Carbon Monoxide Market

Air Products terminated a project in Texas for the production of carbon monoxide on 24 February 2025, citing unfavourable project economics in its filing with the U.S. Securities and Exchange Commission, and that single cancellation explains how this gas really trades. The carbon monoxide market covers CO gas made from natural gas, coal or residues and delivered to chemical, metal and electronics users, whether piped from a cold box next door or shipped in cylinders. Douglas Insights sizes it at USD 4.64 billion in 2025, built as 14.62 million tonnes at an average realised USD 317.6 per tonne, and projects USD 7.01 billion by 2035 on a 4.21% revenue CAGR. Volume adds 2.62% a year and price 1.55%. The study belongs to our chemicals and materials coverage, and every step of the build follows the research methodology page.

What does the carbon monoxide market include, from syngas cold boxes to cylinder gas?

Carbon monoxide in this study is CO valued at the point of use: 14.62 million tonnes in 2025 across captive syngas, pipeline and on-site supply, and cylinders and bulk. The scope runs from carbonylation feed for acetic acid to 99.99% electronic grade gas, and it excludes CO burned as fuel inside steel mills.

Three supply modes carry the tonnes: captive syngas made inside an acetyl or isocyanate complex, pipeline and on-site supply from a gas company under long take-or-pay contracts, and cylinders and bulk for laboratories, carbonyl makers and fabs. Grades run from industrial grade for carbonylation to high-purity grade and electronic grade.

Steel-mill offgas holds large amounts of CO, but almost all of it is burned for heat, so it stays outside the 14.62 million tonne base.

Who supplies carbon monoxide over the fence, and how concentrated are the cold-box owners?

Douglas Insights estimates the top three gas companies (Linde, Air Liquide and Air Products) hold 41.6% of 2025 carbon monoxide value, about USD 1.93 billion. The rest is made captively inside acetyl and isocyanate complexes in China, Europe and the Gulf.

Linde lists carbon monoxide among its process gases and states that it can be produced by either steam methane reforming (SMR) or autothermal reforming (ATR) of natural gas or other feedstock, according to its annual report on Form 10-K. Its total sales reached USD 33.99 billion in 2025, so carbon monoxide is a small line inside a very large on-site business. Linde also notes that for carbon monoxide the raw materials are largely purchased from outside sources, which puts gas cost straight into contract formulas.

Air Liquide sells carbon monoxide as a packaged gas and classifies it as Repr. 1A and Acute Tox. 3 on its own safety data sheet. Air Products runs a Gulf Coast carbon monoxide pipeline network in Texas, yet the Texas project exit shows it now walks away from CO plants whose returns do not clear its hurdle. Celanese, a major buyer on the demand side, generally purchases carbon monoxide under long-term contracts and warns that some plants have single sources of supply for raw materials such as carbon monoxide.

Methanex supplies methanol, the other half of the acetic acid molecule, and Twelve makes fuel from captured CO2 with electrolysers. Douglas Insights puts Linde at 17.3% of 2025 carbon monoxide value, Air Liquide at 13.9% and Air Products at 10.4%.

Which carbonylation segment makes the money: acetic acid, phosgene or formic acid?

Acetic acid and acetyls lead carbon monoxide demand with 61.7% of 2025 value, or USD 2.86 billion. Methanol carbonylation consumes roughly 0.87 tonnes of CO per tonne of methanol and runs at very large single-train scale, so one acetyl site can absorb more CO than a whole country of cylinder users.

Acetic acid and acetyls hold USD 2.86 billion because acetic acid takes about 11% of world methanol demand, according to the Methanex July 2026 investor presentation, which puts 2026 methanol demand near 100 million tonnes. That is 11.0 million tonnes of methanol, and at the 28 to 32 molecular weight ratio it pulls about 9.62 million tonnes of carbon monoxide, 65.8% of our volume base. The value share is lower than the volume share because Chinese coal-based acetyl plants make CO at the cheapest cost in the market.

