Plasma vs Oxy-Fuel vs Laser Cutting: Which to Use
Choosing between plasma, oxy-fuel, and laser is one of the highest-leverage decisions a fab shop makes, because it drives capital outlay, per-foot operating cost, edge quality, and how much secondary machining you can eliminate. Each process wins in a specific window of thickness, metallurgy, volume, and tolerance, and the wrong pick either bleeds margin on consumables or forces you to outsource work you could have kept in-house. This guide breaks down the metallurgy and physics behind each method, the real numbers on gas, consumables, and capital, and gives you a decision framework you can apply on your next quote. Local distributors and suppliers across the US stock the gases, consumables, and cylinders referenced here, and a good one will size the whole package for your duty cycle.
How Each Process Actually Cuts Metal
The three methods sever metal by fundamentally different physics, and that physics dictates what each can and cannot do. Understanding the mechanism is the fastest way to stop misapplying a process.
Oxy-fuel (oxy-acetylene, oxy-propane, or oxy-propylene) is a chemical reaction, not a melting process. A preheat flame brings the steel surface to its kindling temperature of roughly 1,600 to 1,800 degrees F, then a stream of high-purity cutting oxygen (99.5 percent or better) oxidizes the iron in an exothermic reaction. The cut only works because iron oxide (FeO) melts lower than the parent steel and the oxygen jet blows the molten slag out of the kerf. This is why oxy-fuel cuts carbon and low-alloy steel and essentially nothing else.
Plasma uses an electric arc constricted through a copper nozzle to ionize a gas (air, O2, N2, or H35) into a plasma jet at 20,000 to 30,000 degrees F. The arc melts the metal and the high-velocity gas ejects it. Because it relies on electrical conductivity and melting rather than oxidation, plasma cuts any electrically conductive metal, including stainless, aluminum, copper, brass, and titanium, in addition to carbon steel.
Laser cutting focuses a high-power beam, today almost universally a fiber laser in the 3 to 12 kW range, to a spot a few thousandths of an inch across, delivering enormous power density. An assist gas (O2 to react and add energy, or N2 to blow molten metal clear for an oxide-free edge) clears the kerf. The tiny focused spot is why laser holds tight tolerances and a narrow heat-affected zone.
Thickness Ranges and Metals: What Each Handles
Matching the process to material and thickness is where most shops go wrong. The practical operating windows overlap in the mid-range, and that overlap is exactly where cost and quality trade-offs decide the winner.
Oxy-fuel owns heavy plate. It cuts carbon steel from about 1/4 inch up to 12 inches routinely, and specialized rigs handle 20 inches or more, all with modest capital. Below roughly 3/16 inch the preheat warps and blows through thin sheet, so oxy-fuel is a poor fit for light-gauge work. It cannot cut stainless or aluminum because chromium oxide and aluminum oxide are refractory and do not support the oxidation reaction.
Plasma is the versatility champion in the thin-to-medium range. Air plasma systems handle roughly 26 gauge up to 1 inch for quality cuts, with sever capacity to 1.5 or 2 inches on larger machines. High-definition (HD) plasma at 200 to 400 amps pushes quality cuts to 1.5 to 2 inches on carbon steel. Critically, plasma is the only one of the three that cuts stainless and aluminum across a wide thickness band without exotic setup.
Laser excels on thin to medium material with unmatched precision. Modern 6 to 12 kW fiber lasers cut carbon steel up to about 1 to 1.25 inches, stainless up to roughly 1 inch with N2, and aluminum up to 3/4 to 1 inch. Below 1/2 inch laser is in its sweet spot, holding tolerances plasma and oxy-fuel cannot approach. Reflective metals like copper and brass are cuttable with fiber lasers but demand care.
| Attribute | Oxy-Fuel | Plasma | Laser (Fiber) |
|---|---|---|---|
| Carbon steel range | 1/4 in to 12+ in | 26 ga to 1.5-2 in | Foil to ~1.25 in |
| Stainless steel | No | Yes, to ~1.25 in | Yes, to ~1 in (N2) |
| Aluminum | No | Yes, to ~1.25 in | Yes, to ~0.75-1 in |
| Copper / brass | No | Yes (conductive) | Yes, with care |
| Best thickness window | 1 in and up | 3/8 to 1.5 in | Under 1/2 in |
| Kerf width | 0.030-0.090 in | 0.060-0.190 in | 0.004-0.020 in |
| Typical tolerance | +/- 0.030-0.060 in | +/- 0.015-0.045 in | +/- 0.002-0.010 in |
Cut Quality, Kerf, HAZ, and Speed
Edge quality determines your downstream labor. A dross-free, square edge you can weld or paint straight off the table is worth real money versus one that needs grinding. The ISO 9013 standard classifies thermal-cut edge quality by squareness (perpendicularity tolerance u) and mean surface roughness (Rz), and it is worth quoting your customers against it.
