How Demand Charges Change a Commercial Solar Decision

Commercial solar demand charges should be modeled from the facility’s actual load and tariff, not estimated as a percentage of annual solar production. Solar may lower a monthly measured peak when the array is producing during the determining interval. It may have little effect when the peak occurs in the evening, during weak production, or when a tariff ratchet, minimum, contract demand, or power-factor rule sets billing demand instead.

That distinction can change the right project scope. A business may still have a strong solar case based on energy used on site even if it assigns little demand value. Another facility may need a different layout, an operating change, or a separately evaluated storage case. The first decision is therefore not “How much will solar save on demand?” It is “Do the tariff and the data support any demand-charge value at all?”

Demand charges change the decision, not just the bill estimate

Commercial electric bills can contain several different cost determinants. Energy charges usually depend on the quantity of electricity consumed in kilowatt-hours. Demand charges depend on a measure of power in kilowatts or kilovolt-amperes. Fixed charges, riders, taxes, minimums, export credits, and other provisions may remain or change under their own rules.

Solar interacts differently with each component. One kilowatt-hour of solar used on site can displace one kilowatt-hour of grid import during that interval, subject to the account’s metering and tariff. But a demand charge may be set by one short interval, a defined on-peak window, a historical lookback, a contract amount, or a combination of those rules.

Use this decision table before assigning demand value to a proposal:

Question	Evidence required	What it can change
What schedule controls the account?	Current bill, tariff, riders, service contract, utility confirmation	Interval, time window, rate, minimum, ratchet, and eligibility
What set billed demand?	Interval data and a reconstruction of the billing calculation	Whether the visible bill line came from this month’s peak or another rule
Was solar producing then?	Time-aligned facility load and modeled PV production	Solar-only measured-demand effect
Does a floor override that effect?	Prior-month peaks, contract demand, minimums, and power-factor data	Difference between measured and billing demand
Is the effect durable?	Multiple months and downside cases	Whether value belongs in a decision model
Would another action address the peak better?	Operations, efficiency, charging, storage, and cost scopes modeled separately	Proceed, redesign, stage, or wait

This page owns that tariff-and-data test. The separate commercial solar ROI guide owns the complete cash flow, tax, financing, lifecycle cost, NPV, IRR, payback, and sensitivity analysis.

Start with kW, kWh, measured demand, and billing demand

Four terms prevent most demand-charge misunderstandings.

Kilowatts and kilowatt-hours answer different questions

Kilowatts (kW) describe a rate of power at a moment or averaged over an interval. A facility running a 100 kW load for one hour uses 100 kWh. The same facility running 200 kW for half an hour also uses 100 kWh. Energy is identical, but the peak power is not.

Kilowatt-hours (kWh) describe energy accumulated over time. Annual solar production is expressed in kWh, but annual kWh does not identify the facility’s highest monthly demand interval. Two buildings can consume the same monthly energy and have very different peaks because one runs steadily while the other starts large loads together.

When interval data report energy rather than average demand, convert it using the interval duration:

Average interval kW = interval kWh ÷ interval hours

A 15-minute interval is 0.25 hour, so 25 kWh in that interval represents 100 kW average demand. A 30-minute interval is 0.5 hour, so 50 kWh also represents 100 kW. Treating interval kWh as kW creates a fourfold or twofold error.

Measured demand is not always billing demand

Measured demand is the meter-derived demand under the schedule’s measurement rule, such as the maximum integrated 15-minute or 30-minute demand in a billing period.

Billing demand is the number the tariff says to charge. It may be the measured demand, but it may instead be the greatest of several values: the current peak, a percentage of an earlier peak, contract demand, a stated minimum, or a power-factor-adjusted value. Time-of-use schedules can calculate separate on-peak and off-peak demands.

That means a proposal can model a lower current-month net-load peak yet overstate bill value because a historical ratchet or contract floor still controls. Always reproduce the billed number before calculating a solar case.

Some schedules use kVA or adjust for power factor

Solar proposals commonly discuss real power in kW, while some tariffs bill apparent power in kVA or adjust demand when average power factor falls below a threshold. Those quantities cannot be substituted casually.

The correct question is not whether panels “improve the demand charge.” It is how the current tariff defines demand, what the meter records, and whether the proposed system changes the applicable determinant. Electrical and tariff reviewers should resolve kW, kVA, reactive-power, and power-factor treatment explicitly.

Current South Carolina tariffs show why one rule does not fit every account

South Carolina does not have one commercial demand-charge formula. The provider, schedule, riders, meter, contract, and effective date all matter. Confirm the name on the bill and use the South Carolina utility directory to reach the applicable provider documents. A city name is not proof of utility service, and a utility name alone is not proof of rate schedule.

