Your utility bill has two parts that most people focus on and one part that often matters more. The energy charge — the dollars per kWh you consumed — gets the attention. The demand charge — the dollars per kW based on your peak 15-minute demand interval — quietly determines a large fraction of what commercial buildings actually pay.
For a 100,000 sq ft commercial office building in Oregon, the demand charge component can run $2,500–$5,000 per month, representing 25–40% of the total electric bill. That number is set by a single 15-minute window somewhere in the billing period. Understanding exactly how this works changes how you think about HVAC scheduling.
The Mechanics: How Your 15-Minute Peak Gets Measured
Your utility measures electrical consumption at your meter in 15-minute intervals, 24 hours a day, every day of the billing period. At the end of the month, they find the highest 15-minute average power draw (measured in kilowatts) during that period — or, if you're on a time-of-use tariff, the highest interval during the designated peak demand window (often 4 PM to 8 PM on weekdays).
That single peak interval value is multiplied by the demand charge rate — which in Oregon's PGE commercial tariff schedules runs roughly $8–$18/kW depending on rate schedule and season. The result is your demand charge for the entire month.
The implications are asymmetric in a way that trips up facilities managers who think about energy in terms of kWh: it doesn't matter how efficiently you operated the other 2,975 intervals in the month. One interval at 200 kW sets the same demand charge whether the rest of your month was 100 kW or 180 kW. You pay for your peak, not your average.
What Creates a Demand Peak in a Commercial Building
Demand peaks in office buildings have recognizable patterns. The two most expensive and most preventable:
Morning HVAC recovery. After a night or weekend setback, the BMS brings all systems to full operation simultaneously at 6 or 7 AM. Air handlers ramp up, chillers start, reheat coils activate across multiple zones. The simultaneous startup draws the building's near-maximum electrical load for 30–60 minutes. In a building with a 250 kW HVAC capacity, the morning startup might pull 200+ kW for 45 minutes — enough to set the month's demand charge on an otherwise moderate-consumption day.
Coincident with tariff peak window. In the summer cooling season, buildings running full HVAC cooling at 3–5 PM are doing so exactly when demand charges are highest (the utility's peak window, typically afternoon hours). A building that could pre-cool to comfort temperature before 3 PM — using its thermal mass to coast through the peak window — would draw much lower demand during the highest-cost interval.
There are secondary contributors: elevator banks starting simultaneously, electric vehicle charging loads without managed timing, exterior lighting on shared circuits with HVAC. But HVAC is the dominant driver for most commercial buildings, representing 40–60% of total electrical consumption and an even higher share of demand peaks.
The Ratchet Clause: When One Bad Month Follows You
Some commercial utility tariffs include a ratchet clause — a provision that sets a minimum billing demand based on a percentage of your highest demand in the past 11 or 12 months. On PGE's Schedule 29 large commercial tariff, for example, the billing demand cannot fall below a percentage of the maximum recorded demand over the preceding 11 billing periods.
What this means in practice: if your building sets a demand peak of 350 kW during a heat event in August, your minimum billing demand for the next eleven months may be set at 60–70% of 350 kW — meaning you pay for 210–245 kW minimum even in January when your actual peak might be 140 kW. A single uncontrolled peak demand event has twelve months of billing consequences.
We're not saying the ratchet clause is the primary reason to manage demand — the direct per-month charge is usually larger than the ratchet differential. But it illustrates how the economic structure of commercial tariffs punishes demand spikes in a way that is non-obvious from reading just the rate schedule.
Time-of-Use Demand vs. Flat Demand: What Rate Schedule Are You On?
Not all demand charges work the same way. There are two main structures you'll encounter on Oregon commercial tariffs:
Flat demand. Your peak 15-minute interval anywhere in the billing period sets the demand charge. The clock doesn't matter — a demand spike at 2 AM costs the same as one at 4 PM. Buildings on flat demand schedules benefit from staggering equipment startups to avoid coincident peaks, but time-of-use pre-cooling strategies don't directly apply.
Time-of-use (TOU) demand. Only intervals falling within the designated peak demand window (e.g., 7 AM to 10 PM weekdays, or a narrower afternoon window) count toward the peak demand calculation. Off-peak intervals are ignored. This structure creates the pre-cooling opportunity: if you can move your HVAC load from 3–6 PM to 10 AM–1 PM, you reduce the intervals that count for demand charges even if you consume the same total kWh.
If you're unsure which structure your building is on, look at your utility bill for a line item called "Demand (kW)" or "On-Peak Demand." If the charge is the same rate regardless of time, you're likely on a flat demand schedule. If you see separate line items for on-peak demand and off-peak demand, or a time qualifier in the demand charge description, you're on TOU demand.
Calculating Your Demand Charge Exposure
Here is a quick calculation framework for any commercial building:
Pull your last 12 monthly utility bills. Find the demand charge line on each. Divide that number by the kW figure shown for that month's peak demand to find your effective demand charge rate. Then identify the months with the highest demand peaks — those are the months where pre-conditioning intervention has the highest dollar impact.
For a building paying $14/kW in demand charges and setting a monthly peak of 180 kW, the demand charge is $2,520 that month. If a pre-cooling strategy could reduce the peak demand interval to 130 kW, the savings on demand charges alone would be $700 in that month. Multiply that across 8–10 months per year where HVAC loads are a meaningful demand driver, and you're looking at $5,600–$7,000 per year in demand charge avoidance from one intervention.
Those numbers are specific to that building's rate schedule and demand profile. Run the same calculation on your own bills before evaluating any optimization approach — the math either works clearly or it doesn't.
Why kWh Efficiency Projects Often Miss Demand Charges
Most energy efficiency programs — LED retrofits, high-efficiency equipment upgrades, occupancy-based lighting controls — focus on kWh reduction. These are legitimate improvements. But they often don't move the demand charge much, because demand charges are determined by peak interval power draw, not total consumption.
A building that replaces its fluorescent lighting with LED (reducing consumption by 30%) and also has an uncontrolled Monday morning HVAC ramp-up at 6 AM will have a lower energy charge but approximately the same demand charge. The two metrics respond to different interventions.
Demand charge reduction requires managing when loads occur — specifically, flattening the peak demand intervals by moving discretionary loads out of the highest-cost windows. That is a scheduling and control problem, not an equipment efficiency problem. It's also why we focus on predictive pre-conditioning rather than hardware: the marginal cost of demand charge avoidance through scheduling is much lower than the marginal cost of reducing it through equipment upgrades.
The Optimization Window
The opportunity for buildings on time-of-use tariffs is specific: the gap between when comfort conditioning needs to be achieved (before occupancy) and when the tariff peak window opens (typically afternoon for commercial buildings). A building that needs to be at 71°F by 8 AM doesn't need to run its chillers at 4 PM — it needs to cool aggressively at 5–7 AM, then maintain. On a hot day where the afternoon cooling load is significant, the opportunity is to pre-cool as far below target temperature as comfort allows before the peak window, using the building's thermal mass to absorb afternoon heat without running chillers at full load during the highest-rate period.
That strategy requires knowing three things simultaneously: the exact tariff schedule, the building's thermal response time, and tomorrow's weather forecast. When those three inputs are available and correctly weighted, the 15 minutes that currently set your demand charge can be moved to a time window that costs significantly less.