Laser produces the best edge of the three: kerf as narrow as 0.004 to 0.020 inch, a heat-affected zone (HAZ) often under 0.010 inch, near-square walls, and nitrogen-cut stainless edges that are oxide-free and ready to weld. Nitrogen-assist laser edges frequently need zero secondary finishing. Speed on thin material is dramatic: a 6 kW fiber will run 16 gauge mild steel at 1,000+ inches per minute.
Plasma sits in the middle. Kerf runs 0.060 to 0.190 inch and the cut has a characteristic bevel (typically 2 to 5 degrees, correctable with bevel-compensating torch heads) plus a wider HAZ than laser. HD plasma tightens this considerably. Plasma is fast in the 1/4 to 1 inch band, often outrunning oxy-fuel by 3 to 5 times on 1/2 inch plate, and modern systems cut dross-free within their optimized speed window.
Oxy-fuel gives a good, square edge on thick plate with a wide but predictable HAZ, and it is the slowest process. On 1 inch steel expect roughly 12 to 16 inches per minute versus much faster plasma. Its advantage is that edge squareness holds even at 6, 8, or 12 inches where the other two simply cannot reach. Preheat also introduces the most thermal distortion on thinner parts, another reason to keep it on heavy sections.
- Kerf: The width of material removed by the cut. Narrow kerf (laser) means tighter nesting and less waste; wide kerf (plasma, thick oxy-fuel) must be programmed into your part offsets.
- HAZ: Heat-affected zone. The band of parent metal whose microstructure changed from cutting heat. A larger HAZ can harden edges (a concern on hardenable alloys and for later machining or fatigue-critical parts).
- Dross: Re-solidified molten metal clinging to the bottom edge. Correct amperage, speed, and standoff eliminate it on plasma; N2-assist laser avoids it entirely on stainless.
- Bevel: Plasma inherently produces a slight angled edge from the swirling arc; account for it or use a beveling head. Laser and oxy-fuel run much closer to square.
Gas, Consumable, and Operating Costs
Per-foot operating cost is where these processes diverge sharply, and it is the number that decides profitability on production runs. Capital cost is a one-time hit; consumables and gas bleed every hour the machine runs.
Oxy-fuel has the lowest operating cost by far. Acetylene runs roughly 200 to 300 cubic feet in a common cylinder, and a full oxygen/acetylene setup with cutting oxygen consumes gas measured in a few dollars per hour. Tips are cheap (5 to 20 dollars) and last a long time. Watch acetylene safety: never draw above 15 PSIG per NFPA 51 and CGA guidance, because acetylene is unstable at higher pressure. Propane and propylene are cheaper fuel alternatives when preheat matters more than pierce speed.
Plasma consumables are the hidden cost. Electrodes and nozzles are the wear items, and on a 200 to 400 amp HD system a matched electrode/nozzle/swirl-ring/shield set can run 40 to 120 dollars and last anywhere from 1 to several hours of arc-on time depending on pierce count. Air plasma is cheapest to feed (shop air plus power), while O2 and N2 plasma buy better edges at higher gas cost. Budget consumable cost per foot deliberately; frequent piercing on thin nested parts burns electrodes fast.
Laser has high fixed cost but low consumable cost once running. Fiber lasers have no lamps or resonator gases to replace, so the recurring spend is assist gas and electricity. Nitrogen is the swing factor: high-pressure N2 cutting (up to 300+ PSI) on thick stainless consumes gas rapidly, and shops running volume often justify an on-site nitrogen generator or bulk liquid tank over cylinders. Oxygen-assist on carbon steel is far cheaper than N2. Optics (protective cover slides) are a minor consumable; the real cost is the electric draw of a 6 to 12 kW machine plus chiller.
Capital Cost and When to Outsource Laser
Capital is the gate that keeps most shops out of laser and sends thin, precise work to a service center. Be honest about your annual volume before you sign a lease on a machine that needs to run two shifts to pencil out.
A complete oxy-fuel outfit (torch, regulators, hoses, cart, cylinders) costs roughly 300 to 1,500 dollars for a hand rig; a CNC oxy-fuel plate table runs tens of thousands. A handheld air plasma unit (30 to 65 amp) is 800 to 3,000 dollars, a shop-grade 100 amp machine is a few thousand, and a full CNC plasma table with HD power supply, water table, and fume extraction lands anywhere from 25,000 to well over 150,000 dollars. Fiber laser cutting systems start around 250,000 to 400,000 dollars for a 3 to 6 kW machine and climb past 1 million for high-power, automated, tube-and-sheet cells.
The outsourcing math is straightforward. If you need laser tolerances or oxide-free stainless edges only occasionally, buying a laser is hard to justify against a machine that must run near capacity to amortize. Job shops and laser service centers exist across the US precisely to fill this gap, and turning around laser-cut blanks in a few days is common. Outsource laser when: annual laser-hours are low, parts are thin and tolerance-critical, or you need clean edges on stainless without a finishing department.
