Battery Storage for Commercial Demand-Charge Management

A facility manager opens the monthly electric bill and finds a line near the top for roughly $3,800 labeled demand charge. The energy charge below it, the part that pays for every kilowatt-hour the building actually consumed, is smaller than that single line. Nobody left anything running all month. The building ran the way it always runs. One afternoon, for about fifteen minutes, several large loads happened to turn on at once, and that brief overlap set a number the utility then billed against for the whole period.

That is the problem this post is about: the short peaks that drive demand charges and what a battery can and cannot do about them. The savings are real for the right site. They are also entirely a function of your rate structure, so we will keep every dollar figure directional and spend the time on the mechanism.

The peak that set your whole demand charge

Commercial and industrial rate schedules usually bill on two separate meters. One tracks energy, the total kilowatt-hours you draw over the month. The other tracks demand, the single highest rate of draw the utility recorded during any short interval, usually fifteen minutes (EIA). The energy meter is a story about the whole month. The demand meter is a story about your worst fifteen minutes.

Picture a light industrial site. Compressors, an HVAC plant, a couple of production lines, and office load. On a normal interval the building might pull 180 kW. Then the air conditioning kicks hard on a hot afternoon at the same moment a compressor cycles and a line starts up, and for one interval the draw touches 320 kW. That 320 kW is your demand for the month. Multiply it by a demand rate that can run anywhere from a few dollars to well over twenty dollars per kW depending on your utility and tariff, and you get a charge that has almost nothing to do with how much electricity you used and almost everything to do with when you used it fastest.

This is why two facilities with identical monthly kWh can land on very different bills. The one that spreads its load pays less. The one that stacks big loads into the same moment pays for a peak it felt for a quarter of an hour.

What a demand charge actually measures

Demand is power, measured in kilowatts. It is the rate at which you pull energy at an instant, averaged over the utility's interval. Energy is power applied over time, measured in kilowatt-hours. A 100 kW load running for one hour uses 100 kWh. The same 100 kW running for six minutes still registers as 100 kW of demand for that interval, even though it only consumed 10 kWh.

That gap between a short burst and a sustained draw is the whole game. Utilities meter demand because they have to build and maintain enough generation, wires, and transformers to serve your maximum pull, not your average one (EIA). The capacity has to be there for the worst interval even if that interval happens once a month. The demand charge is how they recover the cost of standing that capacity up for you.

Electricity prices for commercial and industrial customers reflect not only the energy they consume but the capacity the system must keep available to meet their highest level of demand, which is one reason large customers often face charges tied to peak power rather than total usage alone.

Paraphrased from the U.S. Energy Information Administration, Factors Affecting Electricity Prices

Read your own tariff sheet and you will usually find the demand component spelled out as a rate per kW, sometimes split into more than one type. Some schedules bill a facility-wide maximum regardless of time. Others bill a separate peak only for demand set during defined afternoon hours. The distinction matters a great deal once you start thinking about a battery, because a battery can only fight the peaks it is told to watch for.

Why the energy meter and the demand meter disagree

A common instinct is to attack a high bill by using less electricity. Turn things off, retrofit lighting, tighten schedules. All of that lowers the energy charge, and it is worth doing. It can leave the demand charge almost untouched.

The reason is that demand does not care about the total. It cares about coincidence. If you cut your lighting load by 40 kW but that lighting was never running during the fifteen-minute window that set your peak, your demand charge does not move. Meanwhile a single startup event that overlaps with your existing base load can lift your billed demand by 100 kW for the month, and you paid for that overlap even though it barely registered on your energy consumption (EIA).

So the two meters can send opposite signals. A building can trim energy use month over month and watch its demand charge hold steady or even climb, because the peak is set by timing, not volume. Any plan to cut a demand charge has to target the shape of the load curve at its tallest point. That is exactly the point a battery is built to attack.

How a battery shaves a peak

Peak shaving is simple in concept. A control system watches your real-time draw from the grid. When that draw starts climbing toward a threshold you set, the battery begins discharging, feeding part of the building's load from stored energy instead of the grid. The meter sees only the net draw from the utility, so if the building is pulling 320 kW and the battery is supplying 120 kW of it, the grid records 200 kW. Your demand for that interval, and potentially for the month, is 200 kW instead of 320.

