One 15‑minute spike can cost more than every other kilowatt‑hour you buy
Pull out last month’s electric invoice. If your building is on a commercial or industrial tariff, two dollar values dominate: one labeled something like Energy, often charged in cents per kilowatt‑hour, and a second called Demand or Maximum kW, charged in dollars per kilowatt. Study the math and you will discover a painful fact: many customers pay more for their worst fifteen‑minute power spike than for all the energy they consume all month. In New York City a single 900‑kilowatt HVAC surge on a July afternoon can trigger a $50,000 demand charge—even if average load cruises at 300 kilowatts the rest of the billing cycle. Demand charges are the silent budget killer because they penalize instantaneous power, not monthly energy. They are also the piece batteries solve better than any other technology. This deep dive explains what demand charges are, why utilities bill them, and how two‑ to four‑hour lithium‑ion batteries shave 10–30 percent off total electric cost while giving your property backup power, ESG credits, and higher NOI.
What exactly is a demand charge?
Utilities must size generation, transmission, and distribution equipment to handle the single worst fifteen‑ or thirty‑minute period of load in an entire month. That capacity sits idle most of the time, but investors and regulators still expect a return on the capital. Rather than bury those fixed costs inside the energy rate, utilities create a separate line item: the demand charge. Think of it as a toll for reserving a private lane on the freeway even if you only use it once. The toll is calculated by recording your building’s highest average kilowatts over one interval—fifteen minutes for most large‑customer tariffs, thirty minutes in some Midwestern cooperatives—and multiplying that value by a dollar‑per‑kilowatt price. The price can be low teens in inexpensive regions, high twenties on many West Coast and Mid‑Atlantic tariffs, and north of sixty dollars on the Hawaiian and New York City grids.
Utilities justify demand charges with two arguments:
• Cost recovery for fixed infrastructure. Transformers, substations, wire, and spinning reserves must be big enough for your worst spike, so you should pay your fair share of that capability.
• Price signal to flatten peaks. If customers cap their own peaks, the utility can delay new infrastructure, keeping rates lower for everyone. Demand charges therefore incentivize self‑smoothing.
Unfortunately for building owners, that “incentive” shows up as a painful bill line because most facilities have unpredictable spikes—chiller restarts after a brief outage or a simultaneous elevator bank run at shift change. Those events are hard to schedule manually. The rest of this article shows how batteries automate the fix.
Why demand charges now rival or exceed energy charges
Demand charges have existed for decades, but they used to be a manageable slice of the invoice. Three forces have pushed them to center stage:
Grid modernization cost. Utilities are investing billions in wildfire hardening, EV‑charger upgrades, and renewable interconnections. Regulators prefer to allocate those fixed costs to customers who use the most capacity, so dollar‑per‑kilowatt rates climb faster than cents‑per‑kilowatt‑hour energy rates.
Energy efficiency success paradox. LEDs, VFD drives, and better HVAC controls shrink monthly energy consumption, but they rarely reduce the single highest fifteen‑minute power draw. As the energy numerator drops, demand becomes a larger share of the bill pie.
Time‑of‑use demand tiers. New tariffs such as PG&E’s E‑20 and Duke’s OPT G add separate demand charges for different time windows—on‑peak, part‑peak, off‑peak. A building can pay three demand charges in the same month if it spikes in each window.
The result is a bill where demand routinely exceeds forty percent of cost and sometimes hits sixty percent. That is why battery peak shaving returns double‑digit internal rates of return in many territories even before energy arbitrage is considered.
Hidden billing rules that further amplify demand charges
Not all kilowatts are billed equally. Four common clauses amplify the pain:
• Ratchet clauses. Many tariffs lock billing demand at fifty to seventy‑five percent of last year’s worst peak. One bad summer day sets a twelve‑month floor.
• Coincident peak allocation. Regional grid operators such as ERCOT and PJM charge large customers based on their share of the grid’s single highest hour each season. Miss that hour and you still pay.
• Contract demand escalation. Some utilities force you to nominate an expected peak. Exceed it a few times and the baseline ratchets up permanently.
• Standby reservation fees. If you self‑generate, the utility reserves capacity for when your generator fails and bills you a monthly standby demand fee.
Why these rules exist
Ratchets smooth revenue: a utility’s debt payments do not fall just because a factory idles production in November. Coincident charges prevent free‑riding; the grid must be sized for system‑wide peaks, not individual peaks. Contract demand protects transformer life—overloading equipment even once accelerates wear. Standby fees ensure a cost recovery path for infrastructure kept on call.
Why they surprise budget owners
Fine print hides them. Ratchet language may be one sentence on page six of a tariff. Coincident charges appear a year after the actual peak, long after managers forget the event. If you lease space, the reimbursement formula might pass demand penalties straight to the tenant, souring relationships. Most financial models also assume a linear relationship between energy savings and bill savings, ignoring nonlinear adders. Failing to account for hidden rules leads to under‑sized solutions and missed savings targets.