The following examples were checked on August 10, 2026. They illustrate material differences; they are not a substitute for the account’s current filed tariff.

Current example	Measurement basis	Rules that can override or reshape the current peak	Solar decision consequence
Dominion Rate 20	Maximum integrated 15-minute demand, possibly rolling, stated in kVA	Seasonal lookbacks, contract demand, and a 75 kVA floor	Midday net-load reduction may not equal billing-demand reduction
Dominion Rate 21	Separate on-peak and off-peak 15-minute kVA demand	Winter lookback, contract/minimum relationships, current on-peak windows	Late-day and winter morning windows can have weak solar coincidence
Duke Energy Carolinas LGS	Maximum 30-minute demand	70% lookback, 50% of contract demand, and 30 kW minimum	A historical peak or contract amount can remain binding
Santee Cooper GA-25	Maximum integrated 30-minute kW demand	30% of the greatest measured demand in the preceding 11 months	The current peak can fall without billing demand falling as far

Dominion Energy South Carolina uses several different business structures

Dominion’s official South Carolina rates and tariffs page lists general-service, time-of-use, medium, large, and experimental business schedules. The relevant schedule must come from the account, not from a generic commercial label.

For example, Rate 20 Medium General Service, effective for bills rendered on and after the first billing cycle of July 2026, defines billing demand as the greatest of the current integrated 15-minute kVA peak, different seasonal percentages of earlier peaks, contract demand, or 75 kVA.

Rate 21 General Service Time-of-Use Demand, also effective in July 2026, separates on-peak and off-peak demand. Its current on-peak demand hours include 5 p.m. to 9 p.m. year-round and an additional 6 a.m. to 9 a.m. period from October through April. A system can produce substantial annual energy yet have limited output during an evening or winter-morning determinant.

The current Rate 24 Large General Service Time-of-Use schedule uses its own on-peak/off-peak relationships, minimums, and power-factor treatment. It would be unsafe to transfer a Rate 20 assumption to a Rate 24 account.

Duke Energy Carolinas and Duke Energy Progress are separate utilities

Duke Energy Carolinas’ current Schedule LGS, effective March 1, 2026, defines monthly billing demand as the largest of the current maximum 30-minute demand, 70% of the maximum 30-minute demand during the previous 12 billing months including the current month, 50% of contract demand, or 30 kW. The schedule can also correct demand when average monthly power factor is below 85%.

Its Schedule SGS has different applicability thresholds and a 50% historical lookback. Duke Energy Progress uses a separate set of South Carolina schedules. The SC Office of Regulatory Staff tariff directory lists Duke Energy Carolinas and Duke Energy Progress separately for a reason. Confirm the utility entity before opening a tariff model.

Santee Cooper direct retail has its own current demand definitions

Santee Cooper’s current GA-25 Small General Service schedule, effective April 1, 2025, measures maximum integrated 30-minute kW demand and sets billing demand as the greater of the current measured demand or 30% of the greatest measured demand in the preceding 11 months. Its GB, GT, and GL schedules have their own applicability and billing structures.

Santee Cooper is considering changes for 2027 and 2028, but its 2026 commercial rate-study page identifies those as proposed changes. They are not the current 2026 tariff. A model should not apply a proposed balanced-demand method before it becomes effective for the account.

Santee Cooper also supplies wholesale power to other entities. Its direct-retail schedule does not automatically govern an electric-cooperative account. The SC Office of Regulatory Staff explains that cooperatives are primarily self-governing. Municipal and cooperative customers need their own provider’s current documents.

Why solar may lower one peak and leave another in place

Solar changes net load when it is producing behind the meter:

Net facility load = facility load − contemporaneous solar production

That relationship must be calculated at the tariff’s relevant interval. It cannot be reconstructed from monthly totals alone.

Imagine a hypothetical facility with a 300 kW baseline peak at 5:30 p.m. and a second 295 kW interval later in the evening. If solar output is 20 kW during the first interval and zero during the second, the solar-case values become 280 kW and 295 kW. The determining peak moves to the later interval. The array delivered useful energy, but it reduced the monthly maximum by only 5 kW in this simplified example—not by its midday output or nameplate rating.

