Commercial demand charges are becoming the deciding variable in U.S. storage economics — not the federal tax credit. Commercial power prices rose again in 2026, regulators approved a new generation of large-load tariffs that floor billed demand at a percentage of contracted capacity, and several utilities began paying customers for verified on-site curtailment. For 2027 projects the rule is blunt: size the battery to the tariff sheet, not to the load curve.
That shift matters to anyone specifying, importing, or distributing commercial and industrial (C&I) storage in the United States. A system that looks correctly sized on an interval-data chart can lose half its projected savings if the utility bills a demand floor, applies a ratchet, or excludes the site from load-control riders.
Why 2027 Is a Tariff Year, Not a Tax-Credit Year
Retail prices keep climbing. The U.S. average commercial price reached 14.47 cents per kWh in August 2026, up 4.3% year over year, while the industrial average hit 9.64 cents, up 7.3%, according to the EIA Electric Power Monthly released October 23, 2026. Those are blended energy prices; demand-related components sit on top.
Load growth is now a planning constraint. EIA expects U.S. electricity consumption to grow another 3.1% in 2027, driven largely by data centers and other commercial loads, and residential customers are paying nearly 5% more for power this year (EIA, via Utility Dive, November 9, 2026). Supply cannot close the gap quickly: Berkeley Lab’s GridTracker work shows the median project built in 2025 spent more than five years between interconnection request and commercial operation, and only 13% of capacity that requested interconnection between 2000 and 2020 had come online by the end of 2025.
Merchant revenue is not the answer either. In ERCOT the average battery earned 84% less in 2025 than at its 2023 peak as more storage chased the same services (Modo Energy, cited in the same Utility Dive analysis). With roughly 52 GW of utility-scale battery storage operating in the U.S. as of mid-2026 — including 8.3 GW added in the first half of 2026 (EIA Today in Energy, August 7, 2026) — wholesale spreads are no longer a reliable foundation for a business case. What remains is the retail bill, and the retail bill is increasingly shaped by tariff design rather than nameplate capacity.
What Commercial Demand Charges Actually Bill
Commercial demand charges are $/kW fees applied to the highest average power draw recorded during a billing interval — typically 15 or 30 minutes — not to total energy consumed. Two facilities with identical monthly kWh can pay wildly different amounts depending on how spiky their load is. Four mechanics compound:
- Billing demand definition. Some tariffs bill the highest interval of the month; others bill a rolling maximum over 11 or 12 months.
- Ratchet provisions. A ratchet holds billing demand at a historical percentage of a prior peak, so one bad month can price the next eleven.
- Component separation. Generation, transmission, and distribution demand are often billed separately, at different rates and intervals.
- Time differentiation. On-peak demand may carry a multiple of the off-peak rate, which changes when the battery should discharge.
For a low-load-factor operation — a warehouse with a short shipping window, a plastics plant with batch process loads, a cold-storage site with compressor staging — commercial demand charges can dominate the bill. The exact share depends entirely on the tariff, which is why a savings model built from someone else’s utility bill is close to worthless.
The Structural Change: Minimum Billing Demand
The most consequential 2026 policy development for C&I storage is not a subsidy. It is a billing floor. Berkeley Lab and The Brattle Group reviewed 55 large-load tariffs for their August 2026 update, Electricity Rate Designs for Large Loads. Minimum billing demand — a floor on the kW used for billing regardless of actual consumption — appears in most of them, at a median of 80% of contracted demand. Individual tariffs are stricter:
- Dominion Energy Schedule GS-5 sets a multicomponent floor of 85% of contracted demand for distribution, 85% for transmission, and 60% for generation.
- Indiana Michigan Power requires billing demand no lower than 90% of the greater of contract capacity or the prior 11-month maximum.
- Colorado Springs Utilities bills 100% of contracted demand for the first 10 years, dropping to 68% of the trailing 12-month maximum afterward.
