Other Electric Power Generation (U.S.) — NAICS 221118

An investor primer for public-market and private investors. Federal statistics are as-reported; forward-looking statements are labeled as judgments.

---

1. Overview

The North American Industry Classification System (NAICS) code 221118 is a small, technical "catch-all" bucket for U.S. electricity generation that does not fit the seven named fuel codes — a residual home for things like tidal and wave power. As a formal Census category it is tiny: 46 firms and about $41.5 million of receipts in 2022 [3]. But investors reach for this code as the closest label for a much bigger story — grid-scale electricity storage (batteries and pumped-storage hydro), which the federal statistics mostly file elsewhere. That storage fleet is one of the fastest-growing parts of U.S. energy infrastructure: utility-scale batteries passed 26 gigawatts (GW) by the end of 2024 and are still compounding [5]. Public-market investors get exposure through diversified owner-operators and one storage-equipment maker; private investors buy or build the projects themselves, or fund the developers and infrastructure platforms that do. This primer covers both routes and is honest about where the official numbers stop and the market begins.

Scope note: because the storage assets investors care about are usually classified under their owners' primary industry (a utility, an independent power producer, a manufacturer, or a fund), the pages below use two clearly labeled lenses — the narrow federal statistics for 221118, and the broader physical storage market.

---

2. What it is and how it is structured

Formal scope. NAICS 221118 covers establishments that operate electric-generation facilities "except hydroelectric, fossil fuel, nuclear, solar, wind, geothermal, biomass" — converting other energy forms (the Census example is tidal power) into electricity delivered to the grid [1]. It is the eighth, residual member of industry group 22111 (Electric Power Generation). The seven codes it excludes are its nearest neighbors [1]:

Also excluded: the "wires" businesses — 221121 (Electric Bulk Power Transmission and Control) and 221122 (Electric Power Distribution) — and, in manufacturing, 335910 (Battery Manufacturing). The people who make batteries and the contractors who build the plants are not in 221118; only the asset owner-operator that dispatches power to the grid is [1].

Why storage does not fit cleanly. A battery creates no new energy — it stores electricity and returns part of it later, so the U.S. Energy Information Administration (EIA) treats it as a secondary source, not primary generation [10]. As a result, a battery's Census industry usually follows its owner's main business (a regulated utility, an independent power producer, a developer, a manufacturer, or an investment fund). Pumped-storage hydro likewise sits formally under hydroelectric (221111) [1]. The practical upshot: the 221118 business statistics do not measure the storage market, and this primer sizes the two separately. (The two research reports underlying this primer disagreed on this point; the reading here — 221118 as a narrow residual, storage counted mostly elsewhere — is the one consistent with the tiny official figures.)

Ownership mix. The investable storage universe spans four buckets:

• Investor-owned utilities (IOUs) and their unregulated arms — e.g., NextEra Energy, AES, Dominion, Duke, Southern subsidiaries.
• Independent power producers (IPPs) and merchant developers — Vistra, Plus Power, Jupiter Power, LS Power, and dozens more.
• Government, public-power, cooperative, and federal owners — especially of the legacy pumped-storage fleet (below).
• Private capital / infrastructure funds — Brookfield, Global Infrastructure Partners (now part of BlackRock), and similar sponsors buying contracted projects and platforms.

Batteries skew toward IPPs and private capital; pumped storage skews toward utilities and government. An EIA producer-type proxy put the battery-heavy residual category at roughly 85% IPP-owned versus about 10% electric-utility-owned, while pumped storage was about 84% utility-owned [38].

---

3. How big it is

Lead with the federal statistics — they are small, and they undercount by design.

The 2022 Economic Census receipts of ~$41.5 million were about 0.03% of the roughly $146 billion the same Census reported for all U.S. electric-power generation [3]. Note the two surveys are not directly comparable — the Economic Census (firms, receipts; 2022) and County Business Patterns (establishments, employment; 2023) differ in coverage and reference year [2][3]. Both undercount the economic weight of storage for the reason in §2: the largest assets roll up under their parent utilities, IPPs, and funds. Read these figures as a floor, not the industry's true size. (Among the taxable firms captured, employment did rise off a low base — roughly 446 in 2021 to 1,411 in 2023 — a faint fingerprint of the build-out [2].)