Several conditions can change the determining interval:

• HVAC load can persist as solar production declines.
• A cloud event can temporarily reduce PV output during a facility peak.
• Motors, refrigeration, process equipment, elevators, pumps, or compressors can start together.
• EV charging can coincide with an existing operational peak.
• A shutdown and restart can produce a short high interval.
• Weekend or holiday operations may count if the tariff does not exclude them.
• A new tenant, shift, production line, or electrified load can move the peak.
• Multiple meters may set separate peaks even when the owner views the property as one campus.

Array orientation, inverter loading, and layout can change the production profile, but none produces a universal demand benefit. Test each design as a time series against the same facility load and tariff. Do not select a west-facing layout solely from a general statement about afternoon value; site constraints, production losses, the actual peak window, and other project economics still matter.

Validate interval data before believing a demand model

The U.S. Department of Energy’s utility-rate evaluation guidance says electric cost depends on both load magnitude and shape and the bill structure. DOE recommends reviewing bills, obtaining interval data when available, and understanding the site’s actual rate options. That sequence is essential here.

Collect at least 12 complete bills plus the best meter interval data available. Use more history when the tariff has a lookback, the business is seasonal, operations changed, peaks are erratic, or the facility is planning a meaningful new load. Twenty-four months can help separate a recurring pattern from one abnormal year, but more data do not fix the wrong meter or tariff.

Build a data manifest

For every file, record:

• legal account name, account number, service address, meter ID, and utility;
• rate schedule, riders, effective dates, and contract demand;
• file source, export date, and whether values are actual or estimated;
• interval length, units, timestamp convention, time zone, and daylight-saving treatment;
• billing-period start and end dates;
• missing, duplicate, negative, zero, reset, or obviously abnormal values;
• meter replacements, account changes, outages, vacancies, shutdowns, and temporary loads; and
• planned changes to occupancy, shifts, equipment, HVAC, fleet charging, or production.

Check interval timestamps and units

Determine whether each timestamp marks the start or end of an interval. A row labeled 2:00 may represent energy from 1:45 to 2:00 or from 2:00 to 2:15. A one-interval shift can move solar production into or out of the determining peak.

Check time zones and daylight-saving transitions. A fall transition can create a repeated clock hour; a spring transition can omit one. Do not delete or duplicate values merely to make a file look regular. Preserve the utility’s convention, document the transformation, and compare it with billing evidence.

Confirm whether the file reports kW, kWh, kVA, kvar, or another quantity. If it reports energy per interval, apply the correct duration conversion. If a tariff uses rolling demand, simple fixed-block data may not reproduce the utility maximum exactly; document that limitation and request the appropriate data.

Reconcile energy and demand to actual bills

Before adding solar, reconstruct the no-solar bill.

1. Sum interval energy inside each exact billing period and compare it with billed kWh.
2. Calculate maximum demand within the tariff’s applicable windows.
3. Apply lookbacks, contract demand, minimums, and power-factor or kVA rules.
4. Compare calculated measured and billing demand with the bill.
5. Reconcile known fixed charges, riders, fuel adjustments, taxes, and other components separately.

A material mismatch is a stop signal, not a reason to add a hidden calibration factor. Investigate missing intervals, estimated meter values, a wrong schedule, billing-calendar alignment, unit conversion, time-zone shifts, power factor, multiple meters, or a tariff provision the model omitted.

The U.S. Energy Information Administration reinforces this point. EIA says its published average electricity prices are revenue divided by sales and are not individual utility rates. Operating revenue can blend energy, demand, customer, fuel, tax, and other charges. Dividing a commercial bill by kWh creates a blended historical ratio; it does not encode the demand formula.

Model the solar-only demand effect interval by interval

After the baseline reconciles, create a compatible solar production series. It should use the proposed site, layout, DC and AC size, equipment, weather source, loss assumptions, and time convention. Annual production is a useful output, but demand analysis needs the timestamps.

NREL’s commercial PV demand-charge fact sheet explains that PV affects demand only when production coincides with the building peak. Its simplified method estimates a range for relatively simple monthly demand charges and warns that actual results vary with daylight, load, production, and tariff complexity. Current South Carolina schedules with time windows, lookbacks, contract demand, minimums, or kVA rules require a more detailed treatment.

NREL’s System Advisor Model commercial documentation supports flat or time-of-use energy charges, fixed or time-of-use demand charges, tiers, and fixed charges, and it requires electric-load data when demand charges or tiers are used. A sophisticated tool is not evidence by itself. The user must still encode the current tariff correctly, supply valid data, and reconcile results.

