Most of a commercial electric bill comes in two kinds of charge. The energy charge pays for how much electricity a site used. The demand charge pays for how fast it drew that electricity at its busiest moment.
The first is intuitive. The second is where multi-site teams get surprised, because two sites using the same energy can pay very different bills.
kW and kWh in one picture
kWh measures energy: the total amount used over the billing period. kW measures demand: the rate of use at a point in time.
A useful picture is water. kWh is how many gallons went through the pipe this month. kW is how wide the pipe had to be for the busiest few minutes. The utility builds wires, transformers and generating capacity for the widest moment, not the average, and the demand charge is how it recovers that cost.
In this example, two stores each use 30,000 kWh in a month. One draws a steady 60 kW for 500 hours. The other draws 150 kW in a short burst every morning and much less the rest of the day. Same energy, very different demand.
How the utility measures demand
The National Renewable Energy Laboratory explains that demand charges are typically based on the highest average electricity usage in a defined interval, usually 15 minutes, during a billing period.
Two details in that sentence matter. It is an average over the interval, so a spike lasting seconds counts for less than a load that runs the full 15 minutes. And it is the single highest interval in the period, so one bad quarter hour sets the charge for the month.
Your tariff defines the rest: the interval length, whether demand is measured all day or only in peak hours, and how billed demand is calculated from metered demand.
Why demand can dominate a bill
The same NREL analysis looked at more than 10,000 utility tariffs in 48 states and found that demand charges “can account for anywhere from 30% to 70% of a customer’s electricity bill.”
That range is wide because sites differ. A site that runs steadily around the clock spreads its demand charge across a lot of kWh. A site with a sharp short peak pays a similar demand charge on far fewer kWh, so demand becomes the larger part of its bill.
One site’s month, priced
Here is one site’s month, with clearly illustrative round prices. Check your current tariff for real ones.
| Line | Quantity | Example price | Charge |
|---|---|---|---|
| Energy | 40,000 kWh | $0.08 per kWh | $3,200 |
| Demand | 200 kW | $15 per kW | $3,000 |
| Total | $6,200 |
In this example, demand is almost half the bill. Now suppose that on one July afternoon three rooftop units, a walk-in cooler defrost and an EV charger all run in the same 15 minutes, and the peak reaches 260 kW. The energy barely changes. The demand charge rises by 60 kW times $15, or $900, for that one quarter hour.
Run your own numbers in our demand charge calculator. It shows the demand share of a bill, the load factor and the cost of each 10 kW of peak, and it runs in your browser without sending anything.
Load factor tells the story
Load factor compares a site’s average demand to its peak. The formula is kWh divided by the peak kW times the hours in the period.
In the example above, 40,000 kWh divided by 200 kW times 720 hours gives about 28%. A site running near flat would sit much higher. A low load factor means the site is paying for capacity it uses briefly, which is the signature of a site where demand charges deserve attention.
Load factor is also the quickest way to compare sites across a portfolio. Sort your sites by it. The ones at the bottom are usually where a peak is worth chasing.
Ratchets, minimums and peak windows
Billed demand is not always the metered peak. Tariffs add rules, and they vary by utility and schedule.
- Ratchets. Some schedules bill demand at the higher of this month’s peak or a set share of the highest peak in prior months. A summer peak can keep costing through winter.
- Minimum demand. Some schedules set a floor on billed kW, or a minimum bill.
- Peak-window demand. Time-of-use schedules may charge demand only, or more heavily, during on-peak hours. A peak at 6 a.m. may cost less than one at 4 p.m.
- Power factor adjustments. Some tariffs raise billed demand when a site’s power factor falls below a set level.
Read the “determination of demand” section of your schedule before you plan anything. It tells you which peak counts, and for how long.
Two common misreads
Cutting kWh does not always cut demand. A lighting retrofit that lowers overnight load reduces energy charges, but if the monthly peak happens on a weekday afternoon with every system running, billed demand may barely move. Measure where the peak is before you count on a project to lower it.
Turning things off at night does not help a daytime peak. Night setbacks save energy. They can also make the morning peak worse, because every unit starts at once to recover the setpoint. The fix may be a staggered recovery rather than a deeper setback.
Both misreads come from looking at monthly totals. Interval data shows when the peak happens and what else was running, which is the only way to know whether a change will touch it.
Reading demand on the bill
Most demand-metered bills show at least two kW figures, and some show more.
- Metered or actual demand: the highest interval the meter recorded.
- Billed demand: the kW the utility charged, after any ratchet, minimum or adjustment.
- On-peak demand: on time-of-use schedules, the highest interval inside the peak window.
When billed demand is higher than metered demand, the tariff rule behind the gap is the thing to understand first. It may mean last summer is still setting this winter’s bill.
What usually sets the peak
Peaks tend to come from loads that start together. Common sources across commercial portfolios:
- Rooftop units all starting at once after a night setback.
- EV chargers plugged in at shift change.
- Kitchen equipment and dish machines at dinner.
- Air compressors, dock doors and forklift battery chargers in a warehouse.
- Electric heat on the coldest morning of the month.
Often the first fix is scheduling, not equipment: staggered start times, a charger limit, or a sequence for kitchen start-up. Equipment changes, controls and storage belong with a licensed engineer or contractor, and any decision to install should rest on their design.
Why multi-site teams see it differently
With one building, a demand charge is a line on one bill. With 40, it is a pattern. Sites on the same schedule, with similar size and hours, should have similar load factors. The outliers tell you where to look.
A portfolio view also shows which sites share a cause. If every store’s peak lands within 15 minutes of opening, the fix is one change to an opening procedure, applied 40 times.
It also changes who owns the problem. A demand peak is set by operations, billed to facilities and paid by finance. Each team sees one piece. Putting the peak time, the load behind it and the dollar cost on one page is usually what gets a change made.
Where to start this week
- Pull peak kW and kWh from the last 12 bills for each demand-metered site.
- Calculate load factor per site and sort the list.
- For the bottom five, request interval data from the utility.
- Find the day and time of each month’s peak.
- Ask the site team what runs at that hour.
That is the beginning of tariff analysis: reading the schedule, the interval data and the operations together, so the answer comes from the tariff rather than a guess.
Plan the peak, site by site
The Demand Charge Plan does this work in depth for one site: 12 months of interval data, the tariff’s demand rules, the loads behind each month’s peak, and a written plan showing what each change would have meant on last year’s bills. If you are not sure which site to start with, send three bills for a free Three-Bill Read.