The broader measure that matters — the physical storage market (a different lens, labeled as such):

• Utility-scale batteries exceeded 26 GW of power capacity at year-end 2024, after adding 10.4 GW in 2024 (a 66% jump), yet still only about 2% of the 1,230 GW U.S. utility-scale fleet [5]. One gigawatt is 1,000 megawatts (MW); one MW is roughly the instantaneous draw of ~750 homes.
• The EIA reported a record 15 GW of batteries added in 2025, with developers planning another 24 GW for 2026 — plans, not completions [6].
• Pumped-storage hydro, the legacy backbone, is 43 plants, ~22 GW, and ~553 gigawatt-hours (GWh) of stored energy, with a median duration near 12 hours versus about two hours for the early battery fleet [7].
• Observed 2024 battery construction spending averaged $1,469 per kilowatt (kW), roughly $16.3 billion of included cost across ~11 GW installed [8].

Bottom line on size: the formal 221118 industry is a rounding error; the storage market it stands in for is a tens-of-billions capital story growing at double digits.

---

4. The investable universe

There is no pure-play "221118" public stock. Listed exposure comes bundled inside diversified utilities, merchant power companies, and one equipment integrator. Market data below are as of the July 17, 2026 close [36]; storage is a fraction of each company's revenue, which none break out cleanly (so no clean per-company storage-revenue figure exists).

(P/E = price-to-earnings; n/m = not meaningful — FLNC posted a trailing net loss of ~$41.8 million [14].) Other listed names with material — but small-relative-to-total — storage or pumped-hydro exposure include Constellation Energy, GE Vernova, Tesla, Brookfield Renewable, Dominion, Duke, and Southern Company subsidiaries. Diversified, indirect access is available through exchange-traded funds (ETFs) such as GRID (First Trust Nasdaq Clean Edge Smart Grid, 0.56% expense ratio) or IGF (iShares Global Infrastructure, 0.39%); note that a "battery" ETF like BATT is mostly upstream materials and electric-vehicle exposure — only ~7% was classified as energy storage in mid-2026 [37].

Major private and other owners (reached by direct investment, co-investment, or debt rather than a ticker):

---

5. How the money works

Storage owners earn money in one of two very different ways, and some do both.

(a) Regulated rate-base model (utility-owned storage). When a utility such as Florida Power & Light or Dominion owns the asset, a state public utility commission (PUC) — or the Federal Energy Regulatory Commission (FERC) for transmission-function assets — folds the prudent capital cost into the utility's rate base and lets the utility earn an allowed return on equity (ROE) on the equity portion, plus recovery of depreciation, taxes, and operating costs. Recent decisions cluster near 10%: the California PUC authorized 2026–2028 ROEs of 9.98% (PG&E), 10.03% (Southern California Edison), and 9.93% (SDG&E) [21]. The earnings engine is simply more approved capital → more rate base → more earnings — a low-volatility, bond-like model, though the PUC can disallow costs it deems imprudent.

(b) Competitive / merchant model (most standalone batteries). A merchant battery earns a stack of market revenues that shifts as more batteries compete:

• Ancillary services (fast frequency regulation and reserves) — the richest early revenue, but it saturates quickly. In the Texas market (ERCOT, the Electric Reliability Council of Texas), battery revenue averaged ~$70/kW annualized in the first half of 2024, and ancillary-service revenue later fell roughly 90% from its peak by late 2025 [22].
• Energy arbitrage — charge when power is cheap (midday solar), discharge into the evening peak. This is now the dominant use: the EIA found arbitrage was the primary use for 41% of U.S. battery capacity in 2024, with frequency regulation second at 24% [10].
• Capacity payments — being paid to be available. In the PJM Interconnection (the mid-Atlantic grid operator), the 2025/2026 capacity auction cleared at a record $269.92 per MW-day (up roughly 9× year over year), far higher in constrained zones (Dominion $444.26); analysts attributed ~63% of the increase — about $9.3 billion — to data-center load [23].
• Contracts — to tame that volatility, developers increasingly sign power purchase agreements (PPAs) or tolling agreements, where an offtaker pays a fixed fee for dispatch rights, converting choppy merchant revenue into bond-like cash flow that lenders and infrastructure funds will finance [24].