For each billing month:

1. Preserve the validated baseline interval load.
2. Align the modeled PV series to the same time convention and resolution.
3. Subtract contemporaneous PV from load to create the solar-case net-load series.
4. Identify solar-case measured demand in every applicable tariff window.
5. Reapply historical lookbacks, minimums, contract demand, kW/kVA, power-factor, and on/off-peak relationships.
6. Record the baseline and solar-case determining timestamps, facility load, solar production, measured demand, billing demand, and controlling rule.
7. Keep exports and energy credits separate from the demand calculation unless the tariff explicitly links them.

Do not apply one demand-reduction percentage to every month. Do not multiply array kW by a generic capacity credit and carry it through the full project life. Do not use the best month as the annual assumption.

Test the historical floor explicitly

Consider a hypothetical tariff with a 70% lookback. A prior peak of 280 kW creates a 196 kW floor. If solar lowers the current measured peak from 210 kW to 180 kW, billing demand in this simplified example remains 196 kW. The measured peak fell by 30 kW, but billing demand fell by only 14 kW.

That result does not mean the array failed. It means the tariff delays or limits the billed effect. The actual percentage, lookback period, seasons, and other floors must come from the controlling schedule.

Use downside cases rather than a single weather year

At minimum, test:

• weak PV output during the baseline determining interval;
• the next-highest facility interval becoming the new peak;
• a later-day or winter-morning peak;
• an operating schedule change;
• planned HVAC, process, tenant, or EV load;
• a shutdown/restart event;
• degradation and availability assumptions appropriate to the production model; and
• a future tariff case presented as sensitivity, not certainty.

The result should be a monthly range and a list of controlling conditions. Only then should the validated utility-value difference flow into the broader financial analysis.

Keep operations changes and battery storage as separate cases

Solar is not the only way to change a facility peak. An energy audit may identify simultaneous equipment starts, scheduling conflicts, inefficient HVAC control, unmanaged charging, or another operational cause. Some changes can lower load without changing the solar array; others may be unsafe, impractical, or disruptive. Facilities and operations staff must define what can move and what cannot.

Model at least these cases separately when relevant:

Case	What changes	Evidence needed
Baseline	Nothing	Reconciled bills, load, and tariff
Solar only	Contemporaneous grid import	Time-series PV model and current tariff
Operations or efficiency	Facility load shape	Engineering/operations plan, schedule, persistence, and cost
Solar plus operations	Both series	Interaction modeled without double counting
Storage alternate	Dispatchable charge/discharge	Controls, power, energy, reserve, losses, degradation, tariff, and installed scope

A battery may be modeled for demand management, but it does not automatically eliminate demand charges. Dispatch must anticipate or respond to the relevant peak while respecting available power and energy, state of charge, efficiency, degradation, operating reserve, charging constraints, and competing uses. A ratchet, contract amount, or minimum may remain.

The separate commercial battery cost and quote-scope guide owns equipment ratings, controls, electrical integration, code review, commissioning, warranty, lifecycle obligations, and price. Keep those costs and constraints visible rather than inserting an ideal battery trace into a solar chart.

Require a demand-charge evidence schedule in the proposal

A proposal should let another qualified reviewer reproduce the demand result. Ask the bidder to provide a demand-charge exhibit with:

• account, meter, utility, schedule, riders, contract demand, and source dates;
• interval-data period, resolution, units, time convention, and cleaning log;
• month-by-month energy and demand reconciliation to actual bills;
• baseline measured demand and billing demand, with the controlling rule;
• solar-case measured demand and billing demand under the same rule;
• determining dates, times, facility loads, and modeled PV output;
• treatment of lookbacks, minimums, contract demand, kVA, and power factor;
• production model, weather source, layout, equipment, and loss assumptions;
• downside cases and unresolved data or tariff questions; and
• separate solar-only, operations, and storage results where applicable.

Use the commercial solar proposal checklist to compare the rest of the technical, price, schedule, warranty, and acceptance scope. Demand evidence is one exhibit, not the entire project decision.

Red flags include:

• “solar reduces demand by X%” with no tariff or interval calculation;
• annual PV kWh multiplied by a blended bill rate;
• one demand interval assumed for every South Carolina utility;
• a city used to guess the provider;
• measured demand presented as billing demand without applying floors;
• a model that cannot reproduce actual bills before solar;
• clean-looking data with no gap, duplicate, or timestamp record;
• a battery trace with no power, energy, control, degradation, or cost constraints;
• the best month repeated through the year; or
• an unconditional demand-reduction or savings claim.

Decide whether to proceed, redesign, or wait

Proceed to detailed design when the current tariff is confirmed, the baseline reconciles, the solar-only effect is reproducible, downside cases are disclosed, and unresolved rules are immaterial or assigned to a named reviewer.

