Industrial Gas Manufacturing (U.S.) — NAICS 325120

A Histometrics industry primer for public-market and private investors.

1. Overview

Industrial gas manufacturing takes ordinary air, natural gas, and a few chemical by-product streams and turns them into purified, packaged commodities: oxygen, nitrogen, argon, carbon dioxide, hydrogen, helium, and specialty mixtures. These gases are inputs to almost everything else — steel, chips, hospitals, food, welding, chemicals, and increasingly clean-energy projects.

For investors the appeal is that this is a boring business that behaves like a toll road. Gas is heavy, cheap per pound, and uneconomic to ship far, so each region is effectively served by whoever owns the local plant. Combine that with 10-to-20-year "take-or-pay" supply contracts (the customer pays for a minimum volume whether or not it uses it) and you get utility-like, recurring revenue with real pricing power [4][5].

There are two clean ways in. The public route is a short list of very large listed producers — most accessibly Linde and Air Products in the U.S. [5][10][14]. The private route is not about building an air-separation plant (that takes hundreds of millions of dollars and a global balance sheet); it is about owning the fragmented tail — regional cylinder-and-welding-gas distributors, cryogenic equipment makers, or merchant carbon-dioxide and specialty niches — which private equity has been actively rolling up [9][19].

2. What it is and how it's structured

NAICS (North American Industry Classification System) code 325120 covers establishments that manufacture industrial organic and inorganic gases in compressed, liquid, and solid form — including oxygen, nitrogen, argon, carbon dioxide, hydrogen, nitrous oxide, acetylene, and dry ice. Chlorine is classified elsewhere, as are ethane and butane made from refined petroleum [1][4].

Production runs on three core technologies: cryogenic air separation units (ASUs), which chill air until oxygen, nitrogen, and argon separate out by boiling point (this dominates by volume); steam methane reforming (SMR), which splits natural gas into hydrogen; and by-product capture of carbon dioxide from ammonia and ethanol plants [4][21]. Large plants may simultaneously supply pipeline customers and make liquid product for the merchant market — plant utilization, product mix, and local demand density matter more than the manufacturing cost of any single molecule [5]. Gas reaches customers three ways, and each is a different business:

• On-site / tonnage — a dedicated plant built next to one huge consumer (a steel mill, refinery, or chemical complex) or a pipeline feeding an industrial cluster. Linde reports on-site contracts typically run 10–20 years with minimum purchases and price escalators; Air Products describes large on-site contracts of 15–20 years, small on-site contracts of 10–15 years. The most capital-heavy and most annuity-like channel [5][10].
• Merchant / bulk liquid — cryogenic liquid trucked to mid-size users and stored in leased on-site tanks. Merchant contracts generally run 3–7 years. Route density and delivery cost drive the economics [5][10].
• Packaged / cylinder — high-purity, specialty, and welding gases in cylinders, sold (often with welding hardware) through distributor branches and welding-supply stores. Contracts typically run 1–3 years or through purchase orders. The most fragmented and highest-margin-per-unit channel [5][10].

Transport economics create local rather than purely national markets. Liquid oxygen and nitrogen have relatively short economical delivery radii because cryogenic trucking is expensive; argon travels farther, while helium can move internationally in specialized containers. Pipeline systems, spare capacity, cylinder fleets, and dense delivery routes therefore constitute a real moat — the product may be chemically fungible, but dependable delivered supply is not.

What 325120 excludes (and why it matters for the numbers below): other basic inorganic chemicals sit in NAICS 325180; extracting and processing natural gas (from which helium is often recovered) sits in NAICS 211130; and — importantly — the wholesale distribution of cylinder gas and welding supplies is classified under wholesale trade (NAICS 423/424), not manufacturing. The machinery itself (ASUs, cryogenic tanks) is machinery manufacturing (NAICS 333). So a large slice of the everyday "industrial gas economy" is counted outside 325120.

Ownership is unusually concentrated and unusually foreign. The U.S. leaders are subsidiaries or units of a handful of global champions: Linde (Irish-domiciled, run from Connecticut), Air Liquide (French), Air Products (U.S.), Nippon Sanso/Matheson (Japanese), and Messer (German family-owned) [6][17][18].

3. How big it is

Our federal figures for NAICS 325120:

Two things stand out. First, this is an extraordinarily capital-intensive, automated industry: about $17.3 billion of receipts against only ~13,800 workers is roughly $1.25 million of revenue per employee — machines and long-lived plants do the work, not headcount [1][2]. Second, 546 plants owned by just 93 firms means the big players each run many sites; the top four already hold two-thirds of revenue [1][3].

