A 350 kW DC fast charger running at full power for fifteen minutes can add more to a monthly electric bill than the session that caused it earns in revenue. That is the demand charge problem covered in DC Fast Charging ROI: utilities bill commercial customers on their single worst 15-minute peak of the month, not on how much energy they actually sold. A battery sitting between the grid connection and the chargers is one real fix for that problem. It is also being sold right now as a default add-on to sites that do not need it yet.
What a battery-buffered charger actually does
A battery-buffered DCFC system puts a battery pack between the utility meter and the charger. The battery trickle-charges from the grid at a modest, steady draw between sessions, then discharges at high power the moment a car plugs in. The grid only ever sees the battery's slow recharge rate, not the vehicle's peak draw. The car gets the same charging speed either way.
A 2024 NREL technical assistance case study (OSTI document 89492) done with the Joint Office of Energy and Transportation for a state DOT evaluating a grid-constrained fast-charging site, and its companion fact sheet (OSTI 89493), put a number on this: an appropriately sized battery-buffered system can cut the grid service capacity a site needs by roughly 50 to 80 percent compared with a station powered entirely from the grid, with the same charging experience for the driver. NREL describes that range as an approximation derived from its reference tables for a 150 kW corridor station at different grid-supplied power levels, and says the underlying assumptions are available on request. This research pass confirmed the range through NREL's published document summaries but could not open the full PDFs, so pull the original before citing site-specific sizing.
Reducing required interconnection capacity is not the same measurement as reducing the billed coincident peak that sets a demand charge, though the two move together in practice: a battery sized to shave a site's grid-service requirement by a given percentage generally shaves the metered peak the utility bills on by a similar amount, because both are driven by the same underlying vehicle draw the battery is absorbing. The math below treats the two as roughly interchangeable for that reason, not because they are measured identically. Curtailing charging for payment during grid events is another way to ease the peak; see demand response for EV chargers.
ADS-TEC Energy sells commercial hardware built on this same battery-buffered idea; the company's ChargeBox, with about 140 kWh of integrated storage, is marketed as delivering up to 320 kW to a vehicle from as little as 50 kW of grid input, and ADS-TEC claims savings of up to 65 percent on peak-power demand charges versus conventional DC chargers (per its 2024 press release). This research pass could not reach ads-tec-energy.com, so verify the specific product and current spec sheet with ADS-TEC directly before citing it in a proposal. Treat the savings percentage as a vendor's own claim, not an independently audited result: it is the kind of number a system integrator should reproduce against your actual tariff and load profile before you sign a contract, not a promise to take at face value.
The demand-charge math a battery is actually fighting
Use the site's own published demand-charge benchmarks for a four-port, 350 kW managed-peak DCFC site (from DC Fast Charging ROI) rather than a vendor's example, since these are the numbers this site already verifies against real utility tariffs:
| Utility demand rate | Annual demand charge, no battery |
|---|
| $5/kW/month (Southeast, parts of Midwest) | $21,000 |
| $10/kW/month (Midwest, Mountain West) | $42,000 |
| $15/kW/month (California and Northeast secondary) | $63,000 |
| $25/kW/month (California IOU peak, NYC, Boston suburbs) | $105,000 |
Apply the NREL range's conservative-end figure, a 50 percent cut to coincident peak, to each tier, and the annual demand-charge savings from the battery run from about $10,500 at the $5/kW tariff up to about $52,500 at the $25/kW tariff. Apply ADS-TEC's marketed 65 percent reduction instead and the range moves to roughly $13,700 to $68,300 a year. Either way, the same battery, sized to the same 350 kW coincident peak, produces wildly different dollar savings depending entirely on which utility territory the site sits in.
That is the piece vendor content usually skips: the battery's technical performance does not change by tariff, but the annual savings do, by a factor of five between the top and bottom rows of that table. This article does not have an independently verified battery-buffered system cost to divide those savings into (see the verification note below), so it cannot pin down an exact payback year for your site. What it can say is that whatever payback a $25/kW market achieves on a given battery and installation cost, a $5/kW market on the identical hardware takes roughly five times as long to reach. Vendors quoting a single "3 to 5 year payback" figure, a range that shows up repeatedly in industry cost guides, are almost always quoting the high-tariff end of that range. Read your own utility's actual demand-charge rate off your tariff sheet (see How to Read Your EV Tariff Sheet) and get a system quote before assuming a vendor's payback example transfers to your site. A payback built on a $25/kW California example does not transfer to a $7/kW Tennessee site.