Phosgene for isocyanates and polycarbonate is worth USD 1.09 billion, 23.4% of value. Every tonne of phosgene needs 0.28 tonnes of CO, and phosgene sits upstream of isocyanates and polycarbonate resin, so this segment tracks the Polyurethane Market closely. Purity rules here are tighter, and that lifts the realised price per tonne.

Formic acid and other carbonylation adds USD 455.0 million, 9.8% of value, covering formic acid, dimethylformamide, propionic acid and methyl formate. Metals and electronics is the smallest segment at USD 236.8 million and 5.1%, but it grows fastest at 6.94% a year, because nickel carbonyl refining and electronic grade CO for semiconductor etch command prices many times the pipeline average.

Why is carbon monoxide demand still rising with acetic acid and isocyanate capacity?

Carbon monoxide volume grows 2.62% a year in our base case, from 14.62 million tonnes in 2025 to 18.94 million tonnes in 2035. Four drivers supply those 2.62 points: acetic acid, phosgene chains, other carbonylation and metals with electronics, each tied to a named downstream product rather than to general industrial output.

Acetic acid capacity is the largest driver, worth 1.31 points of the 2.62-point volume leg. Acetic acid demand follows vinyl acetate monomer, purified terephthalic acid and acetate esters, and each new carbonylation train is built with a dedicated CO source. Celanese states that carbon monoxide is typically purpose-made in close proximity to the acetic acid unit, as its 10-K for 2025 sets out, so every acetyl expansion books CO volume on the day it starts. Douglas Insights expects acetyls to add 2.10 million tonnes of CO demand between 2025 and 2035, about half of the 4.32 million tonne total.

Isocyanate and polycarbonate growth contributes 0.74 points. Methylene diphenyl diisocyanate (MDI) goes into rigid insulation foam and toluene diisocyanate (TDI) into flexible foam, and both are made through phosgene, which is generated on site from CO and chlorine rather than shipped. Segment value rises from USD 1.09 billion to USD 1.70 billion at 4.58% a year, faster than acetyls, because new MDI trains are built at world scale.

Formic acid and other carbonylation adds 0.31 points. Formic acid is used in silage preservation, leather tanning and de-icing, while dimethylformamide serves solvents and fibres. Those chemicals need high-purity grade CO, and their volume grows in step with feed additive and coatings output at about 3.1% a year on our model. Formic acid plants are small, often 50,000 to 100,000 tonnes a year, so each one adds only a modest CO load, but several start every year.

Metals and electronics supply the last 0.26 points. Nickel refineries use CO in the carbonyl process to produce high-purity nickel powder and pellets, and chip fabs buy electronic grade CO for dielectric etch. The share is small at 5.1% of value, but the growth rate of 6.94% is the highest in the study, lifting the segment from USD 236.8 million to USD 463.2 million. Each new logic fab that adds advanced etch steps raises electronic grade CO demand without any change in chemical output. The four contributions, 1.31, 0.74, 0.31 and 0.26 points, add up to the 2.62% volume leg exactly.

What limits carbon monoxide growth: pipeline-bound plants, toxicity and coal-route overcapacity?

Three restraints remove 0.86 points from carbon monoxide volume growth, holding it at 2.62% instead of 3.48%. They are plants tied to one customer, Chinese coal-route overcapacity and the cost of handling a toxic, odourless gas outside closed chemical parks.

Single-customer economics remove 0.38 points. A CO plant only works when one buyer signs a long contract next door, and when that buyer delays, the plant dies. The Texas project that Air Products terminated is the clearest recent case. Celanese itself flags single-source supply of carbon monoxide as a risk factor, which tells suppliers that buyers want redundancy they rarely pay for.

Coal-route overcapacity in China removes 0.31 points. Chinese acetyl and formic acid plants run on coal gasification syngas, and when acetic acid margins compress the operators cut rates, which flattens CO tonnage for several quarters. Toxicity and handling costs remove the last 0.17 points: carbon monoxide carries a 50 ppm permissible exposure limit in the United States, and cylinder users outside chemical parks pay for detection, ventilation and training, as covered in our Industrial Gas Detection Market study.

Where is carbon monoxide consumed, and which region adds the most tonnes?