This is the single most important number in a storage proposal. Suppose a facility contracts 1,200 kW and the battery shaves its monthly peak from 1,200 kW to 700 kW. Under a straight demand charge the saving is 500 kW. Under an 80% floor, billed demand cannot fall below 960 kW, so the addressable reduction is 240 kW — less than half of what the load curve promised. A battery captures full value only by reducing contracted capacity with utility approval, or by operating under a rider that pays separately for curtailment. Both require equipment a utility will accept as firm, controllable capacity. Commercial demand charges are billed monthly, but the floor is set once — at contract signing.
A worked example: how the floor halves the savings
The following is an illustrative model, not a quote. Assumptions: 1,200 kW monthly peak, 400 kW average load, a $14/kW-month demand charge, and an 80% minimum billing demand on 1,200 kW contracted.
- No billing floor. A 500 kW battery shaves peak to 700 kW. Savings: 500 kW × $14 × 12 = $84,000 a year.
- 80% floor. Billed demand cannot fall below 960 kW, so the same battery yields 240 kW of billable reduction: 240 kW × $14 × 12 = $40,320.
- Floor plus arbitrage. Add 1,000 kWh of daily throughput at a 7-cent on-peak/off-peak spread over 300 days — about $21,000 a year — and the gap narrows but does not close.
The second number belongs in an investment committee memo. The first is what too often appears in a vendor proposal, and the difference is one line of tariff language.
How the 2026 Tariff Playbook Treats On-Site Storage
Tariffs are splitting into three groups, and each group decides whether commercial demand charges can be managed at all. Those that allow behind-the-meter resources: AEP Ohio’s Schedule DCT (25 MW minimum) lets customers interconnect on-site resources and offset contract capacity, but requires equipment able to instantaneously curtail load equal to or greater than on-site generation output, or the utility may suspend service. Evergy Missouri’s Large Load Power Service offers an optional Demand Response & Local Generation Rider paying bill credits for verified curtailment.
Those that restrict them: Florida Power & Light’s Large Load Contract Service requires an 85% minimum load factor and bars customers from load-control riders. East Kentucky Power Cooperative pairs interruptible service with a per-kW credit that scales with annual interruption hours — and a penalty of five times the firm demand charge for non-performance.
Those that treat interruption as a product: Colorado Springs Utilities’ interruptible rate allows up to 100 hours of full interruption per year, with energy and demand credits and no non-performance penalty. Load-factor eligibility adds another filter, with thresholds from 60% to 85% and one tariff at 92% for nine of the prior 12 months.
How Tariff Elements Change Commercial Demand Charges and Payback
The table summarizes design elements documented in the 2026 Berkeley Lab review and translates each into its effect on commercial demand charges and project returns.
| Tariff element | What the 2026 evidence shows | Effect on a storage business case |
|---|---|---|
| Monthly demand charge | Core component of nearly every large-load tariff; often split by generation, transmission, and distribution | Primary savings driver; model each component, not one blended $/kW |
| Minimum billing demand | Median 80% of contracted demand; examples at 85%, 90%, 100% | Caps addressable kW; shaving below the floor produces zero savings |
| Demand ratchet | Rolling 11–12 month lookbacks in several tariffs | Raises the value of consistent dispatch; one missed peak erases a year |
| Minimum contract duration | Median rose from 5 years (pre-2025 tariffs) to 12 years (proposed since January 2025) | Locks in contracted MW; on-site capacity that trims it gains value |
| Load factor threshold | 60%–85% typical; one tariff at 92% | Rewards load flattening over pure arbitrage; affects eligibility |
| BTM resource clause | Allowed with mandatory instantaneous curtailment (AEP Ohio DCT) | Requires certified controls, telemetry, response-time evidence |
| Demand response rider | Bill credits for verified curtailment (Evergy Missouri) | Adds a revenue line independent of the demand charge |
| Interruptible credit | Per-kW credits scaled by interruption hours; penalties up to 5x firm demand | Higher ceiling, real performance risk; needs automated dispatch |
| Direct cost assignment | Upfront contributions for interconnection and network upgrades | Raises capital cost; storage can reduce upgrade scope if sized early |
| Collateral requirement | Dominion GS-5 requires $1.5 million per MW of contracted capacity | Contracts large enough to trigger collateral favor lower contracted MW |
Sizing Storage to a Tariff: Five Steps
- Collect real inputs. Twelve months of 15-minute interval data, the full tariff sheet with riders, and the customer’s contract demand — not a bill summary.