The undercount caveat. Read the $17.3 billion as gas manufactured in the U.S., not the size of the U.S. industrial-gas economy. Cylinder distribution, welding-supply retail, and cylinder rental are booked under wholesale/retail trade, not 325120; captive gas made inside a customer's own plant is not sold at all; and the multinational parents book much of the corporate profit outside this narrow U.S. manufacturing code. Industry trackers that add distribution put "U.S. industrial gas production" nearer $18 billion and rising [20]. Globally the market is roughly $110–120 billion, with the top four suppliers holding over 80% [6][7].

4. The investable universe

Unlike most industries, the pure-play public options are few — and two of the four global leaders are foreign-listed.

Major private / other owners in the U.S.:

• Messer Americas — U.S. bulk and cylinder business built from divested Linde/Praxair assets; the Messer parent is German family-owned, with sovereign-wealth fund GIC as a minority partner. Messer describes itself as the world's largest privately held industrial-gas specialist [18][22].
• Airgas / Matheson — U.S. operating arms of listed foreign parents (Air Liquide, Nippon Sanso). Air Liquide describes Airgas as its wholly owned U.S. industrial-, medical-, and specialty-gas supplier [9][17][23].
• Regional independents — hundreds of small packaged-gas and welding-supply distributors, steadily being consolidated; Linde bought one of the largest, nexAir (~$400M sales), in 2023 [19].
• Equipment adjacency — cryogenic-equipment makers such as Chart Industries (GTLS) supply the tanks and ASUs but are not gas producers; a way to play the build-out without owning molecules.

Bottom line: if you want a listed operator, LIN and APD are the accessible U.S. large caps; broad materials/industrials index funds hold both. Everything smaller tends to be private. Note that neither is a pure U.S. NAICS 325120 security — both are global, and Linde also owns an engineering business.

5. How the money works

Owners make money less on the gas itself than on locked-in volume, local density, and return on capital. This is a capital-, electricity-, and logistics-intensive infrastructure business, not merely chemical commodity resale. The economics that matter:

• Take-or-pay contracts. On-site deals run 10–20 years with minimum-purchase and cost pass-through terms, so revenue holds up even in downturns. Linde alone estimates roughly $59 billion of future minimum-purchase and plant-sale consideration on its books, plus a ~$7.1 billion backlog of projects under construction [5].
• Local pricing power. Because low-value gas can't travel far, each metro is effectively an oligopoly or duopoly. National concentration looks only "moderate" (HHI ~1,299) [3], but local concentration is much higher — that's the real moat.
• Segment profitability. Linde's 2025 Americas segment reported $15.2 billion of sales and a 31.2% operating margin; Air Products' fiscal-2025 Americas segment reported $5.1 billion of sales and a 29.6% margin. These margins reflect mature networks and contract portfolios, not the economics of a greenfield plant or an average independent distributor [5][10].
• Channel mix. Cylinder/packaged gas carries the fattest per-unit margins; bulk liquid is a route-density game; on-site is thin-margin but ultra-stable annuity revenue. Linde's 2024 Americas split — roughly $6.4B packaged, $4.6B merchant, $3.2B on-site — shows the balance [5].
• The main cost. Linde calls energy its single largest cost item: electricity powers air separation, natural gas is both fuel and hydrogen feedstock, and diesel supports distribution. Pricing formulas, surcharges, tolling, and cost pass-through provisions reduce exposure but do not eliminate basis, timing, or unrecovered-cost risk [5].
• Energy pass-through distorts reported growth. Air Products' fiscal-2025 consolidated sales fell 1%: volume contributed negative 4%, price positive 1%, and energy pass-through positive 2%. Its gross margin fell 110 basis points, partly because pass-through revenue carries little incremental profit. Investors should examine volume, price, and pass-through separately rather than treating reported sales growth as demand growth [10].

Because capacity is expensive and slow to build, the industry rarely floods itself — a big reason returns stay steady across cycles.

