US Residential Solar-Plus-Storage Hits 46% Attachment Rate as Q2 Storage Sets 18.9 GWh Record

The battery is no longer an accessory to residential solar in the United States — it is becoming the default. The national residential solar-plus-storage attachment rate reached 46% in Q2 2026, up from 41% in the same quarter of 2025, according to the American Clean Power Association's US Energy Storage Monitor. In the same quarter, grid-scale storage additions hit a quarterly record of 18.9 GWh, per Wood Mackenzie and the Solar Energy Industries Association. Two markets, one signal: storage is no longer the optional half of solar-plus-storage.

That matters for anyone modeling these systems. When nearly half of new residential installs include a battery, the modeling question shifts from "should I add storage?" to "how do I size it correctly against the tariff it will actually operate under?" The answer is different in every utility territory — and it is where most residential proposals underperform their own projections.

What the Attachment Rate Record Actually Says

An attachment rate of 46% means that for every 100 new residential solar systems installed nationally, 46 included a battery. In Q2 2025 that figure was 41%; two years before that it was closer to a quarter. The trajectory is unambiguous and it is being driven by three structural forces rather than by consumer preference alone:

• Export compensation reform. Net-energy-metering successors — net billing, avoided-cost export, and instant-use-only tariffs — lowered the value of exported solar, making self-consumption the dominant revenue stream. A battery captures that value; a bare PV system increasingly does not.
• Time-of-use rate spread. Peak-to-off-peak price ratios of 3:1 or more in major markets reward storage arbitrage and demand-charge management in ways that a solar-only system cannot access.
• Resilience demand and virtualization. Backup power remains the emotional driver, but utility and aggregator virtual power plant (VPP) programs now pay enrolled batteries for grid services, adding a second revenue stack.

The 41%→46% jump in a single year is not a marketing artifact. It reflects a market where the marginal economics of pairing have improved precisely as the economics of not pairing have weakened.

The Grid-Scale Side: 18.9 GWh in One Quarter

Utility-scale storage is running the same race at a larger scale. The 18.9 GWh deployed in Q2 2026 was a quarterly record, following 9.7 GWh in Q1 — a near-doubling in three months. Wood Mackenzie and SEIA attribute the acceleration to longer-duration systems and sustained interconnection activity, with year-to-date storage additions contributing to a combined solar-plus-storage share of roughly 70% of all new generating capacity added in the first half of 2026.

The commercial significance is not just volume. Longer duration means the business case has shifted from short-duration peaking (capacity and frequency response) toward energy-shifting — the same service a residential battery performs behind the meter, but at 100–1,000x the scale and against wholesale price signals rather than retail tariffs.

Why Attachment Rate Is a Modeling Problem, Not a Market Fact

A 46% attachment rate tells you batteries are being bought. It tells you nothing about whether they are correctly sized. And that is where the gap between a proposal's projected savings and the owner's first-year bill tends to open.

The reason is that residential storage sizing is tariff-specific, not generic. A 10 kWh battery that performs well under a peak/off-peak TOU rate can be substantially oversized — or badly undersized — under a demand-charge rate or a solar buyback program with a low export rate. The engine that decides the optimal power-to-energy ratio is the tariff structure plus the load shape, not the battery catalog.

Three modeling inputs drive the result:

• 15-minute load and PV profiles. Hourly data hides the very peaks (EV charging, HVAC cycling, cooking) that a battery is meant to shave. Sub-hourly resolution is required for demand-charge accuracy.
• Export compensation logic. Whether excess generation earns retail, avoided-cost, or zero determines whether the optimal control strategy is self-consumption, arbitrage, or export-limited. Modeling the wrong one inverts the recommendation.
• Degradation over the tariff horizon. A battery sized to a first-year savings number will underdeliver by year 8–10 as capacity fades; a proper projection carries SOH decay through the full warranty term.