- Reconstruct billed demand. Identify which interval drives the charge, whether transmission and distribution use different windows, and whether a ratchet applies.
- Calculate the floor. Multiply minimum billing demand by contracted capacity. The gap between peak and floor is your addressable kW; that gap is the entire pool of commercial demand charges the system can remove, and everything above it is unbillable.
- Size power, then energy. Power must cover the addressable kW plus control margin. Energy must cover the peak window duration and daily cycle count implied by the dispatch strategy. Check continuous C-rate and usable depth of discharge at the warranty boundary, not the datasheet maximum.
- Design for acceptance. Where the tariff ties on-site resources to curtailment capability, the proposal needs a stated response time, telemetry protocol, and test procedure the utility will sign off on. This is where commercial and industrial energy storage systems with native demand-response inputs separate from retrofit kits.
The pre-quote tariff audit: ten checks
Each item below changes either the size of commercial demand charges or the site’s eligibility to reduce them.
- ☐ Confirm the applicable rate schedule and every rider by name and effective date.
- ☐ Record the minimum billing demand percentage and how it is calculated.
- ☐ Identify ratchet lookback length and whether it resets annually.
- ☐ Separate generation, transmission, and distribution demand charges.
- ☐ Confirm the demand interval (15 vs 30 minutes) and meter recording method.
- ☐ Check load-factor eligibility if the site sits near a tariff boundary.
- ☐ Read the behind-the-meter clause: required curtailment capability, and who tests it.
- ☐ Verify whether demand-response or load-control riders are available, restricted, or prohibited.
- ☐ Capture minimum contract duration and any collateral scaled to contracted MW.
- ☐ Model savings against the floor, ratchet, and dispatch window — then stress-test at 70%.
Where the Policy Is Heading in 2027
Tariff templates are standardizing. The Pennsylvania Public Utility Commission published a final order in April 2026 on a large-load tariff template, recommending that load-serving entities assess contributions in aid of construction that recover all interconnection costs, including network upgrades (Berkeley Lab, 2026). Standardization makes tariff risk easier to price — and easier for competitors to replicate.
Contract terms are lengthening. The median minimum contract duration rose from 5 years to 12 years for tariffs proposed since January 2025, which makes every kW of commercial demand charges more expensive to carry over the contract life, and increases the value of on-site capacity that can credibly reduce contracted MW.
Export compensation keeps weakening. pv magazine USA reported on November 3, 2026 that storage-first models and VPP integration now define the post-ITC solar market, with declining export rates pushing projects toward self-consumption and dispatchable value (pv magazine USA, November 3, 2026). Demand-charge management is the most direct form of dispatchable value a C&I site controls. Reliability rules are tightening in parallel: FERC Order No. 901 directs NERC to develop standards for inverter-based resources, including batteries, through 2030, so systems with field-updatable controls will age better than fixed-firmware designs.
Tariff rules change frequently and vary by jurisdiction. This is technical and commercial analysis, not tax, legal, or regulatory advice; confirm current terms with the serving utility before committing capital.
What this means for distributors and OEM programs
If bills are decided by tariff mechanics, hardware differentiation has to move there. Specify continuous rather than peak capability: demand-charge duty is daily cycling, often twice a day, so continuous kW, overload duration, partial-load efficiency, and high-cycle warranty terms matter more than a headline rating. Ship controls that can prove performance — Modbus TCP, SunSpec-compatible registers, dry-contact demand-response inputs, timestamped event logs — because tariffs that allow behind-the-meter resources increasingly require demonstrated curtailment.
