Demand Charges Explained: How to Cut Your Commercial Electric Bill

Demand charges can be 30-70% of a commercial electric bill. Here is exactly how they are calculated and seven proven ways to shrink them.

What Is a Demand Charge?

A demand charge is a fee on your commercial electricity bill based on the highest rate of power your facility pulled at any single moment during the billing period, not on how much total energy you consumed. So what is a demand charge in plain terms? The energy charge (in $/kWh) is like the total distance you drove this month. The demand charge (in $/kW) is the top speed you hit, even for a few minutes.

Utilities bill this way because they must build and maintain generation, transmission, and distribution capacity sized to your peak draw, not your average. A plant that averages 200 kW but spikes to 600 kW when three compressors and a chiller start together still forces the grid to stand ready for 600 kW. The demand charge recovers the cost of that reserved capacity.

For most commercial and industrial accounts, demand charges run roughly 30% to 70% of the total electric bill, a share that surprises managers who watch only the cents-per-kWh line. Cutting peak demand is therefore the highest-leverage move on the whole invoice. Breaking your bill into its components is the single most valuable hour you can spend this quarter and the foundation for every demand charge management strategy below.

How Peak Demand Is Measured: Intervals and Ratchets

The mechanics decide which tactics actually work. Your meter does not capture an instantaneous spike. It averages power over a fixed demand interval, most commonly 15 minutes, though 30-minute and 60-minute intervals are still used by some utilities. Your billed demand is the single highest interval average recorded during the billing period.

That averaging window is good news. A motor inrush lasting two seconds barely moves a 15-minute average. But four large loads running together for 20 minutes will set a new peak. The goal of demand charge management is to make sure your big loads never coincide inside one interval.

The demand ratchet: the trap that surprises people

Many tariffs include a demand ratchet: your billed demand for the month is set to the higher of (a) this month's actual peak, or (b) a percentage, typically 75% to 95%, of the highest peak recorded over the previous 11 or 12 months. One uncontrolled 15-minute spike in July can therefore inflate your demand charge through the following spring, even in months when real usage is low. Ratchets make a single bad interval extraordinarily expensive and are the strongest argument for active peak management.

Coincident vs. non-coincident demand

Non-coincident demand is your facility's own highest interval. Coincident demand is your draw at the moment the utility's system peaks, such as a hot August weekday afternoon. Some tariffs bill on coincident demand, which is precisely why demand response programs, which pay you to drop load during grid peaks, exist.

How to Read the Demand Line on Your Bill

Pull your most recent invoice and find these line items. Labels vary by utility, but the structure is consistent.

These ranges are typical U.S. illustrative figures; your actual tariff governs, and rates differ widely by region and rate class. To put a dollar value on shaving your peak, drop your billed kW and rate into our free Demand Charge Savings calculator, which shows monthly and annual savings for any reduction you achieve.

Do the math first. If your demand charge is $18/kW-month and you bill 500 kW, that single line is $9,000 per month, or $108,000 a year. Cutting the peak by just 10% (50 kW) returns about $10,800 a year, before touching a single kilowatt-hour of consumption. That is why peak shaving usually beats energy-efficiency tweaks on a pure-dollar basis.

Strategy 1: Peak Shaving and Load Shifting

Peak shaving means reducing your highest demand intervals; load shifting means moving flexible loads to off-peak periods. Together they are the highest-leverage, lowest-cost lever most facilities have, because they attack the billed quantity directly.

• Stagger large motor and equipment starts. Sequencing two 150 kW compressors 10 minutes apart instead of together keeps both out of the same demand interval. This is often just a control-system programming change, essentially free.
• Interlock non-critical loads. Use a demand limiter or building automation system to shed water heaters, battery chargers, EV chargers, or secondary HVAC when real-time demand approaches a setpoint.
• Shift batch processes off-peak. Charging, pumping, grinding, or curing that need not run at 2 p.m. can run overnight, flattening the demand curve and often capturing cheaper time-of-use energy too.
• Pre-cool or pre-heat. Driving HVAC harder before the on-peak window and coasting through it shifts demand without sacrificing comfort, a long-recognized demand-limiting technique in ASHRAE guidance.

An ISO 50001 energy management system formalizes this with a plan-do-check-act loop and an energy baseline, turning one-time wins into a sustained program. Even without certification, tracking energy performance indicators against a baseline is what keeps peaks from creeping back up.

Strategy 2: Fix Power Factor and Reactive Demand

Inductive loads (motors, transformers, fluorescent ballasts, welders) draw reactive power (kVAR) that does no useful work but inflates the current the utility must deliver. The ratio of useful power (kW) to total apparent power (kVA) is your power factor (PF): PF = kW ÷ kVA. A facility full of lightly loaded motors can sit at a PF of 0.75–0.85.

Low power factor hurts two ways. First, many tariffs levy a direct penalty, or bill demand on kVA instead of kW, when PF falls below a threshold (commonly 0.90 or 0.95). Second, poor PF wastes capacity in your own cables and transformers, adding voltage drop and heat.

The fix is power factor correction: installing capacitor banks, or automatic stepped capacitor controllers, near the inductive loads to supply reactive power locally. Correcting from 0.80 to 0.95 cuts apparent power (kVA) for the same real load by about 16% (1 − 0.80/0.95), which can directly reduce a kVA-billed demand charge and remove penalties. Use our Power Factor Correction calculator to size the required kVAR and estimate savings for your starting and target PF.

One safety note: capacitor banks store energy and can interact with harmonics from variable-speed drives. Size and commission them with a qualified engineer, follow NFPA 70E for safe work practices, and base PPE on an arc-flash incident-energy study (IEEE 1584 method) for the affected switchgear. On harmonic-rich systems, detuned reactors may be needed to avoid resonance.

Strategy 3: VFDs, Efficient Motors, and Soft Starts

Motors are the largest single end use of electricity in industry, and how they start and run shapes your demand curve. Three upgrades pay off:

• Variable frequency drives (VFDs). On centrifugal pumps and fans, the affinity laws mean shaft power scales with roughly the cube of speed, so trimming flow to 80% can cut shaft power to about half (0.8³ ≈ 0.51). VFDs also ramp motors up smoothly, eliminating the inrush spikes that nudge demand intervals upward. Model the energy and demand impact with our VFD Energy Savings calculator before you buy.
• Premium-efficiency motors. When replacing a failed motor, specify a higher IE class. The IEC 60034-30-1 standard (closely aligned with NEMA Premium) defines the IE efficiency bands; moving up a class trims continuous draw and the heat it produces.
• Soft starters. Where a VFD is not justified, a soft starter limits inrush current at startup, protecting both equipment and your demand profile.

The biggest demand wins come from pairing a VFD with a right-sized motor on a high-runtime fan or pump: you cut both kWh and the peak that motor contributes to.

Strategy 4: Lighting, HVAC, and Building-Level Efficiency

Every kilowatt you remove from your baseline lowers the floor your peaks build on, so permanent efficiency upgrades quietly shrink demand month after month.

• LED retrofits. Replacing legacy HID or fluorescent lighting typically cuts lighting energy by 50–70%, and because lights usually run during the on-peak window, that reduction lands squarely on demand. Our LED Retrofit Savings calculator estimates the connected-load reduction, energy savings, and payback. Occupancy sensors and daylight harvesting compound the win.
• HVAC optimization. Economizers, demand-controlled ventilation, optimized start/stop, and tightened setpoints follow ASHRAE guidance and cut the single largest swing load in most commercial buildings.
• Thermal and battery storage. Ice or chilled-water storage made overnight, or a behind-the-meter battery, discharges during peak hours to flatten demand: effectively automated peak shaving for facilities with predictable load shapes.
• Compressed-air and process tuning. Fixing leaks, lowering pressure setpoints, and properly sequencing compressors trims one of the most expensive and overlooked loads in any plant.

Track these against a documented baseline. The EN 15341 maintenance KPIs and the ISO 55001 asset-management framework help connect equipment condition to energy performance, so degraded bearings, fouled coils, or clogged filters, each of which raises power draw, get caught before they inflate your peaks.

Strategy 5: Demand Response and Tariff Optimization

Beyond shrinking your own peaks, you can get paid to manage them and confirm you are on the right rate.

Demand response

Demand response (DR) programs pay commercial customers to curtail load during grid emergencies or system peaks. You commit to dropping a set number of kilowatts when called, by idling non-critical equipment, dimming lighting, or running a generator, and you receive capacity payments plus per-event energy payments. For facilities with flexible load, DR turns peak management into a revenue line rather than just a cost avoided. Programs are run by utilities and grid operators (ISOs/RTOs), often through aggregators known as curtailment service providers.

Tariff and rate-schedule review

Utilities offer multiple rate schedules, and you are not always on the cheapest one for your load shape. Worth checking:

• Time-of-use (TOU) demand rates that bill demand only during on-peak hours, ideal if you can shift load to nights and weekends.
• High-load-factor schedules that reward steady, flat consumption with lower demand rates.
• Standby or backup charges that apply if you add on-site generation or storage; understand these before investing.
• Whether a ratchet applies, and how many months it persists, so you know the true cost of a single spike.

A rate analysis using 12 months of interval data, which you can request from your utility, often surfaces a better-fit schedule with no capital spend at all: the cheapest savings you will ever find.

Putting It Together: A 5-Step Action Plan

Demand charges reward a structured approach over scattered fixes. Work the list in order; each step funds the next.

1. Measure. Get 12 months of interval (15-minute) data from your utility and break each bill into energy, demand, and power-factor components. You cannot manage a peak you cannot see.
2. Quantify the target. Identify your top peaks and the equipment that caused them. Run the Demand Charge Savings calculator to put a dollar value on each kW you could shave, then prioritize by payback.
3. Capture the free wins. Sequence equipment starts, set demand limits in your automation system, and review your rate schedule. These cost little and deliver the fastest return.
4. Invest where payback is clear. Use the Power Factor Correction, VFD Energy Savings, and LED Retrofit Savings calculators to rank capital projects, then fund the shortest-payback ones first.
5. Sustain it. Adopt an ISO 50001-style review cadence so new peaks are caught early and savings do not erode. A five-minute monthly demand review prevents the one bad interval a ratchet would punish for a year.

Done well, this sequence routinely takes 10–30% off demand-related costs, real money that drops straight to the bottom line without cutting a minute of production. When you are ready to evaluate a specific upgrade or service, get itemized quotes from at least three vendors, ask for a guaranteed-savings or measurement-and-verification clause, and validate every claim against your own interval data before you sign.

Related free calculators

Frequently asked questions