Why a home battery sits idle on peak days

The study is by Sydney Forrester, Galen Barbose and Cesca Ann Miller of Lawrence Berkeley National Laboratory, published open access in iScience in 2022. Its summary is blunt: batteries operated solely to maximize solar self-consumption "provide customer bill savings up to $20–30 per kWh of storage capacity annually, but virtually no grid value." It is a few years old, but the rate structure it studies keeps spreading, so the finding is worth understanding.

The rate that makes a battery chase your own solar

Under classic net metering, solar you send to the grid is credited at the full retail price. Under net billing, which the paper calls the de facto successor in many states, you can still offset your own use with solar as it is generated, but surplus exported to the grid is credited at a lower rate, often tied to the utility's avoided cost. At the time of writing, the authors counted some form of net billing in ten states, with at least five more proposing it.

That gap between what you pay for power and what you get for exports pushes owners toward one goal: export as little as possible. A battery does that by soaking up midday surplus and releasing it into the house later. The study shows why this is a big deal. Without storage, between 47% and 72% of a typical rooftop system's annual output was exported across the households studied.

What the study measured

The core data are metered, hourly electricity use for 1,749 homes across six utilities: Detroit Edison (Michigan), Green Mountain Power and Vermont Electric Cooperative (Vermont), Lakeland Electric (Florida), NV Energy (Nevada) and the Sacramento Municipal Utility District (California). Solar output was simulated for each location with NREL's System Advisor Model. The batteries were modeled as roughly two-hour lithium-ion units, which the authors say describes almost all residential batteries on the market.

They then compared two ways of running the same battery: dispatch for self-consumption, and dispatch against wholesale prices and peak-related grid costs.

Good for the bill, idle for the grid

For the owner, self-consumption saves up to $20 to $30 per kWh of battery capacity a year. At battery prices of $700 to $1,300 per kWh, the authors conclude that buying storage only for self-consumption "clearly is not an economical customer investment at present," with simple paybacks well beyond the expected ten-year battery life. (Their point is that people buy batteries for several reasons, backup power prominent among them, not that batteries are a bad idea.)

For the grid, the result is starker. The authors measured how much the batteries discharged during the system's highest-load hours. Capacity factors were generally under 20% during bulk-system peaks and under 30% during local distribution peaks. Assigning a capacity cost of $50 per kW-year to the top 40 load hours, the peak value of self-consumption batteries was roughly zero for virtually all customers at four of the six utilities. Only NV Energy and SMUD, both with heavy residential cooling loads, showed meaningful positive value, and still well short of what the same batteries could deliver if dispatched for the grid.

The reason is the weather

You might guess the problem is timing: the battery empties in the evening when the grid peak is earlier or later. The authors say that is part of it but "not the primary factor." The main cause is simpler:

"storage operated solely for solar self-consumption tends to stand idle on peak load days."

Homes tend to hit their own peak demand on the same days the whole system peaks. On those days the house uses more of its own solar as it comes in, so, as the authors put it, relatively little surplus is available to charge the battery. A battery whose only job is to store leftover solar has nothing to store on exactly the days that matter most to the grid. The study's peak-day plots show shallow cycling under self-consumption, versus full discharge at the peak hour under market-based dispatch.

There is a second, smaller mismatch on ordinary days. Self-consumption batteries charge mid-morning from surplus solar, while the cheapest wholesale power usually comes in the middle of the night.

What would fix it

The authors trace the inefficiency to three things: the price gap between exports and self-use, prices that do not vary by time, and limits on charging from or discharging to the grid. Their modeling found that the potential energy and peak value under market-based dispatch, combined, was greater than the upper end of the bill savings owners get from self-consumption under net billing. They read that as a sign that both the owner and the grid could come out ahead with better price signals.

The obvious objection is that letting thousands of home batteries discharge to the grid would overload neighborhood circuits. The study addressed this: unconstrained grid discharge does raise exports to distribution networks, but 50% to 70% of the potential market value was achievable "without materially degrading solar self-consumption levels or increasing local grid stress," mainly by targeting discharge at narrow peak periods.

An earlier Dutch study points the same way. Using measured data from 79 households in Amersfoort, Schram and colleagues found batteries sized for self-consumption cut neighborhood peak demand by only 5.7%, but that smarter control could raise the reduction to 22% with simple rules, or 51% with perfect foresight. I read that one at the abstract level only.

The honest limits on this

The household load data are from 2012 and 2013. The peak value is a stylized, uniform calculation that the authors say does not accurately represent what batteries earn in today's capacity markets or peak programs, and they did not model location-specific distribution limits. pv magazine's write-up added an editor's note that the "virtually no grid value" finding applies mainly where net billing uses flat prices; time-varying export rates, as in California, change the picture, though the study still found net billing degrades storage value there. And the study does not tell any single owner what their battery is worth. Rates, programs and climate vary too much.

What to take from it

A home battery's value to anyone besides you depends on its control logic and on the rate it runs under, not on the hardware. If your utility offers a time-varying rate or a program that pays for discharge during peak events, that is the mechanism this study says unlocks value that self-consumption leaves on the table. If not, the battery will mostly do what it was told, which is to wait for spare sunshine.

This is close to home for me: I help design the AI battery-cycling systems for a veteran-owned (HUBZone) energy-storage integrator, run by fellow veterans. Deciding when a battery should charge and discharge is the whole job. I do not own that company and earn nothing from this link. Full policy here. Nothing in this report involves them or any product they sell.