The battery recharges later, during intervals when your draw is low and there is headroom under your target, often overnight or in the early morning. It stores cheap, low-demand energy and spends it during the brief windows that would otherwise set your peak. Across a month the building's total energy use barely changes. The maximum the grid meter ever sees drops, and the demand charge drops with it.

Two things make or break this. The first is the controller. Shaving a peak you cannot predict means the system has to respond within the interval, reading load in near real time and dispatching fast enough to hold the net draw under target. A slow or dumb controller lets spikes slip through, and a single missed interval can reset your billed peak for the whole month. The second is sizing, which decides whether the battery even has the power and the stored energy to cover a peak once it starts. That is its own section below.

Peak shaving versus shifting load across the day

Peak shaving and load shifting get used interchangeably, and they are related, but they answer different parts of the bill. Shaving is about the demand charge. It goes after the single tallest spike. Shifting is about moving energy consumption out of expensive hours into cheap ones, which goes after time-of-use energy charges and, on some tariffs, a time-defined demand charge.

On a time-of-use schedule, the utility prices energy higher during late-afternoon and early-evening peak hours and lower overnight. A battery can charge on cheap off-peak energy and discharge during the costly peak window, so you consume your own stored kilowatt-hours instead of buying expensive ones. In regions moving to export rules that pay less for midday solar sent to the grid, shifting self-consumption into the evening is often where the value sits (CPUC).

A well-run commercial battery usually does both jobs from the same hardware, but they can compete. Energy sitting in the pack that you were saving to flatten a 5 p.m. demand spike is energy you cannot also dump early to chase a time-of-use arbitrage. The dispatch logic has to know which charge is worth more on your specific tariff and prioritize accordingly. On a rate with a punishing demand charge and a modest time-of-use spread, protecting the peak wins, and the reverse favors shifting. This is a tariff question, not a hardware question.

Sizing storage to your peak profile

Sizing a demand-charge battery is a question with two numbers, and both have to be right. Power, in kW, is how much the battery can supply at once. Energy, in kWh, is how long it can keep supplying before it runs empty. A battery rated at 100 kW and 200 kWh can cover a 100 kW shave for two hours, or a 50 kW shave for four.

Start from your interval data. Pull a year of fifteen-minute meter reads if you can get them, and look at the shape of your peaks, not just their height. A peak that spikes 120 kW above baseline for a single interval needs a lot of power and very little energy. One that sits 60 kW above baseline for three hours every afternoon needs less power but far more stored energy. Size for the wrong one and the battery either cannot reach the peak or runs dry partway through and lets the tail reset your billed demand anyway.

Think of it as tiers of ambition. Shaving the top off a sharp, rare spike is the cheapest win, because it asks for power but little duration. Flattening a broad daily plateau costs more, because it demands hours of stored energy every afternoon. Erasing the peak entirely is usually the worst value, because the last few kW of reduction cost as much as the first and buy far less. Most commercial projects target the bulk of the demand savings without oversizing the pack to chase the final slice.

Load that is already flat is the hard case. If your draw hugs a steady line all month with no real spikes, there is no peak to shave, and a battery sized for demand management has little to work with. The best candidates have sharp, predictable peaks against a lower baseline, which is where a modest battery buys a large reduction in billed demand.

Charge types and what storage can do about each

Not every line on a commercial bill is something a battery can touch. It helps to separate the charges that respond to storage from the ones that do not. The table maps the common components, what drives each one, and whether a battery can do anything about it. Treat that last column as directional, because the size of the effect depends on your rate.

The pattern is that storage earns its keep against charges tied to peak kW and against time-priced energy. It does nothing for fixed fees and nothing for total consumption on its own. Knowing which lines on your bill are which tells you most of what a battery can save before you run a single number (EIA).

