Demand Charges and Margin Compression

Across a multi-site portfolio, demand charges behave less like a utility cost and more like an unmanaged liability: set by a handful of 15-minute intervals, locked in by tariff rules, and rarely owned by anyone with the authority to change them.

The line item nobody owns

Most executives think of electricity as a consumption cost. Sites use energy, the utility bills for it, and the number moves roughly with operating hours, weather, and traffic.

Demand charges don't work that way. They are set by the single highest 15-minute interval of power draw at each site in each billing period, and according to the National Renewable Energy Laboratory they can represent 30% to 70% of a commercial electric bill. That portion of the bill has almost nothing to do with how much business a location did that month. It reflects the worst quarter hour of how the site was operated.

That creates an ownership gap. Demand is set operationally, at the site, by opening procedures, equipment condition, and local overrides. It's paid by accounts payable. It surfaces in finance as a variance. And in most organizations, no single function is accountable for it, which means no one has written a policy for it.

How a 15-minute event becomes a 12-month cost

Many commercial tariffs include a demand ratchet: a floor on billed demand, set as a percentage of the highest peak recorded over the prior 11 or 12 months. Floors of 60% to 80% are common.

The practical effect is that a single afternoon can reprice a site for most of a year. This summer made that concrete. PJM reported that demand on July 2 likely broke a record that had stood since 2006, ERCOT set an all-time high on July 22, and the Southwest Power Pool followed on July 27. Sites that spiked during those weeks may now carry a demand floor well above what they will actually use this fall and winter.

From a financial standpoint, that is the problem in its sharpest form. A ratchet converts a variable operating event into a quasi-fixed cost, one that doesn't flex down when traffic softens or the weather cools. Few cost structures are worse for margin than a fixed cost created by accident.

The accumulation effect

At the level of a single site, a demand spike looks like a rounding error. At the level of a portfolio, it compounds.

Illustrative portfolio math: Consider a 150-site operator. Over one summer, 30 of those sites (one in five) each record a single avoidable spike of 25 kW above their normal peak. At $15 per kW, the peak months themselves cost a few hundred dollars per site. Under an 80% ratchet, each affected site then carries an extra 20 kW of billed demand, about $300 a month, until the floor resets.

Across the shoulder and winter months, that adds up to roughly $2,400 to $3,000 per affected site, or $72,000 to $90,000 across the portfolio before counting the peak months themselves. The events that created it lasted a combined seven and a half hours.

For an operator running a 4% net margin, recovering $90,000 in unplanned cost requires about $2.25 million in additional revenue.

That math assumes rates hold steady, and they aren't. As of August 2026, 57 electric and gas utilities across 48 states had rate increase requests pending, with an average requested increase of 13.2%. Rate cases reshape delivery charges, fixed fees, and demand charges alike. In the PJM region, capacity auctions have now cleared at their price ceiling three times running. Every increase in the underlying rate makes each uncontrolled peak more expensive.

Why this compresses margin, specifically

It doesn't scale with revenue. When same-store sales flatten, labor and food costs can be adjusted. A ratcheted demand floor can't. It stays in place regardless of what the location earns.

It undermines forecast accuracy. Demand charges are the most event-driven component of the energy bill. Budgets that treat energy as a single line tend to discover demand-driven variance only after the bills arrive, when the only remaining option is to explain it.

It concentrates in the sites least equipped to absorb it. The locations that spike most often tend to be the ones with aging or poorly maintained equipment. That means demand exposure and customer experience risk frequently show up in the same places. The sites leaking margin through demand charges are often the same sites most likely to have a hot dining room or a failed cooler on the next hot day. (The companion FM piece covers this operational link in detail.)

Why this requires executive policy

Facilities teams can reduce demand exposure site by site. They can't solve it at portfolio scale on their own, because the drivers cut across functions: operations sets opening procedures, local managers control overrides, maintenance budgets determine equipment condition, capital planning decides replacement timing, and procurement or finance typically owns tariff selection.

When a cost is created by several functions and owned by none, it takes a policy to close the gap. For most organizations, that policy has six parts.

• Assign ownership. Name a portfolio-level owner for demand exposure, with finance and facilities jointly accountable for results.
• Report demand separately. Break demand charges out from consumption in monthly energy reporting. Track billed versus metered demand, ratchet exposure by site, and the dates each floor resets.
• Set operating standards. Treat opening sequences, setpoints, and override limits as enterprise operating policy, in the same category as food safety or cash handling, rather than local preference.
• Link demand data to maintenance and capital. Sites with recurring peaks should move up the maintenance priority list and be reviewed explicitly in replacement planning.
• Review tariffs periodically. Rate schedules change, and sites change. A periodic review of rate class eligibility across the portfolio can surface sites billed on schedules that no longer fit their load.
• Budget with the floors you already have. Ratchets are known in advance. Forecasts for the coming months should start from the demand floors already locked in, not from last year's bills.

The cost you can actually control

Executives have learned to treat energy price as an external risk: something to hedge, forecast, and absorb. Demand charges belong in a different category. They are largely set by the organization's own operating behavior.

You can't control the price of natural gas. You can control what your sites do at 11 a.m. on a hot Tuesday in July.

That is what makes the absence of policy so expensive. Across a portfolio, demand charges are one of the few material energy costs that sit mostly within management's reach. The organizations that govern them deliberately will protect margin that their competitors will keep explaining as variance.