The same 15 minutes that sets your demand charge is often the moment your equipment is working hardest. For multi-site operators, that is where cost and customer experience collide.
This summer, the grid set records. PJM, which serves 67 million people across 13 states and Washington, D.C., reported that demand on July 2 likely surpassed a record that had stood since 2006. ERCOT hit an all-time high of 91.1 GW on July 22. The Southwest Power Pool set its own record five days later.
Most facility managers experienced those weeks as a string of hot afternoons, a stack of service tickets, and rooftop units running flat out. What many won't fully see until later is the bill that follows. For sites on tariffs with a demand ratchet, a single 15-minute interval from one of those afternoons may set the minimum billed demand for the next 11 months.
It's September. The heat is easing. The cost of July is not.
Your electric bill has two main components. Energy charges bill you for how much electricity you used, measured in kilowatt-hours. Demand charges bill you for the highest rate at which you used it, measured in kilowatts, typically as the highest average draw during any 15-minute interval in the billing period.
That one interval gets multiplied by your demand rate, commonly somewhere between $8 and $25 per kW, and the result appears as a charge for the entire month. According to the National Renewable Energy Laboratory, demand charges can make up 30% to 70% of a commercial customer's electric bill, depending on the rate structure and how the site uses power.
Then there's the ratchet. Many commercial tariffs set a floor on billed demand equal to a percentage of the highest peak recorded over the previous 11 or 12 months. Floors between 60% and 80% are common.
Illustrative example: A quick-service restaurant normally peaks around 80 kW in summer. On one July afternoon, the late-morning opening sequence overlaps with a struggling rooftop unit, and the site hits 110 kW for 15 minutes. At $15 per kW, that adds $450 to the month's demand charges.
The bigger number comes later. Under an 80% ratchet, the site's billed demand can't fall below 88 kW until the floor resets. If actual winter demand runs around 60 kW, the site is paying for 28 kW it isn't using: roughly $420 a month, for as long as the floor holds.
One quarter hour. Months of cost.
It's tempting to treat a demand spike as a billing problem. Look closer at the interval data, though, and it usually turns out to be an equipment story. The conditions that create a peak are often the same conditions that push equipment toward failure. Three patterns show up again and again.
The opening stack. At many sites, everything comes on at once: rooftop units, fryers and ovens, hood fans, lighting, signage. Every compressor and motor start draws an inrush current several times higher than its normal running current, and startup is the most mechanically stressful moment in a compressor's working life. Stack enough of those starts into the same few minutes and you get a spike on the meter and a daily shock to the equipment.
The struggling unit. A rooftop unit with dirty coils, a low refrigerant charge, or a failing capacitor draws more power, runs longer, and often short cycles. Short cycling multiplies the number of hard starts, which accelerates wear on compressors, contactors, and capacitors. A unit that is quietly degrading often reveals itself in the demand profile well before it shows up as a no-cool call.
The catch-up surge. When a schedule fails, a setpoint gets overridden, or a walk-in door is propped open during a delivery, equipment has to recover. Recovery on a 95°F afternoon means multiple systems pulling hard at the same time. That's peak demand. It's also exactly when a marginal compressor is most likely to trip.
Customers never see a demand charge. What they notice is what happens when strained equipment finally gives out.
A dining room that's 81°F during the lunch rush. A theater auditorium that never pulled down before the 7 p.m. show. A reach-in cooler drifting toward the 41°F cold-holding limit, forcing staff to pull product. A kitchen crew working in heat that makes it hard to keep drive-thru times on pace.
Customers don't experience your portfolio average. They experience the one location they walked into, on the day they walked into it. For multi-site brands, consistency is the promise, and every site-level failure is a small break in it.
That's why demand peaks deserve attention from the facilities side, not just accounts payable. A site with an erratic demand profile is often a site with equipment under stress. And a site with equipment under stress is a customer experience incident waiting for a hot day.
September is the right moment for a post-summer demand review. The data is fresh, the patterns are visible, and there's still time to act before next cooling season.
Demand charges are one of the few costs on your bill that trace directly back to how equipment is operated and maintained at each site. That makes them worth managing for the dollars alone.
The bigger reason is what the peak is telling you. When a site spikes, something is stacked, strained, or struggling. The utility bills you for it this month. If nobody acts on it, your customers may be the next to notice.