Proceed for energy value while assigning little or no demand value when solar fits the facility’s daytime consumption or other goals but the important peaks do not reliably coincide with production, or a floor dominates. A conservative demand assumption can be better than forcing a favorable result.

Redesign or stage the project when another layout, operating schedule, efficiency measure, charging plan, or future load case could materially change the peak. Model each alternative from the same baseline.

Study storage separately when the tariff and load pattern support a defined demand-management objective and the owner is prepared to evaluate the complete controls, electrical, code, commissioning, cost, and lifecycle scope.

Wait when the tariff is unknown, interval data do not reconcile, meter boundaries are unclear, the business expects a major load or ownership change, or a proposal relies on an unsupported percentage.

Sunburst’s commercial solar service supports load, tariff, site, and project-feasibility review for South Carolina facilities. If your team can provide complete bills, interval data, the current rate schedule and riders, operating hours, planned load changes, and any proposals, request a commercial assessment built around the account’s actual demand rules.

How Sunburst models demand for South Carolina businesses

Demand charges are where commercial solar proposals most often overstate the result, so our modeling starts from your rate schedule and your measured intervals rather than an annual kWh total. We identify the billing-demand definition that actually applies to your account, test the peaks against modeled production hour by hour, and report the peaks solar cannot reach — including the winter morning and post-sunset peaks common in South Carolina facilities.

Where the demand component is the real problem, the honest answer is often a storage or storage-ready scope rather than more modules, and we will price both. Our commercial solar service covers the feasibility work, the utility interconnection, construction and commissioning, and our battery storage team handles the dispatch side.

Read next: what utility data an assessment needs and how to audit an ROI model. To have your tariff read line by line, request a commercial assessment — or see commercial solar by city for where we work.

Commercial solar demand-charge FAQ

Does commercial solar always reduce demand charges?

No. Solar changes demand only when it lowers net load during the interval and time window that controls billing demand. A later peak, weak production, historical ratchet, contract demand, minimum, kVA basis, or power-factor adjustment can limit or eliminate the billed effect in a given month.

How much interval data should a business collect?

Start with at least 12 complete bills and the best interval data available. Use more history when the tariff has a lookback, operations are seasonal, peaks vary, or the facility has changed. The data must match the correct meter, units, timestamps, and billing periods.

Can monthly bills alone prove the demand effect?

Bills can identify billed kWh, demand, schedule, and charges, but monthly totals usually cannot show which interval set the peak or how PV would have changed it. Obtain interval data when demand value is material. If it is unavailable, disclose the limitation and use conservative ranges rather than false precision.

Is a 15-minute peak standard in South Carolina?

No. Current examples include integrated 15-minute and 30-minute measurements, and schedules can use rolling intervals, on-peak windows, or other rules. Read the account’s tariff and meter documentation.

Why can billing demand stay high after measured demand falls?

A tariff can bill the greater of current measured demand and a historical percentage, contract amount, or minimum. That lookback or floor may remain binding until its defined period changes. Reconstruct both measured and billing demand month by month.

Can EIA’s commercial electricity price be used in the calculation?

Not as the account tariff. EIA’s average price is a revenue-per-sales statistic that can include several types of charges. It is useful for broad context, not for reproducing an individual facility’s demand, energy, rider, or fixed-charge calculation.

Will a battery eliminate commercial demand charges?

Not automatically. A storage case depends on the tariff determinant, forecast and control logic, power and usable energy, state of charge, losses, reserve, degradation, operating constraints, and remaining floors. Model it separately and price the complete installed and lifecycle scope.

Should a business change rate schedules before installing solar?

Do not assume a change is available or beneficial. Ask the utility or account representative which schedules the facility qualifies for, then compare complete bills under current and proposed load cases. Consider contract terms, riders, minimums, and future operations—not only the demand rate.

Sources and methodology

This guide was researched on August 10, 2026. Current tariff examples were checked against the linked utility schedules and the SC Office of Regulatory Staff directory. The analysis method follows DOE’s bill/rate/interval-data sequence, NREL’s requirement to match PV production with facility peaks, SAM’s interval-load and rate-model structure, and EIA’s distinction between average revenue per kWh and an individual tariff.

• Dominion Energy South Carolina rates and tariffs
• Duke Energy Carolinas Schedule LGS
• Santee Cooper GA-25 Small General Service
• SC Office of Regulatory Staff tariff directory
• DOE FEMP utility-rate evaluation guidance
• NREL commercial PV demand-charge guide
• NREL System Advisor Model commercial documentation
• EIA explanation of average electricity prices and utility rates