Asia Pacific consumes the most carbon monoxide, USD 2.45 billion or 52.8% of 2025 value. China hosts large acetic acid and MDI capacity and makes much of its CO from coal-based syngas at low cost, which keeps regional value below its tonnage share.

Asia Pacific rises from USD 2.45 billion to USD 3.88 billion at 4.71% a year, the largest absolute gain. Europe holds USD 863.7 million, 18.6% of value, but grows only 2.86% because its isocyanate and acetyl sites face high gas costs. North America accounts for USD 831.2 million and 17.9%, growing 3.72%, anchored on the Gulf Coast carbon monoxide pipeline system. The Middle East and Africa add USD 334.3 million and grow fastest at 5.38% a year, because integrated Gulf complexes pair low-cost gas with new isocyanate and acetyl trains. Latin America is the smallest region at USD 162.4 million, growing 3.16% to USD 221.7 million.

The wildcard is India. A single new Indian carbonylation project with its own CO train would add close to 0.5 million tonnes of carbon monoxide demand, about 3.4% of the 2025 world base, on our model.

How much does a tonne of carbon monoxide cost from a partial-oxidation train versus a cylinder?

Douglas Insights puts the average realised carbon monoxide price at USD 317.6 per tonne in 2025, rising to USD 370.4 by 2035 on a 1.55% price leg. The average hides a spread from about USD 190 for Chinese coal syngas to thousands of dollars for cylinder gas.

Feedstock sets the floor. The Henry Hub spot price averaged USD 3.52 per million British thermal units (MMBtu) in 2025, up 56% from 2024, according to the U.S. Energy Information Administration on 9 January 2026. Partial oxidation needs about 0.571 tonnes of methane per tonne of CO, roughly 30.0 MMBtu, so gas alone costs about USD 105.7 per tonne of CO in the U.S. before oxygen, capital charges and the hydrogen credit.

Pipeline contracts index energy to gas and the facility fee to inflation; the 1.55% price leg is mostly that fee plus a shift toward high-purity grade and electronic grade gas.

Which exposure limits and hazard rules govern carbon monoxide handling?

Carbon monoxide carries a 50 ppm 8-hour permissible exposure limit (PEL) from the Occupational Safety and Health Administration (OSHA) and a 35 ppm limit recommended by the National Institute for Occupational Safety and Health (NIOSH). Its immediately dangerous to life or health (IDLH) level is 1,200 ppm, and EU rules add a reproductive toxicity label.

The OSHA chemical data sheet for CAS 630-08-0 also lists a 200 ppm NIOSH ceiling. Under the EU Classification, Labelling and Packaging (CLP) Regulation (EC) No 1272/2008, carbon monoxide is classified as Flammable Gas 1B, Acute Tox. 3 (H331), Repr. 1A (H360D) and specific target organ toxicity after repeated exposure, category 1 (H372). It ships as UN 1016. The Repr. 1A label raises the cost of cylinder use and pushes buyers toward closed pipeline supply.

What if acetyl expansions slip: carbon monoxide scenarios to 2035?

Carbon monoxide reaches USD 7.01 billion in 2035 in the base case, USD 6.09 billion in the slower case and USD 7.95 billion in the faster case. The USD 1.86 billion spread is set mainly by acetyl and MDI start-up timing.

The slower case repeats the Texas cancellation pattern: suppliers pull marginal CO projects, acetyl operators run older plants harder, and volume grows 1.71% with price at 1.02%. The faster case adds a Gulf acetyl train and wider electronic grade demand, lifting volume to 3.48% and price to 1.97%. A single extra point of volume growth adds USD 714.2 million to the 2035 figure, so the volume leg dominates. Published forecasts run from about 4.1% to 5.5% a year, and our 4.21% sits near the lower end because we exclude fuel-use CO and value captive tonnes at cost.

Can CO2 electrolysis and electronic-grade carbon monoxide reshape supply?

Electrochemical carbon monoxide is below 0.1% of 2025 tonnage in our model. Yet it changes the siting rule: a CO plant built on CO2 and power does not need a natural gas pipeline or a coal gasifier next door, only cheap electricity and a captured CO2 stream.