Costs and returns. A four-hour lithium-ion system was benchmarked near $334/kWh by the National Renewable Energy Laboratory (NREL), with ~85% round-trip efficiency [9]; system costs have fallen steeply and are expected to keep falling, which turns thin merchant spreads into equity returns. The single biggest lever is tax credits. The Inflation Reduction Act (IRA) of 2022 made standalone storage eligible for the investment tax credit (ITC) for the first time. Under the technology-neutral §48E Clean Electricity Investment Credit, storage placed in service after 2024 earns a 6% base credit, up to 30% with prevailing-wage and apprenticeship rules, plus +10 percentage points each for domestic content and for siting in an energy community — and the credit is transferable, with elective ("direct") pay for public entities [25]. Storage generally does not earn a production tax credit (PTC), because it creates no primary electricity [25].

---

6. What drives demand

1. Renewables firming. Solar and wind are the largest source of new capacity; batteries shift midday solar into the evening peak, making variable output dispatchable. Wind and solar were projected to rise from ~18% of U.S. generation in 2025 toward ~21% by 2027 [33]. This is the structural engine.
2. Data-center / AI load. After nearly two decades of flat demand (~0.1%/year, 2005–2019), U.S. electricity demand grew ~1.7%/year from 2020–2025, with data centers a principal driver; the EIA projected ~1% growth in 2026 and ~3% in 2027 [33]. This load surge is what drove PJM capacity prices up ~9× [23].
3. Reliability and thermal retirements. Coal and older gas plants retiring increases the need for fast-responding replacement capacity.
4. Electrification of vehicles, heating, and industry, which lifts peak demand.
5. State mandates. California's early storage-procurement target was 1,325 MW; New York set a 6 GW by 2030 goal [34].
6. Falling costs, which compound the volume flywheel.

---

7. Regulation

• FERC governs wholesale markets and interconnection. Order 841 (2018) forced regional grid operators (RTOs/ISOs — regional transmission organizations / independent system operators) to open energy, capacity, and ancillary markets to storage — the pivotal market-access rule for merchant batteries [28]. Order 2023 replaced slow serial studies with first-ready cluster interconnection, and Order 1920 requires 20-year regional transmission planning [29]. Interconnection queues are the binding constraint: at end-2023 U.S. queues held ~1,570 GW of generation plus ~1,030 GW of storage, with a median wait of five years; the queue neared ~2,300 GW total by end-2024 (~890 GW of it storage), even as new storage entries fell ~32% amid policy uncertainty [30].
• State PUCs set allowed ROE and rate-base treatment for utility-owned storage, run procurement mandates, and — after the Moss Landing fire — set new battery safety and emergency-response standards [21].
• EPA (U.S. Environmental Protection Agency) does not regulate a battery like a smokestack (there are no stack emissions), but it shapes fire safety, hazardous-waste and end-of-life handling, and post-incident cleanup; it points to standards such as NFPA 855 (National Fire Protection Association) and UL 9540/9540A for safe design [31].
• NRC (Nuclear Regulatory Commission) does not regulate grid storage — nuclear is a separate code (221113). It matters here only to note the exclusion and to avoid confusing grid batteries with NRC-regulated spent-nuclear-fuel storage [32].
• Trade / Treasury. Section 301 tariffs and Internal Revenue Service guidance on foreign-entity rules (below) now co-determine project economics.

