Most facilities managers know their monthly utility bill is higher than it should be. Fewer can explain, line by line, what's driving it. The demand charge — typically the second-largest line item after energy consumption — is routinely misunderstood, underestimated, and not actioned because the mechanics aren't obvious from the bill format utilities provide.
This is worth fixing. Once you understand how demand charges are calculated, you can see immediately where HVAC operations either compound or constrain them — and why the timing of HVAC staging decisions matters as much as total energy consumption.
The Two Fundamentally Different Things on Your Bill
Commercial electricity bills contain two distinct charge types that measure different things. Energy charges (measured in kWh) represent total electricity consumed over the billing period — the integral of power draw over time. Demand charges (measured in kW) represent peak power draw — the maximum rate of electricity consumption recorded during any single measurement interval.
The distinction is critical. A building that runs 50 kW of load continuously for 720 hours consumes 36,000 kWh and its peak demand is 50 kW. A building that runs 200 kW for 2 hours and 10 kW for the remaining 718 hours also consumes approximately 37,380 kWh — roughly the same energy — but its peak demand is 200 kW. The second building's demand charge is four times higher despite nearly identical energy consumption.
This is not a theoretical edge case. It's an accurate description of what HVAC morning pre-conditioning looks like on a utility meter versus what the rest of the building's load profile looks like. The ratio of peak-to-average demand — the load factor — is frequently the most important number in your building's energy economics, and HVAC startup behavior drives it directly.
How the Peak 15-Minute Interval Is Recorded
Most commercial metering infrastructure records demand in 15-minute intervals. The meter averages the total kW demand across each 15-minute window and logs the result. The demand charge for the billing month is calculated from the single highest of these 15-minute averages — the peak demand interval.
A few utilities use 30-minute intervals; some industrial accounts use 5-minute intervals for specific tariff structures. But 15 minutes is the standard commercial interval, and it's the window size your demand management strategy needs to be designed around.
The practical implication is that a demand event doesn't need to last long to count fully. A 12-minute HVAC cold-start ramp that stays elevated for 12 of the 15 minutes in an interval will capture nearly the full kW draw in that interval's average. Brief spikes of 2–3 minutes are partially damped by the averaging window — which is useful to understand when evaluating which events are worth managing and which are too short to matter.
The Rate Structure: Where the Multiplier Lives
Demand charge rates for commercial accounts vary significantly by utility and tariff class. In the Midwest, typical commercial demand charge rates run $8–22/kW/month, though large commercial and industrial accounts on separate rate classes may see $12–30/kW or higher. Some utilities apply different demand charge rates to on-peak versus off-peak demand intervals — meaning the same 15-minute spike costs more if it falls within defined peak hours (often 9 AM–9 PM weekdays) than if it occurs overnight.
This rate structure matters because it creates a hierarchy of risk. An HVAC startup spike at 7:30 AM — just before the on-peak window opens — generates a demand event at the standard rate. The same event at 9:15 AM triggers the on-peak demand rate, which may be 40–80% higher. Timing is not irrelevant to how expensive a demand event is; it can be the difference between a $140 demand event and a $240 demand event for the same kW draw.
Ratchet clauses add another layer of complexity. Some utility tariffs include a demand ratchet: even if this month's actual peak demand is lower than prior months, your demand charge may be calculated as a percentage (often 80–90%) of the highest demand recorded in the prior 11–12 months. This means a single high-demand month can elevate your billing for up to a year afterward — making demand event prevention more economically significant than it appears in a single-month analysis.
Where HVAC Sits in the Load Profile
In a typical commercial office building, HVAC represents 40–60% of total electricity consumption and an even higher share of peak demand events. Lighting, plug loads, and elevator systems contribute to the load profile, but their draw is relatively constant across occupied hours. HVAC is the variable-staging system — it's the one that cold-starts, ramps quickly, and creates demand spikes.
The morning pre-conditioning transition is the highest-risk window. When a building transitions from overnight setback (78–82°F cooling setpoint, minimal staging) to occupied setpoint (70–74°F), the chiller plant or DX units stage up rapidly. In a 100,000 sq ft office building with a 200-ton chiller and two 80-ton supplemental units, the morning startup can draw 350–450 kW for 15–25 minutes — easily the highest 15-minute average of the month.
Afternoon peak hours create secondary risk, particularly in summer cooling season when outdoor temperatures peak at 2–4 PM. But these afternoon peaks develop gradually as solar load accumulates and outdoor temperature climbs — the HVAC system ramps up incrementally, not all at once. The spike profile is shallower. The morning cold-start is categorically different: it's a step-change from low-draw to high-draw in a short window, which is exactly what demand charge mechanics are most sensitive to.
We're not saying HVAC is the only source of demand events — lighting retrofits and plug load management matter. We're saying HVAC startup behavior is usually the highest-amplitude, most predictable, and most controllable demand event in the building's monthly profile.
Reading Your Own Bill: What to Look For
To assess your building's demand charge exposure, you need three pieces of data that should be available from your utility or your BMS historian:
- Monthly peak demand (kW) by billing period: Track this for 12 months and look for the months where demand was highest. Winter months with cold-snap pre-conditioning and summer months with extreme cooling days tend to be the high-water marks.
- 15-minute interval data: Most utilities provide this through their commercial customer portal or through a green button data download. Plot it. The morning ramp pattern will be visually obvious — a tall, narrow spike in the 6–9 AM window, most pronounced on Mondays and post-holiday return days.
- Your demand charge rate and any ratchet clause: This is in your tariff schedule, which should be available from your utility's commercial rate page. Look for the demand charge rate, whether on-peak and off-peak rates differ, and any minimum billing demand or ratchet provisions.
Once you have 12 months of 15-minute interval data plotted alongside your BMS staging logs, the story is usually clear. The months with the highest demand charges have the most pronounced morning startup spikes, and those spikes correlate directly with overnight temperature drops and return-from-setback transitions.
The Interventions That Actually Move the Number
Demand charge reduction strategies fall into three categories: load shifting (moving consumption to lower-rate periods), load curtailment (reducing total consumption), and demand smoothing (flattening the peak without reducing total consumption). For HVAC, demand smoothing is usually the most practical and lowest-sacrifice approach.
The core mechanism is pre-conditioning: moving HVAC startup to begin earlier, before the peak tariff window opens, so the demand event occurs during off-peak hours. This doesn't save energy — the same BTUs of cooling capacity are delivered, just shifted earlier in the day. What it saves is demand charge dollars, because the same peak kW draw is recorded during off-peak intervals where the demand rate is lower, or before the on-peak window begins.
For this to work consistently, the pre-conditioning start time needs to be calculated daily based on actual forecast conditions — outdoor temperature at startup time, forecast temperature trajectory, expected occupancy level, and building thermal mass recovery rate. A fixed early-start schedule (say, 5:30 AM every day) adds unnecessary off-peak HVAC runtime on mild days without demand charge exposure, creating energy cost even when demand charge avoidance isn't needed. A forecast-driven pre-conditioning window starts early only when the demand risk warrants it.
That distinction — fixed early start versus forecast-conditional early start — is the operational difference between a rule-of-thumb energy strategy and a data-driven one. The building economics typically make a strong case for the latter within the first two or three billing cycles where the two approaches can be compared side by side.