Twelve opened AirPlant One in Moses Lake to make jet fuel and naphtha from captured CO2, water and renewable power. Douglas Insights models e-routes at 0.04 million tonnes of CO equivalent by 2035, too small to move the base case but enough to prove the chemistry at commercial scale. Electronic grade gas is the second shift: it is 0.9% of tonnage in 2025 yet 3.2% of value on our build, and both numbers rise as fabs add etch steps. Readers following low-carbon syngas should also see our Hydrogen Market report, since every reformer that makes CO also makes hydrogen.

Douglas Exclusive: the Carbon Monoxide Contract and Capacity Calendar

The Carbon Monoxide Contract and Capacity Calendar is a Douglas Insights model built from 9 sourced inputs and 5 modelled thresholds, 12 dated entries in all. It sets dated supply moves beside the years our forecast crosses decision marks.

The 9 sourced inputs are the two Air Products releases, the Linde and Celanese annual reports, the EIA gas price, the Air Liquide results, the Twelve opening, the Methanex demand split and the OSHA exposure data. The calendar finds a gap: the seven sourced events between February 2025 and July 2026 were exits, results or feed signals, and none was a new merchant CO plant. Our model still needs 4.32 million extra tonnes of CO by 2035. Douglas Insights reads that as a decade in which buyers, not gas companies, will finance most new carbon monoxide capacity.

What should acetyl and isocyanate buyers do about carbon monoxide supply before 2030?

Buyers of carbon monoxide face 2.02 million tonnes of extra demand by 2030 on our model, 16.64 million tonnes against 14.62 million in 2025, with few announced merchant plants to meet it. Contract strategy matters more than spot price, because almost no CO trades on a spot basis outside cylinders.

Lock redundancy early: a backup syngas connection costs less than a month of lost acetic acid output. Index energy to a gas benchmark, cap the facility fee escalator near the 1.55% price leg, and audit cylinder sites against the 50 ppm limit.

Methodology: how the model turns 14.62 million tonnes of carbon monoxide into USD 4.64 billion?

The carbon monoxide receipt is 14.62 million tonnes times USD 317.6 per tonne, equal to USD 4.64 billion in 2025. Four application segments and 5 regions reconcile to that total within USD 0.1 million, and every growth rate is a product of separate volume and price legs.

Volume comes from 4 application builds. Acetyls use 11% of about 100 million tonnes of methanol, times 28.01 over 32.04, for 9.62 million tonnes. Phosgene, formic acid and metals use the stoichiometric 0.28 tonnes of CO per tonne of phosgene and Douglas Insights capacity estimates by region. Price blends 5 supply-mode bands weighted by tonnes. Cross-checks: acetyl value of USD 2.86 billion over 9.62 million tonnes implies USD 297.8 per tonne, inside the coal and Gulf Coast bands, and our USD 4.64 billion base sits inside the USD 3.66 billion to USD 5.42 billion band of published base-year estimates. Region totals match the global row to USD 0.1 million in 2025 and 2035, and the four segment 2035 values sum to within 0.4% of USD 7.01 billion.

Sources

1. U.S. SEC / Air Products Air Products to exit three U.S.-based projects (Form 8-K exhibit 99.1) (2025)
2. U.S. SEC / Air Products Air Products fiscal 2025 results (Form 8-K exhibit 99.1) (2025)
3. U.S. SEC / Linde Linde plc Form 10-K for fiscal 2025 (2026)
4. U.S. SEC / Celanese Celanese Corporation Form 10-K for fiscal 2025 (2026)
5. Methanex Methanex investor presentation, July 2026 (2026)
6. U.S. Energy Information Administration In 2025, U.S. natural gas spot prices increased from 2024's record low (2026)
7. OSHA Carbon monoxide chemical data (2025)
8. Air Liquide Carbon monoxide safety data sheet (2024)
9. Air Liquide 2025 annual results (2026)
10. Twelve AirPlant One opens in Moses Lake (2026)