---

8. Competitive dynamics and consolidation

Development is fragmented — hundreds of developers can secure land and file interconnection requests — but the durable advantages concentrate: a viable interconnection position, a high-value grid location, permits and community acceptance, a creditworthy offtaker, bankable equipment warranties, low-cost capital and tax-credit monetization, and real-time trading skill. NextEra is the largest U.S. battery owner, with roughly three times the operating capacity of its nearest rival [15]. The equipment/integration layer is a separate contest led by Tesla and Fluence [14]. Infrastructure capital is the marginal buyer of de-risked (contracted) assets — Brookfield's Neoen stake and LS Power's platform build-out are templates [18][19] — so consolidation is likely to continue as small developers need capital and large sponsors want pipelines and operating teams. One unusual dynamic to underwrite: self-cannibalization — batteries crowding the same grid node flatten the very price spreads that drew them, so backward-looking revenue is a poor guide to future returns in fast-growing markets.

---

9. Risks

1. Fire and safety. The January 2025 Moss Landing fire triggered a ~$400 million write-off plus a later ~$155 million impairment for Vistra, litigation, and new state safety rules — the industry's reputational and insurance overhang [13][31].
2. Supply chain, tariffs, and foreign-entity rules. U.S. storage leans heavily on Chinese lithium iron phosphate (LFP) cells (~$73/kWh in China, 2024). Section 301 duties on Chinese storage batteries are rising toward roughly 58–82% by 2026, and new foreign entity of concern (FEOC) rules can strip tax-credit eligibility from China-linked supply chains — raising cost and complicating credit qualification at once [27].
3. Revenue saturation. As more batteries chase the same ancillary and arbitrage pools, per-unit revenue falls (ERCOT ancillary revenue −90% from peak). Returns depend on entering early or locking in contracts [22].
4. Interconnection and permitting. Multi-year queues, restudies, and unexpected network-upgrade costs strand capital [30].
5. Policy reversal. Storage kept its ITC under the 2025 One Big Beautiful Bill Act (OBBBA), but the phase-down after 2033 and tightening FEOC rules create a policy cliff [26].
6. Interest rates. Storage is capital-intensive and financed with debt plus tax equity; higher rates directly compress equity returns.
7. Degradation and duration. Cells fade and need costly augmentation; four-hour lithium-ion cannot cover multi-day shortfalls, and long-duration alternatives (e.g., Form Energy's 100-hour iron-air, targeting under $20/kWh) are still pre-commercial at scale [35].

---

10. How to invest and the outlook

Two ways in.

• Public markets. No pure play exists. The closest exposures are NextEra (NEE) as the largest battery owner wrapped in a regulated utility, AES as a higher-yield renewables-and-storage developer, Vistra (VST) as a merchant/retail power name with battery upside and tail risk, and Fluence (FLNC) as a volatile, unprofitable equipment integrator [11][12][13][14][36]. Regulated owners trade as dividend-growth utilities (bond-like, rate-base-driven; NEE yields ~2.8%, AES ~4.8%); merchant names trade as higher-beta power-price / data-center plays (VST yields ~0.6%, buyback-heavy); the integrator is a growth/momentum bet on deployment volume [36]. ETFs (GRID, IGF) offer diversified, indirect access [37].
• Private markets. More direct exposure comes from buying an operating project (a contracted special-purpose vehicle), funding development, backing a developer platform, co-investing alongside an infrastructure fund, lending project debt, buying or transferring §48E tax credits, or taking municipal-bond exposure to public-power and pumped-storage programs [25]. The decisive variable is the quality and duration of the cash flow attached to each megawatt — a tolled or long-PPA project underwrites like infrastructure; a short-contract merchant project underwrites like a commodity trade.

Near-term drivers (forward-looking). Deployment is set to stay rapid: developers plan ~24 GW of batteries for 2026, with Texas ~53% (12.9 GW), California ~3.4 GW, and Arizona ~3.2 GW — about 80% of the planned total in three states [6]. Data-center load and renewables firming are durable multi-year tailwinds that also lift capacity-market revenue [23][33]. Revenue is maturing away from early, lucrative ancillary services toward arbitrage, capacity, and contracted / tolling structures, and lenders increasingly demand contracted floors [22][24]. Storage's ITC survival through a 2033 construction start is a real de-risking versus wind and solar, but FEOC and tariff friction raise near-term costs and force supply-chain reshoring [26][27]. The frontier is long-duration storage; a genuine cost breakthrough would widen the market from intraday shifting to multi-day firming [35].

