor humidity, not just outdoor dry-bulb temperature. This is one reason why a humidity-aware forecast model produces better pre-cooling recommendations than a temperature-only model.
Third, the demand savings from thermal mass pre-cooling are not unlimited. There's a maximum load-shifting potential per building type and per day. On an extremely hot summer day, the building's heat gains will overwhelm whatever thermal buffer was loaded overnight within 2–3 hours of occupancy, after which the chiller must run at full capacity regardless. Pre-cooling is most effective at reducing the morning peak on moderate days — shoulder-season days, cool-front mornings — where it can carry the building through the entire on-peak window. On severe weather days, it buys time but doesn't eliminate demand events.
How a Thermal Forecast Quantifies the Buffer
To use thermal mass as a deliberate demand management tool, you need to know three things before you start pre-cooling: how much thermal buffer the building currently holds, how fast the forecast conditions will deplete it, and how much additional buffer you can add overnight. A thermal demand forecast answers all three.
The model tracks the building's thermal state — which we estimate from BMS zone temperature readings, outside air conditions, and occupancy history — and projects forward based on forecast weather and occupancy. The output isn't just a demand curve; it's a pre-cooling recommendation that specifies the start time, setpoint targets, and expected demand-saving window, calibrated to the specific building's time constant and the specific day's conditions.
Take a representative scenario: a 7-story office building in Minneapolis with exposed concrete ceilings and a 10-hour thermal time constant, on a mid-October morning with a forecast overnight low of 28°F and a daytime high of 52°F. The model estimates that passive overnight cooling will lower concrete surface temperatures by approximately 3°F below normal setpoint, pre-loading useful thermal buffer without any active HVAC involvement. Combined with an optional 90-minute active pre-cooling window starting at 5:15 AM, the building can enter the 9 AM on-peak window with enough thermal buffer to run chillers at 35–40% staging for the first two hours of occupancy — instead of the 95% staging that a cold-start would require. That reduction, held across the first two on-peak 15-minute intervals, prevents the monthly peak demand event.
This is the value proposition of thermal mass management: it turns building physics into a demand charge reduction mechanism, with no capital cost and no occupant impact. The limiting factor is having an accurate enough forecast to know when the conditions favor it and exactly how to time the pre-cooling window.