6. What drives demand

Volume tracks industrial activity, with several secular layers on top:

• Heavy industry and the manufacturing cycle — oxygen for steelmaking, hydrogen for refining and chemicals, argon/CO2/acetylene for metal fabrication and welding. This is the cyclical core, tied to manufacturing output [4][5].
• Electronics and semiconductors — ultra-high-purity nitrogen, hydrogen, argon, and numerous specialty "electronic" gases for chip fabrication; a strong secular grower, amplified by U.S. fab construction. Data-center, solar, battery, and other advanced-manufacturing investment can generate adjacent demand [4].
• Healthcare — medical oxygen and respiratory gases; steady, defensive demand that spiked during COVID-19 and keeps growing with an aging population [15].
• Food and beverage — carbon dioxide for carbonation and freezing, nitrogen for packaging [21].
• Clean energy / hydrogen — decarbonization is the biggest forward-looking driver, via low-carbon ("blue") and electrolytic ("green") hydrogen and carbon capture. The Department of Energy reports the United States currently produces approximately 10 million metric tons of hydrogen annually, used mainly in refining and ammonia, with possible newer applications in steel, ports, heavy transport, and data centers. Whether suppliers earn attractive returns depends on power and natural-gas economics, carbon capture, tax eligibility, customer commitments, and construction discipline — not simply projected hydrogen demand [10][13][24].

7. Regulation

Industrial gas sits under several regimes at once, and compliance is a genuine barrier to entry:

• FDA (medical gases). Medical oxygen, nitrogen, nitrous oxide, carbon dioxide, helium, and medical air are regulated as drugs by the FDA's Center for Drug Evaluation and Research, with certification, manufacturing, labeling, and safety-reporting obligations [15][25].
• DOT / PHMSA (transport). The Pipeline and Hazardous Materials Safety Administration enforces the Hazardous Materials Regulations (49 CFR) governing cylinder specifications, periodic testing, markings, and the trucking of compressed and cryogenic gas [16][26].
• OSHA (safety). OSHA identifies oxygen displacement, fire, explosion, toxicity, cryogenic injury, and high pressure as core compressed-gas hazards. Its rules govern cylinder inspection, storage, and handling, with gas-specific standards for hydrogen, oxygen, acetylene, and nitrous oxide [27][28].
• EPA and energy policy. Hydrogen producers can face EPA greenhouse-gas reporting, while facilities holding regulated flammable or toxic gases above applicable thresholds can fall under Risk Management Program requirements. Air permits govern hydrogen and CO2 operations; federal clean-energy tax credits (for clean hydrogen and carbon capture) are a swing factor in whether big decarbonization projects pencil out — and are themselves subject to political change [13][29][30].
• Antitrust. The FTC's treatment of the Praxair–Linde merger required divestitures in multiple bulk oxygen and nitrogen geographic markets and in the national bulk-argon market, finding that the transaction would otherwise reduce competition materially — confirming both high concentration and the product-by-product, geography-by-geography nature of competition [8][18].

8. Competitive dynamics and consolidation

This is a textbook global oligopoly, and it got that way through a rapid consolidation wave:

• 2016 — Air Liquide acquired U.S. packaged-gas leader Airgas for about $13.4 billion, making it #1 in North America [9].
• 2018 — Praxair and Linde AG merged (~$70+ billion) to create today's Linde plc, the world's largest producer, cutting the majors from four to three plus regionals. The FTC required divestitures in multiple geographic and product markets on both sides of the Atlantic [8].
• 2019 — Linde AG's divested Americas assets (with private-equity firm CVC) became Messer Americas; Messer later took full ownership, with GIC as a partner [18].
• Ongoing — the majors keep rolling up independent distributors (e.g., Linde's 2023 purchase of nexAir) to add cylinder density [19].

Competition today is less about price wars and more about winning greenfield on-site anchor contracts and building regional route density. Once a plant and its pipeline are in place, the incumbent is very hard to dislodge. Merchant supply can also be displaced by customer-owned pressure-swing adsorption, membrane, or cryogenic generation — substitution usually changes who owns the plant rather than eliminating the customer's need for gas.

9. Risks

• Cyclicality. Merchant and packaged volumes fall with industrial production, steel output, and manufacturing PMIs; on-site take-or-pay contracts cushion but don't eliminate the swing. Minimum charges and long on-site contracts damp the effect of an industrial downturn, although a permanently closed customer can strand capital [4][5].
• Energy costs. Power and natural gas are the dominant inputs; pass-through lags can compress margins, as Europe's energy shock demonstrated [5].
• Big-project execution and policy risk. Clean-hydrogen mega-projects carry large cost and schedule risk. Air Products' fiscal-2025 project review produced approximately $3.6 billion of project-exit costs, principally noncash asset write-downs and contract-termination obligations — a caution flag on how quickly these economics can turn. That episode illustrates the difference between the relatively defensive installed gas franchise and speculative, very large energy-transition developments [10][13].
• CO2 supply fragility. Merchant carbon dioxide is largely a by-product of ammonia and ethanol plants; their outages cause periodic shortages that hit beverage and food customers, and U.S. merchant CO2 supply is expected to tighten over the coming decade. USDA has documented disruptions caused by ammonia- and ethanol-plant shutdowns, source contamination, driver shortages, and seasonality between fertilizer production and summer beverage demand [21][31].
• Helium and rare-gas scarcity. Helium is geologically constrained and prone to shortages, an ongoing supply headache. USGS estimates that 2025 U.S. Grade-A and gaseous helium sales were 81 million cubic meters, valued at $970 million. U.S. net import reliance is 52% for neon, 93% for krypton, and 98% for xenon — semiconductor, medical-imaging, aerospace, and research demand therefore intersects with concentrated or geopolitically exposed supply [32].
• Capital intensity and interest rates. Multi-hundred-million-dollar plants with multi-year build times make the group sensitive to financing costs and to a single anchor customer's credit.