Judgment: NAICS 221118 is a statistically tiny, definitionally quirky Census box that stands in for one of the most important growth stories in U.S. energy infrastructure. Treat the opportunity as a barbell — regulated rate-base storage (utility owners, bond-like) at one end, merchant battery platforms (contracted-but-volatile, tax-credit-levered) at the other — with equipment integrators as a separate, higher-beta bet on deployment volume.

---

Sources

1. U.S. Census Bureau, 2022 NAICS Manual — 221118 Other Electric Power Generation (definition; adjacent/excluded codes). census.gov/naics
2. U.S. Census Bureau, County Business Patterns 2023 (national file, NAICS 221118: 123 establishments; 1,411 employees; $227.7M annual payroll; 2021–2023 trend). census.gov/programs-surveys/cbp.html
3. U.S. Census Bureau, 2022 Economic Census — Industry Statistics & Concentration, NAICS 221118 (46 firms; 52 establishments; ~$41.5M receipts; CR4 50.5%, CR8 70.9%, CR20 94.3%; HHI 881.4; ~$146B all electric generation). data.census.gov, table EC2200BASIC.
4. U.S. Small Business Administration, Table of Small Business Size Standards, eff. Mar. 17, 2023 (NAICS 221118 = 650 employees). sba.gov
5. U.S. Energy Information Administration, "U.S. battery capacity increased 66% in 2024," Today in Energy, Mar. 2025 (>26 GW; +10.4 GW; ~2% of 1,230 GW). eia.gov/todayinenergy/detail.php?id=64705
6. U.S. Energy Information Administration, "New U.S. electric generating capacity expected to reach a record high in 2026," 2026 (15 GW added 2025; 24 GW planned 2026; TX 12.9 GW/53%, CA 3.4, AZ 3.2). eia.gov/todayinenergy/detail.php?id=67205
7. U.S. Department of Energy, U.S. Hydropower Market Report — 2023 Edition (43 pumped-storage plants; ~22 GW; ~553 GWh; ~12-hour median duration). energy.gov
8. U.S. Energy Information Administration, Construction Cost Data for Electric Generators Installed in 2024 ($1,469/kW; ~$16.3B; ~11 GW). eia.gov/electricity/generatorcosts
9. National Renewable Energy Laboratory, 2024 Annual Technology Baseline — Utility-Scale Battery Storage (~$334/kWh 4-hour; 85% round-trip efficiency; 15-year life). atb.nrel.gov
10. U.S. Energy Information Administration, "Utility-scale batteries are more commonly used for price arbitrage," 2025 (batteries a secondary source; arbitrage 41%, frequency regulation 24%). eia.gov/todayinenergy/detail.php?id=66164
11. NextEra Energy, Form 10-K FY2025 (NextEra Energy Resources ~5,177 MW battery; largest U.S. owner; FPL rate base). sec.gov
12. AES Corp., Form 10-K FY2025 (AES Clean Energy ~10,961 MW U.S. renewables; 550 MW storage under construction; 46 GW pipeline). sec.gov
13. Vistra Corp., Form 10-K FY2025 (Moss Landing fire, Jan. 2025; ~$400M write-off; ~$155M impairment). sec.gov
14. Fluence Energy, Form 10-K FY2025 (6.8 GW deployed; 9.1 GW backlog; 5.6 GW under O&M; ~$41.8M trailing net loss). sec.gov
15. Tamarindo, "Who are the top 5 US storage companies by operating capacity?" (NextEra ~3× nearest rival). tamarindo.global
16. Plus Power, Company & Project Portfolio (1,650 MW / 4,150 MWh; 11+ GW in development). pluspower.com
17. Jupiter Power, Solutions & Portfolio (5,000+ MWh; 12+ GW pipeline; backed by Global Infrastructure Partners / BlackRock). jupiterpower.io
18. LS Power, REV Renewables launch & company history (2.4 GW; 615 MW California batteries; 1,620 MW PJM pumped storage; ~3 GW acquired 2025). lspower.com
