Demand Charges vs. Capacity Charges

A demand charge is a delivery fee the utility bills for the highest 15- or 30-minute average power draw a facility hit in the billing month, in dollars per kW. A capacity charge is a supply cost, passed through by the retail supplier, that pays generators to be available during the grid's peak; it is set by what the facility was drawing during the grid's few highest-demand hours the previous summer, and it applies for a full year. Both reward reducing peak load, but they respond to different peaks, so a facility can lower one without touching the other.

We get asked to explain this more than any other pair of line items, usually by a controller who has just found both on a bill and cannot see why reducing one did not reduce the other. The short version is above. The rest of this piece shows how each is calculated, what each costs on a typical commercial account, and which one is worth attacking first.

Demand charges: your peak, your utility, every month

The utility sizes its wires, transformers and substations to the highest load its customers will put on them, not the average. The demand charge recovers that cost from the customers who cause it. The utility's interval meter records average power in each 15-minute (sometimes 30-minute) window; the highest window in the billing cycle is the billed demand, in kW, and it is multiplied by the tariff's demand rate.

Three features matter:

• It resets every month. A bad 15 minutes in July costs you for July. A good August is billed on August.
• It is often ratcheted. Many commercial tariffs bill the greater of this month's peak and a percentage, commonly 60 to 80 percent, of the highest peak in the past eleven months. One spike can set a floor under the bill for a year.
• It is in the delivery section of the bill. Switching suppliers does not change it. Only the facility's own behavior or a change of rate class does.

Our note on demand charges goes through a bill line by line. The point here is that this charge is about the facility's own worst interval, whenever it happens.

Capacity charges: the grid's peak, your supplier, all year

In the organized markets of the Northeast and Mid-Atlantic, PJM, NYISO and ISO-NE, the grid operator runs a capacity market that pays generators to be available when the system peaks. The cost is allocated to retail customers through a capacity tag, also called a peak load contribution or installed capacity tag. The tag is the facility's average demand during the grid's handful of highest-demand hours in the previous summer: five hours in PJM, one in NYISO, one in ISO-NE. It is set once a year and it applies to every month of the following delivery year.

The supplier multiplies the tag by the capacity price the grid operator's auction produced and recovers it in the supply rate, either as a separate line, as a pass-through, or buried in a fixed price. In PJM the auction price for the 2025-26 delivery year cleared at $269.92 per MW-day and for 2026-27 at $329.17 per MW-day, the auction's price cap, which is why capacity has become the fastest-growing component of supply cost across Pennsylvania, New Jersey, Maryland, Ohio, Illinois and the District. Our note on PJM capacity charges covers the auction.

Three features matter here too, and they are the mirror image of the demand charge:

• It is set by hours you do not choose. The grid operator identifies the peak hours after the summer ends. A facility that was running flat out on those afternoons carries the tag for the next year regardless of what it does in any other month.
• It does not reset monthly. One summer's behavior sets twelve months of cost.
• It is in the supply section. How a supplier handles it is a contract term. A fixed all-in price absorbs the tag at signing; a pass-through contract re-prices it when the tag or the auction price changes.

Side by side

What each costs on a real account

Take a cold storage warehouse in New Jersey with a 600 kW monthly peak and a 500 kW average draw during last summer's five PJM peak hours. If the utility's demand rate is $12 per kW, the demand charge is about $7,200 a month, $86,400 a year. If the capacity price for the delivery year is $329.17 per MW-day, the capacity cost is 0.5 MW multiplied by $329.17 multiplied by 365, about $60,000 a year, before the supplier's reserve margin adjustments and losses push it higher. Both are large. They are set by different afternoons.

Which one to attack first

The demand charge responds to any reduction in the facility's own monthly peak: staggering compressor starts, shifting a batch process off the busiest hour, adding controls so HVAC and process load do not peak together. Every month that the peak comes down, the bill comes down, subject to the ratchet.

The capacity tag responds only to reductions during the grid's peak hours, which are a few hot weekday afternoons between June and September. A facility that can predict those afternoons and curtail during them, which is what demand response programs and peak-alert services exist to do, resets its tag for the following year. Cutting load on a cool day in October does nothing to it.

The practical order is: first, make sure the contract handles capacity the way the facility wants, fixed or passed through, because that is a signature rather than an engineering project. Second, pursue demand charges through operational changes, because the payback is monthly. Third, if the facility can curtail on peak afternoons, enroll in a program that pays for it and reduce the tag at the same time. Our peak load management work covers the second and third steps; the capacity charges explainer covers the first.