How solar and storage stack on a commercial bill

Solar and a battery attack different parts of the bill, which is why they work well together on a commercial site. Solar produces energy during the day, offsetting kilowatt-hours you would otherwise buy and lowering the energy charge directly (DOE). On its own, though, solar is an uneven partner against demand charges. Its output rises and falls with the sun, so a passing cloud at the wrong moment can drop production just as a load spikes, and your demand peak can reappear even with panels on the roof.

That is where the battery completes the picture. Solar fills the pack with low-cost daytime generation, and the battery then dispatches that stored energy on command, holding the net grid draw down during the exact intervals that set the demand charge regardless of what the sun is doing. The panels cut the energy charge across the day. The battery cuts the demand charge at the peak and shifts surplus solar into expensive evening hours (DOE).

The economics have one more layer. On the commercial side, both solar and storage can qualify for the federal Investment Tax Credit, which lowers the installed cost and shortens payback (SEIA). Storage has become a standard part of new commercial solar design rather than an add-on, which the industry's own tracking of installed capacity reflects (SEIA). The credit changes the payback math. It does not change where the savings come from, which is still your rate structure and the shape of your load.

Ratchets, windows, and the fine print that moves savings

Two clauses buried in commercial tariffs decide whether a battery pays off, and both are easy to miss. The first is the demand ratchet. On a ratcheted tariff, your billed demand for a given month is not just this month's peak. It is the higher of this month's peak and some percentage of your highest peak over the past eleven or twelve months. Set one bad spike in July and you can keep paying against a fraction of it through the following spring, even in months when your actual demand was far lower.

A ratchet raises the stakes on every peak. It also raises the value of a battery, because preventing a single spike does not just save one month, it keeps the floor from locking in for a year. Any sizing study on a ratcheted account has to weigh the cost of one missed interval far more heavily than on a plain tariff.

The second clause is the peak window itself. On a time-of-use demand charge, only the intervals inside a defined window count toward the demand peak, often a few hours in the late afternoon. A battery on that tariff has an easier job, because it only has to defend a known block of time rather than stay ready around the clock, and it can recharge freely outside the window. Export rules add a related wrinkle: where sending solar to the grid now pays less than it once did, keeping that energy in the battery for self-consumption during peak hours is usually worth more than exporting it (CPUC). Read these clauses before sizing anything. They change the answer.

Is your facility a fit for peak shaving

Before you commission a study, you can gauge whether your site is a strong candidate or a weak one. Work down this list with a copy of your tariff sheet and a recent bill in hand.

• Does your bill carry a demand charge of more than a few dollars per kW, shown as a separate line from the energy charge (EIA)?
• Are your peaks short and spiky against a lower baseline, rather than a flat line all month that leaves nothing to shave?
• Can you get interval data, the fifteen-minute meter reads, from your utility to see the real shape of your load?
• Does your tariff include a ratchet clause that lets one peak set a floor for future months, raising the value of preventing spikes?
• Do your peaks fall inside a defined time-of-use window, which lets a smaller battery do the job?
• Do you also have time-of-use energy spreads wide enough that the same battery can shift load and earn twice (EIA)?
• Is there roof or ground space for solar, so the project can stack energy and demand savings together (DOE)?

More yeses point to a stronger project. A facility with a steep demand rate, sharp peaks, a ratchet, and usable interval data is close to a textbook fit. A flat-load site with a tiny demand charge is one where a battery will struggle to pay for itself on demand savings alone.

Demand-charge storage mistakes that erase the savings

The projects that disappoint usually go wrong in a handful of predictable ways. Most of them trace back to treating a battery like an energy product when the bill is a power problem.

• Sizing from monthly bills instead of interval data. A monthly total hides the shape of your peaks. Without fifteen-minute reads you cannot tell a sharp spike from a long plateau, and you will size the wrong number (EIA).
• Buying energy capacity but not power. A big kWh pack that cannot discharge fast enough in kW will not catch a short spike. Demand shaving is a power problem first.
• Ignoring the ratchet. Let one interval slip on a ratcheted tariff and you pay against that peak for months. The controller has to be fast and reliable, not merely present.
• Expecting solar alone to cut the demand charge. A cloud at the wrong moment brings the peak right back. Without a battery to dispatch on command, solar is an energy tool, not a demand tool (DOE).
• Chasing the last kilowatt. Trying to erase the peak entirely oversizes the pack, because the final slice of reduction costs as much as the first and saves far less.
• Quoting a fixed dollar figure before reading the tariff. Savings are a function of your rate and your load. Any promise of a specific annual number before someone has seen your interval data and your rate schedule is a guess (SEIA).