10. How to invest and the outlook

Public route. For most investors this means Linde (LIN) or Air Products (APD) — the two U.S.-listed large caps, both long-standing Dividend Aristocrats [14]. International exposure comes via Air Liquide (Paris, or the AIQUY ADR) and Nippon Sanso (Tokyo) [6][9][17]. Note the trade-off: these are prized for utility-like stability plus growth, so they usually trade at premium earnings multiples and pay modest dividend yields — you are paying up for durability, and reserve valuation and yield judgments for entry timing.

Private route. Direct ownership of gas plants is effectively closed to all but the global majors. Realistic private entry points are (a) regional packaged-gas and welding-supply distributors — a proven roll-up game with private-equity precedent; (b) cryogenic equipment and engineering suppliers; (c) merchant CO2 and specialty-gas niches; and (d) higher-risk project equity in clean-hydrogen and carbon-capture developments. The decisive diligence variables are route density, proximity to reliable supply, customer and cylinder retention, contract duration and escalation, minimum commitments, utilization, maintenance capital, safety history, source redundancy, and return on invested capital [9][13][19].

Outlook (forward-looking judgment). The base business should keep compounding at low-to-mid single-digit volume growth tied to industrial production, with pricing discipline driving margin gains — the pattern of the last decade. Semiconductors, healthcare, and U.S. reshoring are secular tailwinds. The wildcard is clean hydrogen and carbon capture: potentially a large new demand layer, but one whose returns hinge on project execution and on federal incentives that are not guaranteed. Air Products' pullback and $3.6 billion in project-exit charges is a reminder to treat the decarbonization upside as optionality, not a certainty. Near-term, watch the manufacturing cycle, power costs, and whether clean-hydrogen project economics firm up.