Distributors who can quantify commercial demand charges — floor and ratchet applied — close faster than those who hand over a datasheet. Our wholesale and distribution program and OEM/ODM manufacturing services are built around that: configurable commercial energy storage systems, LiFePO4 battery modules, and hybrid inverters with the documentation package utilities ask for. For the timing question on interconnection and program eligibility, see our behind-the-meter storage policy calendar.
Frequently Asked Questions
Five questions come up most often when a C&I buyer compares a storage proposal against a real tariff sheet.
What is a demand charge on a commercial electric bill?
A demand charge is a monthly $/kW fee based on the highest average power draw recorded during a billing interval, usually 15 or 30 minutes. It recovers the utility’s cost of being ready to serve your peak, which is why it is billed separately from kWh consumption and why two sites with identical energy use can pay very different amounts.
How much can a battery actually reduce commercial demand charges?
It depends on the gap between your monthly peak and the tariff’s billing floor. If billed demand cannot fall below 80% of contracted capacity, only the portion above that floor is addressable. A site peaking at 1,200 kW with a 960 kW floor has 240 kW of billable reduction available, regardless of battery size — which is why sizing starts with the tariff, not the load curve.
Why does minimum billing demand cap storage savings?
It sets a floor on the kW you are billed for even when actual demand is lower. Berkeley Lab’s 2026 review found a median floor of 80% of contracted demand across 55 large-load tariffs, with individual tariffs at 85%, 90%, and 100%. Shaving peak below the floor reduces consumption but not the bill unless contracted capacity is formally reduced.
Do large-load tariffs allow behind-the-meter batteries?
Some do, with conditions. AEP Ohio’s Schedule DCT permits on-site resources to offset contract capacity but requires equipment able to instantaneously curtail load equal to or greater than on-site generation output, or the utility may suspend service. Evergy Missouri offers a rider that pays for verified curtailment. Others, including Florida Power & Light’s Large Load Contract Service, bar customers from load-control riders entirely.
How do I verify a storage proposal’s demand-charge savings?
Ask for four things: the tariff schedule and riders used, the billing-demand definition including any ratchet or floor, the interval data behind the peak, and the dispatch logic with its response time. Then re-run the model with the floor applied and stress-test savings at 70%. If the proposal cannot produce those inputs, the savings figure is an assumption, not a calculation.
The Bottom Line
In 2027 the difference between a storage project that pays back and one that disappoints will usually come down to tariff mechanics — a billing floor, a ratchet, a load-factor threshold, a curtailment clause. Commercial demand charges are where those mechanics appear on the invoice, and they deserve the same modeling rigor as equipment cost. Start with the schedule, then size the system.
If you need hardware that can pass a utility’s curtailment test, and a partner organized around tariff-first proposals, request a quote and tell us which tariff you are working against.
Sources
- Mims Frick, N., Cappers, P., Kavlak, G., Lam, L., Hledik, R., Grove, J., and Rotella, S. Electricity Rate Designs for Large Loads: Evolving Practices and Opportunities — 2026 Update. Berkeley Lab and The Brattle Group, August 2026. eta-publications.lbl.gov
- U.S. Energy Information Administration. Electric Power Monthly, data for August 2026, released October 23, 2026. eia.gov/electricity/monthly
- U.S. Energy Information Administration. Short-Term Energy Outlook, October 2026. eia.gov/outlooks/steo
- U.S. Energy Information Administration. “Battery storage capacity averaged 70% growth over the last three years.” Today in Energy, August 7, 2026. eia.gov/todayinenergy
- Utility Dive. “The next wave of energy storage capacity is already on the grid.” November 9, 2026. utilitydive.com
- pv magazine USA. “Storage-first models and VPP integration define post-ITC solar market.” November 3, 2026. pv-magazine-usa.com