The VPP Overlay Changes the Optimization Target

The fastest-moving change in residential storage economics is aggregation. Utility and third-party VPP programs increasingly pay enrolled home batteries for capacity, frequency support, and peak reduction — turning a single-asset self-consumption problem into a portfolio dispatch problem.

For a single homeowner, a VPP adds a revenue stream worth, in many markets, a few hundred dollars per year, on top of bill savings. For the aggregator, it means thousands of batteries must be dispatched as one resource without violating any individual site's backup reserve. The engineering constraint that governs is the same on both sides: the state-of-charge window that preserves both the customer's backup floor and the aggregator's deliverable capacity.

Sizing Implications: What 46% Should Teach an Analyst

When half the market is pairing, the differentiator stops being "does the system have a battery" and becomes "is the battery the right size for this tariff, this load, and this degradation path." A few practical consequences for engineers and consultants:

• Model the tariff, not an average rate. A flat average price will recommend a different (usually smaller) battery than the actual TOU or demand-charge tariff the site operates under. Use the real rate schedule, including seasonal and weekend differentiation.
• Test both self-consumption and arbitrage control. In some territories self-consumption wins; in others, with high peak spreads, arbitrage dominates. The optimal EMS strategy should be an output of the model, not an assumption baked in.
• Run the sizing across the full horizon. A degradation-aware optimum is frequently larger than a first-year optimum, because it must still deliver the service in year 12 when nameplate has faded.
• Quantify the VPP option separately. Keep backup-reserve sizing and market-service sizing as distinct runs, then find the operating window that satisfies both.

A worked example: why average rates mislead

Consider a single-family site in a market with a 4:1 peak-to-off-peak spread. The household uses roughly 30 kWh/day, of which about 40% falls in the 4 p.m.–9 p.m. peak window. A solar-only system under a reformed export tariff earns avoided-cost compensation of roughly $0.05/kWh for its midday exports, while the evening load is served at a peak rate near $0.45/kWh. The annualized value of moving one kWh from midday export to evening self-consumption is therefore on the order of the full peak-to-export spread — a multiple of what the bare PV system realizes.

Now size the battery two ways. A first-year optimum shaves the peak window at the highest marginal value, which typically lands the system in the 8–12 kWh range for this load. A degradation-aware optimum, held to the same availability requirement through a 10-year warranty, frequently lands larger — because the system must still cover the evening peak in year 10 at perhaps 80–85% of nameplate. The engineering point is not the specific number; it is that the two methods yield different answers, and only the second one matches what the owner will experience.

This is exactly the class of problem a hybrid simulation platform is built for: 15-minute or hourly profiles, a manufacturer-specific degradation table, tariff-driven dispatch, and a 25-year financial projection that shows the owner where the battery pays back — and where it would not have.

What to Watch Next

Three signals will determine whether the Q2 records are a step-change or a plateau. First, whether the attachment rate keeps climbing as the residential solar market itself contracts — a rising share of a shrinking base still nets growth, but only up to a point. Second, whether longer-duration grid-scale systems continue to displace short-duration peaking, which shifts the modeling emphasis from power to energy. Third, whether VPP participation becomes standard rather than optional in utility programs, which would make fleet-level dispatch modeling a baseline requirement rather than a differentiator.

For now, the direction is clear. The two flagship US storage markets — behind the meter and in front of it — both posted records in the same quarter, for the same underlying reason: storage is the asset that converts intermittent generation into dispatchable value, and the tariffs on both sides of the meter increasingly reward exactly that.

The Trajectory Is Set

Residential attachment near 46% and grid-scale additions at 18.9 GWh per quarter are two views of the same transition. Storage is moving from an add-on sold on resilience to a core asset sized on economics — retail tariffs behind the meter, wholesale prices in front of it, and increasingly the same dispatch intelligence governing both.

For the analyst, the practical takeaway from the Q2 records is not the headline percentage. It is that the optimization target has matured: the question is no longer whether to include storage, but how to prove — numerically, across the full asset life — that the system you sized is the one that actually pays back.

Sources