Model it against your real tariff before you buy

Everything above points to one conclusion. A battery is worth exactly what your rate structure and your load curve let it save, and you cannot know that figure until someone runs the real data. A facility with sharp afternoon peaks on a steep, ratcheted demand charge can see a large reduction from a modest battery. The building next door, with a flat load and a token demand rate, might see almost nothing from the same hardware. Same equipment, opposite outcomes, decided by the tariff.

Enact builds for this problem on the commercial side. The platform runs energy analysis and financial modeling that accounts for demand charges and time-of-use rates, so a design is sized against your actual load profile and your actual rate schedule rather than a rule of thumb (Enact commercial solar and storage). That means a proposal shows what a given battery does to your billed demand, how solar and storage stack on your bill, and how the federal credit changes the payback, all on your numbers (SEIA).

For solar and storage providers who design these systems for clients, the same modeling feeds into a proposal you can put in front of a facility's finance team, with the demand-charge reduction and the economics laid out clearly (Enact design and proposals). Start from your interval data and your tariff. The peak that set your demand charge is a number you can see coming next month, and with the right battery sized to it, a number you can hold down.

Frequently asked questions

What is a demand charge?

A demand charge is a fee on a commercial electric bill based on the highest rate of power you draw from the grid during any short interval in the billing period, measured in kilowatts, usually over fifteen minutes (EIA). It is separate from the energy charge, which bills your total kilowatt-hours, so a brief spike can set a large demand charge even if your overall usage was modest (EIA).

How does a battery reduce demand charges?

A battery lowers a demand charge through peak shaving: a controller watches your real-time grid draw, and when it climbs toward a set threshold the battery discharges to supply part of the load, so the meter records a lower maximum demand (EIA). It recharges during low-demand hours, so your total energy use barely changes while the peak the utility bills against drops. How much you save depends on your rate and the shape of your peaks (EIA).

Do you need solar to use a battery for peak shaving?

No. A standalone battery can shave peaks and shift load on its own, charging from the grid during cheap, low-demand hours and discharging during your spikes. Solar is a strong complement because it fills the battery with low-cost daytime generation and cuts the energy charge, but it is not required for demand management (DOE). Pairing the two lets one project attack both the energy charge and the demand charge (SEIA).

How do I size a battery for demand-charge management?

Size on two numbers: power in kW, which sets how much of a peak the battery can cover at once, and energy in kWh, which sets how long it can hold that cover. Pull a year of fifteen-minute interval data and look at the height and duration of your peaks, because a sharp one-interval spike needs power and little energy while a long plateau needs far more stored energy (EIA). Most projects target the bulk of the reduction rather than erasing the peak entirely (EIA).

Does commercial battery storage qualify for the federal tax credit?

Yes. On the commercial side, storage can qualify for the federal Investment Tax Credit, which lowers the installed cost and shortens payback (SEIA). The credit improves the economics, but the operating savings still come from your demand charge and time-of-use rates, so the eligibility is a reason to model the full picture rather than a savings figure on its own (SEIA).

Why did my demand charge stay high after I cut my energy use?

Because a demand charge is set by your peak kW in a short interval, not your total kWh, so using less overall does not help unless you cut the load during the exact window that set your peak (EIA). A single startup event that overlaps your base load can hold your billed demand high even as monthly consumption falls, which is why demand charges often need a battery rather than efficiency measures alone (EIA).

Sources

1. EIA - Factors Affecting Electricity Prices
2. EIA - Average Price of Electricity (FAQ)
3. DOE - Solar Energy Technologies Office
4. DOE - Solar Performance and Efficiency
5. SEIA - Solar Investment Tax Credit (ITC)
6. SEIA - Solar Industry Research Data
7. CPUC - Net Billing Tariff
8. Enact - Commercial Solar and Storage
9. Enact - Design and Proposals