Traditional fixes—and their built‑in ceilings
Before batteries became cost‑effective, engineers experimented with four primary strategies. Each helps, none solves the problem entirely.
• Operational load shifting. Move energy‑intensive tasks like floor buffing or ice making to nighttime. Downside: dependent on human discipline; one forgotten setting resets the peak.
• Soft‑start equipment upgrades. Replace across‑the‑line motor starters with VFDs. Downside: only affects that motor; peaks often involve multiple loads.
• Rate‑schedule changes. Shift from a pure demand tariff to a time‑of‑use tariff. Downside: most TOU plans still have demand windows, and energy adders can eat the benefit.
• Power‑factor correction. Raise power factor above 0.95 to avoid kVA penalties. Downside: does nothing for real kW peaks.
Most sites adopt these fixes first; when residual peaks keep demand charges stubbornly high, they turn to batteries.
How batteries flatten peaks in practice
A battery energy‑storage system, typically lithium‑iron‑phosphate for safety and life span, connects to the building’s main switchboard or a large subpanel. A controller monitors real‑time load via current transformers. As load climbs toward a programmable threshold—say 600 kilowatts—the controller sends an inverter command to discharge. The battery injects power for the rest of the fifteen‑minute interval so the utility meter never registers more than 600 kilowatts. At night the battery recharges on off‑peak energy costing a fraction of on‑peak rates. Because only a few discharges are needed per day, two‑hour duration is sometimes enough, though four‑hour systems capture both peak and part‑peak windows and often double savings.
Sizing logic in a single paragraph
You size power (kilowatts) to match the gap between your average and peak demand and size energy (kilowatt‑hours) to cover how long that gap lasts. If your normal afternoon load is 400 kW and the spike tops at 800 kW for two hours, a 400 kW / 800 kWh battery caps the meter at 400 kW. Add a thirty‑percent power margin for inverter clipping, round‑trip losses, and cell degradation, and you land on a commercial 500 kW / 1 MWh module—the sweet‑spot product most containerized vendors stock.
Typical savings range
In California, a well‑tuned battery cuts demand charges by $200 to $300 per kilowatt‑year. A 500‑kW system therefore saves $100 k to $150 k annually. After factoring in round‑trip losses and minor increases to energy cost, net savings land between $85 k and $140 k. In higher‑priced Hawai‘i the same battery clears $250 k per year.
The resilience bonus
Although batteries usually export power during peaks, they can instantly flip to island mode during an outage, powering critical loads without the five‑second blackout a diesel generator introduces. Insurance underwriters and data‑center tenants increasingly value that ride‑through capability, sometimes lowering business‑interruption premiums.
Real‑world narrative: cold‑storage warehouse in Southern California
Cold‑storage facilities are perfect demand‑charge candidates: compressors cycle hard during afternoon heat, spiking load by hundreds of kilowatts. One 200 k‑square‑foot warehouse near the Port of Los Angeles averaged 700 kW but spiked to 1,100 kW on hot days. Demand rates under SCE’s TOU 8 tariff were $30 per kilowatt. Monthly demand charges ran $33,000. After auditing interval data, Vista installed a 500 kW / 2 MWh battery. The controller set a 750 kW cap. When two compressors kicked on simultaneously, the battery output 350 kW, keeping the meter under the cap. Over thirty days the highest recorded utility demand was 748 kW, dropping the demand line to $22,400. Energy cost inched up $1,200 from charging losses and nighttime kWh buys, netting a $9,400 monthly reduction. An ESA split eighty percent of savings to Vista, twenty percent to the owner in year one. By year five the split flips, delivering nearly $120 k additional NOI annually.
Why did the project pencil? Demand was forty‑four percent of the bill, peak duration matched a four‑hour battery, and the warehouse’s rate class offered an optional “R” schedule that lowered off‑peak energy prices. Traditional fixes had already been tried: setpoint tinkering ruined product quality, and compressor VFDs had long paybacks. Battery peak shaving provided the only remaining lever.
Non‑obvious battery value streams that sweeten the economics
Energy arbitrage. In time‑of‑use tariffs with 10‑cent spreads, charging on super‑off‑peak and discharging on peak adds another five to ten percent savings over demand reduction.
Demand‑response participation. Utilities pay $75 to $150 per enrolled kilowatt per year for fasts‑hedding assets. Batteries respond instantly and don’t upset tenants like HVAC curtailment, turning the battery from a cost‑cutter into a revenue asset.
Capacity and ancillary markets. ISO‑New England and PJM let batteries bid frequency regulation. Even partial participation can yield $20–$40 per kilowatt‑year, offsetting degradation costs.
ESG scoring. GRESB, ENERGY STAR, and the pending SEC climate‑disclosure framework award points for peak‑reduction technologies that lessen grid stress, boosting investor perception and future capital access.
Insurance premium reductions. Insurers are starting to credit battery‑backed resilience when pricing business‑interruption coverage, shaving OPEX beyond the electric bill.