---

Sources

1. U.S. Census Bureau. County Business Patterns, 2023 — NAICS 325120 (establishments, employment, payroll). 2023. https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau. Annual Integrated Economic Survey, 2023 — NAICS 325120 (receipts). 2023. https://data.census.gov/table?codeset=naics~325120&g=010XX00US
3. U.S. Census Bureau. 2022 Economic Census — Concentration ratios and receipts, NAICS 325120 (firms, CR4/CR8/CR20/CR50, HHI). 2022. https://www.census.gov/programs-surveys/economic-census.html
4. Wikipedia. Industrial gas. 2025. https://en.wikipedia.org/wiki/Industrial_gas
5. Linde plc. Form 10-K, FY2025 (Americas segment, take-or-pay, minimum-purchase and project backlog, channel mix, energy costs). 2026. https://www.sec.gov/Archives/edgar/data/1707925/000162828026011430/lin-20251231.htm
6. Statista. Global industrial gas industry (market size; top-four share; company revenues). 2024. https://www.statista.com/topics/9233/global-industrial-gas-industry/
7. Straits Research. Industrial Gases Market — size and forecast. 2024. https://straitsresearch.com/report/industrial-gases-market
8. FTC. FTC Requires International Industrial-Gas Suppliers Praxair Inc., Linde AG to Divest Assets (merger settlement). 2018. https://www.ftc.gov/news-events/news/press-releases/2018/10/ftc-requires-international-industrial-gas-suppliers-praxair-inc-linde-ag-divest-assets-nine
9. CVC. Messer and CVC Fund VII acquire assets in the Americas from Linde (Airgas and Messer Americas background). 2018. https://www.cvc.com/media/news/2018/2018-07-16-messer-and-cvc-fund-vii-acquire-assets-in-the-americas-from-linde/
10. Air Products & Chemicals, Inc. Form 10-K, FY2025 (revenue, segments, employees, volume/price/pass-through breakdown, project-exit costs). 2025. https://www.sec.gov/Archives/edgar/data/2969/000000296925000055/apd-20250930.htm
11. U.S. Small Business Administration. Table of Small Business Size Standards (NAICS 325120 — 1,200 employees). 2023. https://www.sba.gov/document/support-table-size-standards
12. Air Products & Chemicals, Inc. Air Products Increases Quarterly Dividend to $1.81 Per Share (44th consecutive year). 2026. https://www.airproducts.com/company/news-center/2026/01/0127-air-products-quarterly-dividend
13. Air Products & Chemicals, Inc. Air Products Will Not Proceed with Louisiana Clean Energy Complex; NEOM Green Hydrogen update. 2026. https://www.prnewswire.com/news-releases/air-products-will-not-proceed-with-louisiana-clean-energy-lcec-project-302814443.html
14. Sure Dividend. Dividend Aristocrats in Focus: Linde plc (Dividend Aristocrat 33 years; ~$208.5B market cap); Air Products & Chemicals. 2025. https://www.suredividend.com/dividend-aristocrats-lin/
15. AgencyIQ (POLITICO). FDA carves out a regulatory niche for medical gases (medical gases regulated as drugs by CDER). 2025. https://www.agencyiq.com/blog/in-a-new-final-rule-fda-carves-out-a-regulatory-niche-for-medical-gases/
16. Welding & Gases Today / PHMSA. FDA, DOT, DHS, PHMSA, TSA — inside the regulatory agencies (49 CFR cylinder and transport rules). 2008 (regs current). http://www.weldingandgasestoday.org/index.php/2008/09/fda-dot-dhs-phmsa-tsa-inside-the-regulatory-agencies
17. Wikipedia. Matheson (compressed gas & equipment) — Nippon Sanso Holdings U.S. arm. 2025. https://en.wikipedia.org/wiki/Matheson_(compressed_gas_%26_equipment)
18. Messer Americas / CVC. Messer Completes Acquisition of Joint Venture; GIC New Strategic Partner (Messer Americas ownership). 2023. https://www.messeramericas.com/news/messer-completes-acquisition-gic-new-partner
19. Linde. Linde Buys Out Stake in One of the Largest U.S. Independent Packaged Gas Distributors (nexAir, ~$400M sales). 2023. https://www.linde.com/news-and-media/2023/linde-buys-out-stake-in-one-of-the-largest-u-s-independent-packaged-gas-distributors
20. IBISWorld. Industrial Gas Production in the US — industry size (~$18.4B, 2026). 2026. https://www.ibisworld.com/united-states/industry/industrial-gas-production/459/
21. gasworld / U.S. EPA / Compressed Gas Association. 2025 US merchant CO2 report; Carbon Dioxide Supply Chain Profile (CO2 as ammonia/ethanol by-product; shortage risk). 2023–2025. https://www.gasworld.com/feature/2025-us-merchant-co2-report/2154586.article/
22. Messer Americas. Financial Year 2025: Solid Business Performance. 2026. https://www.messeramericas.com/news/financial-year-2025-solid-business-performance
23. Air Liquide. Airgas: The Backbone of US Energy Needs. 2025. https://www.airliquide.com/stories/integrated-annual-report/airgas-backbone-us-energy-needs
24. U.S. Department of Energy. Hydrogen Overview (U.S. production ~10 million metric tons annually). 2025. https://www.energy.gov/cmei/fuels/hydrogen
25. FDA. Current Good Manufacturing Practice, Certification, Postmarketing Safety Reporting, and Labeling for Medical Gases (final rule). 2025. https://www.fda.gov/about-fda/economic-impact-analyses-fda-regulations/current-good-manufacturing-practice-certification-postmarketing-safety-reporting-and-labeling-0
26. PHMSA. Cylinder Shipment Requirements (49 CFR 173.301). https://www.phmsa.dot.gov/regulations/title49/section/173301
27. OSHA. Compressed Gas Equipment Overview. https://www.osha.gov/compressed-gas-equipment
28. OSHA. 29 CFR 1910.101 — Compressed Gases (General Requirements). https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.101
29. EPA. Subpart P Information Sheet — Hydrogen Production (GHG reporting). https://www.epa.gov/ghgreporting/subpart-p-information-sheet
30. EPA. List of Regulated Substances Under the Risk Management Program. https://www.epa.gov/rmp/list-regulated-substances-under-risk-management-program
31. USDA. Carbon Dioxide — Crops (technical report on CO2 supply disruptions). https://www.ams.usda.gov/sites/default/files/media/CarbonDioxide_Crops.pdf
32. USGS. Mineral Commodity Summaries 2026 (helium sales $970M; neon/krypton/xenon import reliance). 2026. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf