Article

Coincident Peak Demand: Facility Peaks vs. Grid Peaks

Understand coincident peak demand in 2026, compare facility and grid peaks, calculate interval kW, and check demand-charge savings.

Published October 11, 2026

Coincident peak demand is your facility’s electrical demand during the same interval that a utility or grid system reaches its defined peak. Running equipment together can also raise your facility’s own peak, but that is not automatically a coincident peak for billing purposes. Before changing operating schedules, identify whether your bill charges for your highest facility demand, demand during designated system peaks, or both; those rules determine which load reductions can save money.

Facility Peak vs. Coincident Peak: What Changes the Charge

Billing basis or operating comparison	Demand being measured	Relevant interval or condition	Financial implication
Facility maximum demand, often called non-coincident peak demand	Your highest average demand within the tariff’s eligible measurement periods	Confirm the interval length, billing period, and any time-of-use restrictions	Reducing this peak can lower a demand charge, unless a ratchet, minimum, or other adjustment controls billing demand.
System coincident peak demand	Your demand during the utility’s or grid operator’s selected peak intervals	The selected system intervals—not necessarily your facility’s busiest intervals	Reducing demand outside the selected intervals may not reduce this component.
PJM five coincident peaks, or 5CP	Customer demand associated with five selected summer system peak hours	Eligible non-holiday weekdays during June–September; distribution-company allocation rules determine the customer’s final capacity obligation	Lower demand during the selected hours can reduce future capacity costs, subject to allocation factors and contract terms.
Hypothetical U.S. facility: two loads operated together	70 kW facility peak	15-minute fixed intervals; assumed demand rate of USD 15/kW per billing month	USD 1,050 demand-charge component
Same hypothetical facility: two loads operated in separate intervals	50 kW facility peak	Same tasks and energy use; no other interval exceeds 50 kW; no billing ratchet or minimum	USD 750 demand-charge component; conditional reduction of USD 300 for that billing month

The final two rows are an illustrative calculation, not a utility rate quote or a national benchmark. The reduction applies to a simple facility-maximum demand charge. It does not establish savings under an hourly coincident-peak charge.

Timing is the deciding condition: a lower equipment overlap matters financially only when it lowers the demand quantity your tariff or supply contract actually bills.

What “Coincident” Means

A facility peak and a system peak answer different questions. Your facility peak identifies when your own average demand was highest. A system coincident peak identifies your demand when the larger system reached a selected peak, even if your facility was operating below its own maximum. PJM publishes selected system peak dates and hours to support electric distribution companies in calculating customer Peak Load Contributions, or PLCs. Equipment operating at the same time is a separate, related concept. An oven, compressor, and charging station can overlap and increase the meter’s interval-average demand. That overlap can explain a high facility peak, but equipment simultaneity alone does not identify a utility’s coincident peak. For scheduling decisions, use measured operating demand and actual run times rather than treating “probability of simultaneous operation” as a billing formula. Nameplate ratings can help inventory equipment, but they do not demonstrate how much energy each device used during the interval that established the charge.

Why the Distinction Matters

A schedule change can reduce your highest 15-minute demand without reducing your demand across a full system peak hour. Conversely, reducing load during a selected system peak hour can lower a coincident-peak contribution even when your facility’s monthly maximum occurs on another day. The measurement window and the selected timestamps must match the charge being evaluated. That is why a facility should evaluate monthly demand charges and coincident-peak capacity charges separately rather than applying one peak-reduction estimate to every demand-related bill line.

Inputs Needed Before Calculating Savings

Input	Unit or format	Where to obtain it	Why it matters
Interval energy or average demand	kWh per interval or kW	Utility interval-data export	Establishes the measured load profile
Demand measurement interval	Minutes or hours	Applicable tariff and meter documentation	Determines how energy becomes billing demand
Timestamp convention	Interval start/end, time zone, daylight-saving convention	Data-export documentation	Aligns facility readings with billing windows and system peaks
Meter multiplier	Dimensionless	Bill or utility data documentation	Prevents undercounting or double-counting usage
Billing-demand definition	Tariff rule	Current utility tariff	Identifies eligible periods, ratchets, minimums, and adjustments
Demand rate	USD/kW per stated billing period, or another explicitly defined unit	Current tariff or supplier contract	Converts the applicable billing quantity into a charge
Coincident-peak event timestamps	Dates and selected intervals	Utility or grid operator	Identifies the facility readings relevant to coincident demand
PLC or capacity-allocation method	Utility-specific calculation	Distribution company and supplier	Determines how measured peak contributions become billable obligations

Duke Energy’s Ohio business-bill guide explains the distinction between actual demand and billing demand, including the effects of meter multipliers, demand ratchets, and power-factor adjustments. That guide is dated August 2015; use it for the measurement concepts, not as evidence of current 2026 rates or tariff provisions. For current rates, begin with the utility’s official tariff portal and select the actual service location. For example, Duke Energy’s business rate information is location-specific. There is no universal U.S. coincident-peak rate to substitute for your account’s tariff or contract.

How to Find and Verify a 15-Minute Facility Peak

Obtain the complete interval-data export for the billing period. A short sample can explain a peak, but it cannot prove the monthly maximum.
Confirm whether each record contains interval kWh, cumulative kWh, or average kW. For cumulative readings, subtract consecutive readings before calculating interval demand.
Check the interval duration, timestamp convention, time zone, and whether the meter multiplier has already been applied.
Convert interval energy to average demand when needed.
Identify the highest demand within the periods eligible under the tariff.
Apply the tariff’s billing-demand rules and reconcile the result with the bill.
Model the proposed schedule across the complete billing period, including delayed loads and recovery operation. For an interval containing energy (E):

[ P_{\text{interval}}=\frac{E_{\text{interval}}}{\Delta t} ]

where (P) is average demand in kW, (E) is energy in kWh, and (\Delta t) is interval duration in hours. For a 15-minute interval:

[ \Delta t=\frac{15}{60}=0.25\text{ hour} ]

[ P_{\text{interval}}=4E_{\text{interval}} ]

An interval containing 17.5 kWh therefore represents:

[ P_{\text{interval}}=\frac{17.5}{0.25}=70\text{ kW} ]

This is average demand across the interval, not an instantaneous reading. Duke Energy’s Ohio guide illustrates the same distinction: a load profile reaching 36 kW briefly can produce an integrated 15-minute demand of only 28 kW.

A short motor-starting spike is not automatically your billed demand. Integrated demand averages energy use over the measurement interval, so magnitude and duration both matter. Likewise, moving a start time by a few minutes does not establish savings unless the resulting metered averages and billing demand fall. The worked example below assumes fixed 15-minute blocks. Confirm the utility’s measurement method before relying on a schedule aligned to quarter-hour boundaries.

Worked Example: Staggering Two Loads

Assumptions

This hypothetical U.S. commercial facility has:

A constant background load of 20 kW.
Load A drawing 30 kW for 15 minutes.
Load B drawing 20 kW for 15 minutes.
Fixed 15-minute demand intervals.
A flat demand charge of USD 15/kW per billing month.
No demand ratchet, minimum billing demand, power-factor adjustment, or separate coincident-peak charge.
No other interval in the billing period exceeding the peaks shown.
No change in task energy, operating efficiency, or recovery load when the tasks are staggered. The equipment powers are assumed actual operating inputs, not motor output ratings.

Interval Data Before and After Staggering

The following data are synthetic, not a utility-meter export. Times identify interval starts in the same local time zone.

Interval	Original operation	Original energy	Original average demand	Staggered operation	Staggered energy	Staggered average demand
14:00–14:15	Background + A + B	17.5 kWh	70 kW	Background + A	12.5 kWh	50 kW
14:15–14:30	Background only	5.0 kWh	20 kW	Background + B	10.0 kWh	40 kW
14:30–14:45	Background only	5.0 kWh	20 kW	Background only	5.0 kWh	20 kW
14:45–15:00	Background only	5.0 kWh	20 kW	Background only	5.0 kWh	20 kW
One-hour total or maximum		32.5 kWh	70 kW maximum		32.5 kWh	50 kW maximum

The schedule preserves the same two tasks and the same energy use. It reduces the highest 15-minute average because the two loads no longer occupy the same measurement interval.