19. Brookfield Renewable Partners, Form 20-F FY2024 (Neoen ~53% stake, Dec. 2024; ~8 GW). sec.gov
20. Tennessee Valley Authority (Raccoon Mountain 1,652 MW); New York Power Authority (Blenheim-Gilboa 1,160 MW); Oglethorpe Power (Rocky Mountain 1,095 MW). tva.com; dps.ny.gov; opc.com
21. California Public Utilities Commission, Cost of Capital — Authorized 2026–2028 Returns (PG&E 9.98%, SCE 10.03%, SDG&E 9.93% ROE). cpuc.ca.gov
22. Modo Energy / pv-magazine USA, ERCOT battery storage revenues (H1 2024 ~$70/kW; Sept. 2024 arbitrage 58%; ancillary −90% from peak by late 2025). modoenergy.com; pv-magazine-usa.com
23. PJM Interconnection, 2025/2026 Base Residual Auction ($269.92/MW-day; Dominion $444.26); Monitoring Analytics / IEEFA (data centers ~63% of increase, ~$9.3B). pjm.com; ieefa.org
24. Modo Energy, "ERCOT battery energy storage tolling agreements" (contracted dispatch structure). modoenergy.com
25. Internal Revenue Service, Clean Electricity Investment Credit (§48E) (6% base; up to 30%; +10pp domestic content; +10pp energy community; transferability; elective pay; standalone storage eligible; storage generally ineligible for PTC). irs.gov
26. Kirkland & Ellis / Internal Revenue Service, One Big Beautiful Bill Act analyses (storage §48E/§45Y eligible if construction begins before 2033; FEOC / material-assistance rules for post-2025 starts). kirkland.com; irs.gov
27. Benchmark Mineral Intelligence / TWAICE, Section 301 tariffs on Chinese storage batteries (rising toward ~58–82% by 2026; Chinese LFP ~$73/kWh, 2024). benchmarkminerals.com; twaice.com
28. Federal Energy Regulatory Commission, Order No. 841 — Electric Storage Participation in RTO/ISO Markets, 2018. ferc.gov
29. Federal Energy Regulatory Commission, Order No. 2023 (interconnection cluster reform) and Order No. 1920 (long-term transmission planning). ferc.gov
30. Lawrence Berkeley National Laboratory, Queued Up (2024 Edition) (end-2023 ~1,570 GW generation + ~1,030 GW storage; ~5-year median; ~2,300 GW total by end-2024, ~890 GW storage; 2024 storage entries −32%). emp.lbl.gov/queues
31. U.S. Environmental Protection Agency, Battery Energy Storage Systems: Safe Installation and Incident Response, 2025 (fire hazards; NFPA 855; UL 9540/9540A). epa.gov
32. U.S. Nuclear Regulatory Commission, Operating Reactors — What We Regulate (nuclear only; not grid storage). nrc.gov
33. U.S. Energy Information Administration, demand and generation outlooks (~1.7%/yr 2020–25 vs 0.1% 2005–19; ~1% 2026, ~3% 2027; wind+solar 18%→21%). eia.gov
34. California Public Utilities Commission, Energy Storage Program (AB 2514, 1,325 MW); NYSERDA, New York 6-GW Storage Roadmap (2030). cpuc.ca.gov; nyserda.ny.gov
35. Utility Dive / Renewable Energy World, Form Energy iron-air (100-hour; <$20/kWh target). utilitydive.com
36. StockAnalysis / S&P Global Market Intelligence, NEE, AES, VST, FLNC statistics & valuation, July 17, 2026 close. stockanalysis.com
37. First Trust (GRID); Amplify ETFs (BATT); iShares (IGF) — fund fact sheets, 2026. ftportfolios.com; amplifyetfs.com; ishares.com
38. U.S. Energy Information Administration, Electric Power Annual 2024, Table 4.2.A — Existing Net Summer Capacity by Producer Type (battery-heavy residual ~85% IPP / ~10% electric utility; pumped storage ~84% electric utility). eia.gov/electricity/annual