Ownership models and why Energy‑as‑a‑Service often wins
Energy‑as‑a‑Service (ESA): Provider funds the system, maintains it, and takes performance risk. Payment is a share of realized savings. Accounting treatment is usually an operating expense, keeping debt ratios unchanged.
Outright purchase: Building pays full cap‑ex, claims any tax credit, and keeps all savings. Requires capital reserves for inverter and cell replacement around year twelve.
CPACE financing: Long‑term tax assessment pays for the battery at fixed interest. Non‑recourse but lien travels with the property, potentially lowering sale price.
Equipment lease: Fast approvals, higher interest, counts as on‑balance‑sheet debt.
Find out what works for you
CFOs value IRR; asset managers value NOI lift with no balance‑sheet impact. ESA satisfies the latter group: no capital outlay, immediate positive cash flow. IRRs may appear lower because savings are shared, but unlevered returns on invested capital remain infinite—there is no investment. When federal tax credits are certain and the owner has appetite, outright purchase often yields the highest IRR, but the owner bears technical risk. CPACE suits mission‑critical owner‑occupiers who want low, fixed financing and plan to hold the asset at least ten years. Leases bridge temporary liquidity gaps but are usually the least attractive over a twenty‑year horizon.
Step‑by‑step evaluation process you can complete this week
- Export twelve months of fifteen‑minute demand data and verify the meter multiplier.
- Identify the worst peak and calculate its dollar impact by multiplying by the demand rate.
- Check the tariff for ratchet or coincident‑peak language and estimate true savings potential.
- Size a battery using the rule: kilowatt power equals peak minus average, kilowatt‑hours equal that delta times spike duration plus thirty percent.
- Request an ESA term sheet and a capital purchase quote to compare NPV and IRR under different discount rates.
- Model sensitivity: escalate demand rates four percent annually, energy two percent, and strip ten percent savings for degradation.
- Present findings to finance with a one‑page risk matrix: technology, tariff change, counterparty risk.
- If approved, file a non‑export interconnection application; many utilities turn these around in four to six weeks.
- Schedule commissioning at least ten days before the Certificate‑of‑Occupancy or loan draw deadline.
Why this nine‑bullet checklist works
Each bullet removes a common blocker: data unavailability, unrealistic savings assumptions, financing confusion, or interconnection delays. You avoid months of back‑and‑forth by assembling everything once. Lenders love seeing sensitivity modeling; it shows professional risk management. Interconnection, often the long pole, moves into the critical path immediately, preventing schedule slippage.
Future‑proofing: what if demand charges disappear?
Some skeptics argue that as distributed generation proliferates, regulators might eliminate demand charges. Three reasons make that unlikely:
Grid economics rely on fixed cost recovery. Even if photovoltaic penetration grows, substations and wires still need upgrades for electrification of transport and heat. Regulators must collect fixed costs somehow; demand charges are the simplest mechanism.
Tariff reform trends upward, not downward. The latest rate cases in California, New York, and Massachusetts all increase demand components. Where demand charges have been reduced, utilities replaced them with higher customer charges or capacity riders, preserving the revenue.
Performance risk remains on the provider under ESA. If demand charges vanished tomorrow, ESA payment formulas would adjust. Customers would pay based on volumetric energy arbitrage and ancillary revenue instead, keeping risk aligned to the provider. So, sizing a battery for current demand remains a sound hedge even in a changing regulatory landscape.
Key Takeaway: Flatten one spike, boost NOI for decades
Demand charges are the electric bill’s silent heavyweight—often larger than every kilowatt‑hour combined and governed by fine‑print rules that lock pain in for months or years. Traditional fixes chip away but rarely solve the root cause: fleeting, high‑power spikes. A right‑sized two‑ to four‑hour lithium‑ion battery clips those spikes, lowers demand by double‑digit percentages, and unlocks additional revenue streams such as energy arbitrage and demand response. Whether financed through Energy‑as‑a‑Service, CPACE, or outright purchase, the project typically returns capital in five to eight years or produces positive cash flow from day one under ESA. Combine the battery with tariff optimization, power‑factor correction, and smart load shifting, and total bill reduction often exceeds thirty percent—boosting property NOI and asset value far beyond the utility line. The first step is simple: pull your interval data, find the worst fifteen minutes, and let the numbers prove how much you stand to gain.
We can reduce your demand charges for free
Vista Power offers $0 upfront, off-balance sheet energy storage and solar hardware. Our payments are structured as a percentage of actual savings - meaning we will do our best to save you as much money on energy as possible.
As part of our process, we take a look at your energy bill and all available tariffs to you. We then design the energy storage or energy storage + solar system to save you as much money as possible and recommend a rate switch for you if applicable. We can look either at an individual property or an entire portfolio, and select the properties that need the most attention.
All we need are:
- Twelve months-to-date of your energy bills
- Twelve months-to-date of your interval energy use data from your utility's portal
- Optional: How much space you have available (a simple site map can suffice)
Reach out here and let's get started right away!