Hand Calculation

Before staggering:

[ P_{\text{before}}=20+30+20=70\text{ kW} ]

After staggering:

[ P_{\text{after}}=\max(20+30,;20+20,;20,;20)=50\text{ kW} ]

Demand reduction:

[ \Delta P=70-50=20\text{ kW} ]

For the assumed simple demand-charge structure:

[ C_{\text{demand}}=P_{\text{billing}}\times r_{\text{demand}} ]

[ C_{\text{before}}=70\times15=\text{USD }1{,}050 ]

[ C_{\text{after}}=50\times15=\text{USD }750 ]

[ \Delta C=20\times15=\text{USD }300 ]

The estimated reduction is USD 300 in the demand-charge component for that billing month—not USD 300 in energy savings. Both schedules consume 32.5 kWh during the illustrated hour. Fixed charges, taxes, riders, and any energy-price differences are outside this calculation. Use the demand-charge calculator to check the simple billed-kW-times-rate calculation with your own inputs. A calculator result does not establish the utility’s billing demand, replace the tariff, or guarantee realized savings.

Why This Schedule May Not Reduce Coincident Peak Demand

If 14:00–15:00 were the selected system peak hour, both schedules would have the same hourly average:

[ P_{\text{hour}}=\frac{32.5\text{ kWh}}{1\text{ hour}}=32.5\text{ kW} ]

The facility’s 15-minute maximum falls from 70 kW to 50 kW, but its contribution to that full hour remains unchanged. To reduce an hourly coincident-peak contribution, the facility must reduce energy within the selected hour—not merely redistribute it among that hour’s quarter-hour intervals.

Calculating a Five-Event Coincident-Peak Average

PJM’s January 14, 2026 guidance states that its 5CP intervals come from eligible non-holiday weekdays during June through September. The selected peaks use adjusted, unrestricted system load, which can differ from raw metered system load. PJM generally publishes the current year’s selected peaks around mid-October after receiving September data. Use the official PJM 5CP guidance and publication entry point to locate the applicable event file. Match its timestamps to your facility data before calculating a contribution. For example, PJM’s published summer 2025 file identifies its timestamps as hour-ending Eastern Prevailing Time; those are historical 2025 events, not 2026 events or predictions of future peaks. A transparent starting calculation for five measured event demands is:

[ P_{\text{five-event average}}=\frac{P_1+P_2+P_3+P_4+P_5}{5} ]

The NIH’s May 2026 technical bulletin uses this five-event averaging framework to explain PJM peak contributions. Distribution-company allocation and scaling methods still matter when determining the final customer obligation. %20and%20Capacity%20Charges%20-%20A%20Data%20Science%20Approach%20-%20May%202026%20TB_508.pdf)

Consider hypothetical event-hour demands of 80, 90, 85, 95, and 100 kW:

[ P_{\text{before}}=\frac{80+90+85+95+100}{5}=90\text{ kW} ]

If an approved operating plan reduces demand by 20 kW during all five selected hours:

[ P_{\text{after}}=\frac{60+70+65+75+80}{5}=70\text{ kW} ]

The unadjusted five-event average falls by 20 kW. If the same 20 kW reduction occurs during only one selected event, the average falls by:

[ \Delta P_{\text{average}}=\frac{20}{5}=4\text{ kW} ]

These results explain the event-averaging effect; they are not a final PLC assignment or a dollar-savings quote. Obtain the applicable allocation factors, rate units, delivery period, and supplier contract treatment before monetizing the reduction.

A coincident-peak alert identifies a candidate event, not a guaranteed final billing event. Official peak selection occurs after the relevant load data are assembled and adjusted. Evaluate savings against the final selected intervals and the account’s allocation rules.

Compare Scheduling Options Against the Charge You Pay

Operating approach	Facility maximum-demand objective	Coincident-peak objective	Main limitation
Run both tasks together	Can raise interval demand when loads overlap	Can raise the contribution if operation overlaps a selected system interval	Operational convenience may carry a demand cost
Stagger tasks within one hour	Can lower a 15-minute maximum, as in the worked example	Does not lower that hour’s average if hourly energy stays unchanged	Measurement windows determine the benefit
Move discretionary work outside a candidate system peak hour	Must still check for a new facility maximum	Can reduce the contribution if the candidate becomes a selected event	Forecast uncertainty and catch-up operation
Reduce task energy during the relevant interval	Can lower facility interval demand	Can lower demand within a selected system interval	Actual reduction must be verified without compromising required operations

Choose an approach around the charge’s measurement window and the equipment’s permitted operating constraints. Delayed charging, production rescheduling, or other discretionary-load changes should preserve required service, process limits, and manufacturer instructions. Billing analysis is not an electrical service-sizing calculation. It does not authorize smaller conductors, reduced overcurrent protection, or changes to required loads. Any equipment or control modification still requires applicable NEC provisions, local AHJ requirements, manufacturer instructions, and qualified site review.