Grid-Scale Storage Economics: The Revenue Stack

U.S. operators added 8.3 GW of utility-scale battery storage in the first half of 2026, reaching roughly 52 GW of nameplate capacity, with the Energy Information Administration expecting close to 65 GW connected by year end. Batteries now account for 28% of planned utility-scale capacity additions — 24 GW this year against a record 15 GW in 2025. Yet storage remains one of the hardest energy assets for capital to price. A battery does not earn from one contract. It earns from three markets at once, and the mix changes every month.

Three revenue lines, one asset

A grid-scale battery monetizes the same megawatt-hours three different ways, and the discipline of the asset class is knowing which line is paying the bills in any given quarter.

Energy arbitrage is the intuitive one: charge when power is cheap, discharge when it is expensive. The spread is set by the shape of the local demand curve — in California by the solar-driven duck curve, in Texas by summer scarcity and wind volatility. It is the most variable line and the most weather-dependent.

Capacity, where the market offers it, is the contracted line. California's Resource Adequacy program pays batteries to be available when the system operator needs them, independent of whether they ever dispatch. That payment behaves like a contracted revenue stream, which is why lenders treat it differently from everything else in the stack.

Ancillary services — frequency regulation, responsive reserve, regulation up and down — pay a battery to hold capacity in reserve for grid stability. Historically this was the highest-margin line per megawatt, and it is the one that has proven most fragile.

Duration sits underneath all three. Modo Energy's benchmarking found two-hour batteries out-earned one-hour systems in every month over a two-year span, by 15% to 81%, with the widest premium in winter. Duration is not a technical footnote; it determines which of the three lines an asset can meaningfully compete in.

The ERCOT lesson: a revenue line can vanish

Texas ran the cleanest natural experiment in the asset class, and the result should be required reading for anyone underwriting storage.

ERCOT's ancillary services markets are small relative to its energy market. As battery capacity scaled — to roughly 14.96 GW installed by the end of Q1 2026 — those markets saturated. Ancillary revenues fell close to 90% from their 2023 peak. Average annual BESS revenue in ERCOT dropped from about $149 per kilowatt in 2023 to roughly $17 per kilowatt in 2025, according to market benchmarking.

The stack did not disappear; it rotated. ERCOT batteries earned 76% of revenue from energy arbitrage in June 2025, up from 25% a year earlier. The introduction of Real-Time Co-Optimization plus Batteries reshaped dispatch economics again, letting the system operator optimize energy and ancillary awards together.

What remains is volatility that no quarterly report captures well. ERCOT battery revenue ran at an annualized $46,264 per megawatt in January 2026, fell to $15,306 in February, and recovered to $38,145 in April. A three-fold monthly swing is not an anomaly in this asset class. It is the base case, and it is the single reason merchant storage struggles to attract capital that is not staffed to watch it.

CAISO's floor: market design is credit design

California built the opposite structure, and the difference in outcomes is instructive.

Resource Adequacy provides more than half of most CAISO batteries' revenue. Full-stack CAISO BESS revenue averaged $14.83 per kilowatt-month in 2025 where an RA contract was in place — more than four times what merchant wholesale participation earned alone. Merchant revenue hit a record monthly low in December 2025 and had only partly recovered by June 2026, to $2.55 per kilowatt-month. The Slice-of-Day reform is now changing how that capacity value is earned across the day.

Same technology, same chemistry, frequently the same developers — and two entirely different credit profiles, determined by market design rather than by engineering. An institutional allocator looking at "battery storage" as a single asset class is looking at the wrong unit of analysis. The unit is the project inside its market, with its contract position stated explicitly.

What a lender will actually size

Project finance has settled into a reasonably clear set of rules, and they explain the industry's drift toward contracted structures.

Tolling agreements transfer commercial operation of the asset to a counterparty, which takes market price risk in exchange for a fixed fee to the owner. Returns improve as more revenue is contracted, but the benefit levels off around 80% — banks generally do not extend debt against contracted revenue beyond that level, so the last slice of merchant exposure is equity's problem by design.

The sizing gap is concrete. Lenders typically cap senior debt at 1.30–1.40x P50 debt service coverage, or 1.10–1.15x P90, on merchant projects, against roughly 1.20x P90 on contracted toll structures. On the equity side, merchant projects without a long-term capacity contract or toll generally require returns 300–500 basis points above fully contracted equivalents.

Hybrid structures have emerged to split the difference: a floor arrangement guarantees minimum annual revenue per megawatt while the owner retains a share of upside above that floor, commonly in the 50–85% range. That is the same structural logic that makes contracted offtake financeable in other energy verticals, which we covered in power purchase agreements as underwriting collateral — contracted offtake, counterparty credit and tenor remain the three inputs that decide bankability.

The data problem that keeps capital out

None of the above is a technology problem. It is a reporting and access problem, and it is where the capital markets side of this asset class is furthest behind.

An asset whose revenue mix can rotate from 25% arbitrage to 76% arbitrage in twelve months cannot be monitored on a quarterly PDF. The institutions that own storage today — infrastructure funds, utilities, a handful of specialist credit desks — employ analysts who watch nodal prices and ancillary clearing daily. That staffing requirement, far more than minimum check size, is what excludes everyone else. A family office or smaller institution can underwrite a triple-net building from a lease abstract. It cannot underwrite a merchant battery without a data feed it does not have.

Three things change that, and none of them are exotic. First, an asset data feed with the properties a credit feed already has: a defined source of truth, a stated update cadence, a dispute path and independent attestation. Dispatch volumes, state of charge, realized capture spreads and contract status are all machine-readable at the source — they simply are not published to holders in a structured form. We set out what an institutional-grade feed requires in our work on oracle infrastructure.

Second, holder-level reporting that matches the cadence of the asset rather than the cadence of the fund administrator. Monthly revenue-by-product reporting is standard practice inside operators. Publishing it to holders is a plumbing decision.

Third, structures that make smaller positions administratively rational. The cost per holder of onboarding, distributions and reporting is what forces million-dollar minimums on infrastructure equity, not any principle of the asset. A digital capital markets platform lowers that cost — integrated onboarding, an on-chain holder register, standardized distributions — which is the mechanism that makes a lower minimum workable. The broader framing for energy and digital infrastructure assets is in our pillar piece on tokenization for energy and digital infrastructure.

Grid-scale storage is one of the fastest-growing hard-asset classes in the United States and one of the least accessible. The economics are legible to anyone willing to read three revenue lines instead of one. What has been missing is a way to publish those three lines to holders at the speed the asset actually moves, and to hold a position small enough that the analysis is worth doing. Both are solvable.

Have an asset you're thinking about? Register at commertize.com to see the platform — onboarding, KYC, holder dashboard and reporting. Or contact the team and tell us what the asset is; if it isn't a fit, that is a useful answer to get in one conversation rather than three.

Educational only — not legal, tax or investment advice, and not an offer of any security. Any securities offering is made by a sponsor, through documents prepared by the sponsor's counsel, under an exemption that counsel determines.

Have an asset you're evaluating for tokenization? Send the offering memo to deals@commertize.com or start at commertize.com/tokenize, and we will return a written tokenizability and capital-structure memo within 48 hours — free, no obligation.

Confidential review. No cost, no commitment, no calls unless it is a fit.