Why low utilization makes this worse, not better, for the vendor pitch
Here is the counterintuitive part. How to Read Your EV Tariff Sheet makes the point that the demand charge "behaves worst exactly when utilization is low," because the charge is driven by the single worst 15-minute peak, not by total energy sold. A site with only a handful of sessions a day can still hit a full coincident peak if two fast chargers happen to fire at once, and that peak costs exactly as much whether it happens once a week or ten times a day.
So a low-utilization site does have a real demand-charge problem, proportionally the worst kind: if it hits a qualifying coincident peak in most billing months, its annual demand charge can be nearly as high, dollar for dollar, as a much busier site with the same peak magnitude. But that is not the same question as whether a battery is the right fix for it, and this is where the vendor pitch and an honest ROI model diverge. The battery's capital cost does not scale down with session count. A four-port site running at 4 to 8 percent utilization gets the exact same battery bill as one running at 16 to 20 percent, but it has far less total charging revenue to amortize that fixed cost against. The demand charge itself may already be nearly as painful at the low-utilization site; the battery still costs the same six figures either way, and it is the thin revenue base, not a smaller demand-charge problem, that stretches the payback past what the business case can carry. The battery solves the technical problem regardless of utilization. It does not solve the economics of a site that has not yet earned enough total revenue to justify the capital.
There is also a cheaper fix that usually comes first. Load management software, which staggers session starts and caps simultaneous draw across ports, is the standard, lower-cost answer to exactly this problem and, per DC Fast Charging ROI, typically reduces coincident peak by 30 to 60 percent and pays for itself within a year at any site with more than two ports. A battery adds real capability beyond what software alone can do, most importantly at sites where the grid connection itself is the constraint, not just the bill, which is the actual scenario NREL's case study addresses. But if your problem is purely the bill and you have not yet deployed load management, that is the $1,000 to $5,000-a-year fix to try first, not a battery system costing many multiples of that.
When it pencils, and when it is premature
A battery earns its capital at sites that clear three conditions together, not any one of them alone:
A demand rate worth fighting. Territories in the $15 to $25/kW range and above give a battery enough annual savings to work with. Below roughly $10/kW, the absolute dollars saved rarely clear a reasonable payback window regardless of how well the battery performs technically.
Enough session volume to amortize the hardware. A site running at or above the moderate end of the utilization ranges in DC Fast Charging ROI, meaning a well-located corridor or retail site closer to the 16 to 20 percent range than the 4 to 8 percent reference case, generates enough total revenue that a several-year battery payback is a reasonable ask of the business case. A brand-new site still building EV traffic in its first year or two, running near the 8 percent NREL reference case, has not yet earned that.
A real constraint beyond the bill, or a tariff that already justifies it on its own. If the actual barrier is grid capacity, meaning the utility cannot deliver the power a full buildout needs without an expensive service upgrade, NREL's case study is describing your situation directly, and the battery can be the difference between building now and waiting years for a utility upgrade. See Electrical Infrastructure Assessment for how that constraint gets identified before you commit budget.
Vendors selling battery-buffered chargers have every incentive to present the battery as a default line item on any DCFC project, because the technology genuinely works and the case studies are real. The honest treatment is that it works best as a targeted answer to a specific, quantified problem (grid capacity, or a demand charge north of $15/kW at a site with real volume), not a blanket recommendation. Run your own tariff and utilization numbers through the model in DC Fast Charging ROI before you add a six-figure battery system to a site that has not yet proven it can fill its chargers.
Last factually verified: October 19, 2026. Sources: National Renewable Energy Laboratory and Joint Office of Energy and Transportation, 2024 battery-buffered fast charging technical assistance case study (OSTI 89492) and companion fact sheet (OSTI 89493), whose 50 to 80 percent grid-capacity range was confirmed through published document summaries, not the full PDFs (nrel.gov was unreachable to this pass); ADS-TEC Energy's ChargeBox specifications (140 kWh storage, 320 kW output, as little as 50 kW grid input, up to 65 percent demand-charge savings) as stated in its 2024 press release coverage, flagged as vendor claims and not independently audited; an earlier $92,000 dealership savings figure could not be sourced and was removed. The demand-charge rate table is carried over from this site's DC Fast Charging ROI article and was not re-checked against utility tariffs in this pass; the 3 to 5 year payback range is a commonly cited industry figure that could not be tied to a specific verified system cost. evcharginghelp.com is editorially independent and receives no compensation from any company mentioned.