Keep plasma and oxy-fuel in-house for the everyday structural and plate work they dominate; those machines pay for themselves on volume and give you scheduling control. A common, efficient shop configuration is oxy-fuel or CNC plasma for the bread-and-butter carbon steel plus an outsourced laser relationship for precision jobs, upgrading to an in-house laser only when the laser workload proves sustained.
Safety, Standards, and Compliance
Every one of these processes carries regulatory and safety obligations that a professional shop builds into its SOPs, not an afterthought. Cutting also feeds directly into your welding qualification chain, so edge preparation matters beyond aesthetics.
Fumes and eye protection are governed by OSHA. OSHA 1910.252 covers general welding, cutting, and brazing requirements; 1910.253 addresses oxygen-fuel gas welding and cutting equipment specifically, including cylinder storage, hose, and regulator rules; and 1910.1000 sets air contaminant limits that drive fume extraction, critical when plasma-cutting galvanized (zinc fume, metal fume fever), stainless (hexavalent chromium), or coated steel. Match filter shade to the process and amperage per ANSI Z49.1.
Compressed gas handling follows CGA and NFPA guidance. Store oxygen and fuel-gas cylinders separated by 20 feet or a 5-foot half-hour-rated barrier (NFPA 51 / OSHA 1910.253), keep cylinders secured upright, and use proper CGA-numbered connections (for example CGA 510 for acetylene, CGA 540 for oxygen). Cylinders transported over the road are subject to DOT hazmat rules (49 CFR), which your distributor manages on delivery.
Downstream, cut edges feed weld procedures qualified under AWS D1.1 (structural steel), where prequalified thermal-cut edges and roughness limits are spelled out, and filler selection follows the AWS A5.x specs (for example A5.18 ER70S-6 for mild-steel MIG, A5.9 ER308L for 304 stainless, A5.10 for aluminum). Codes such as ASME B31 for piping and API standards for tanks and pressure equipment may also govern how edges are prepared and inspected. A distributor who knows these standards helps you spec gas purity, tips, and consumables that keep the whole chain compliant.
Decision Matrix: Picking the Right Process
Use the matrix below as a first-pass filter, then adjust for your actual volume and finishing capability. The right answer is often to own two processes and outsource the third.
Rules of thumb that hold up on the shop floor: if it is carbon steel over 1.5 inches, oxy-fuel is almost always the economical answer. If it is conductive metal (stainless, aluminum, copper) in the 1/8 to 1.5 inch range at moderate tolerance, plasma is the workhorse. If it is thin, tolerance-critical, high-volume, or needs oxide-free stainless edges, laser wins on total cost despite the capital, provided you can keep it fed. When laser demand is intermittent, outsource it rather than starve an expensive asset.
| If your job is... | Best fit | Why |
|---|---|---|
| Carbon steel plate 2-12 in | Oxy-fuel | Only process that reaches the thickness cheaply, square edge |
| Structural carbon steel 3/8-1.5 in | Plasma | Fast, low capital, dross-free in its window |
| Stainless / aluminum 1/8-1.25 in | Plasma or laser | Oxy-fuel cannot cut these; plasma for cost, laser for edge |
| Thin sheet under 1/4 in, tight tol. | Laser | Narrow kerf, tiny HAZ, best tolerance and speed |
| Oxide-free stainless, weld-ready | Laser (N2) | Clean edge eliminates the finishing department |
| High-mix, low-volume precision | Outsource laser | Capital cannot be amortized at low laser-hours |
| Mobile / field cutting, no power | Oxy-fuel | Portable, needs only cylinders, no electricity |
Frequently Asked Questions
Why can't oxy-fuel cut stainless steel or aluminum?
Oxy-fuel is a chemical oxidation process, not a melting process; it works by burning iron into iron oxide and blowing the slag out with a cutting-oxygen jet. Stainless steel forms a refractory chromium oxide and aluminum forms a tenacious aluminum oxide, both of which have very high melting points and do not support the exothermic iron-oxidation reaction. As a result the oxygen stream cannot sustain a cut, which is exactly why plasma or laser is required for those metals.
What thickness is the crossover point where plasma beats laser?
For carbon steel, laser typically holds the edge-quality and speed advantage up to about 1/2 inch, while plasma becomes more economical and often faster from roughly 3/4 inch to its 1.5-2 inch quality ceiling. The exact crossover depends on your laser power (a 12 kW fiber pushes the point higher than a 4 kW) and your tolerance requirements. If tolerance is loose and volume is high in the 1/2 to 1 inch band, HD plasma usually wins on total cost per part.
How much do plasma consumables really cost per hour?
It depends heavily on amperage and pierce frequency. On a mid-range mechanized system a matched electrode and nozzle set runs roughly 20 to 60 dollars and may last 1 to 3 hours of arc-on time, so plan on a few dollars to low tens of dollars per hour in consumables plus gas. Frequent piercing on nested thin parts is what burns electrodes fastest, so optimizing pierce count and using correct pierce-height settings extends consumable life dramatically.
Is air plasma good enough, or do I need oxygen or nitrogen plasma?
Air plasma is perfectly adequate for general carbon-steel fabrication and is the cheapest to run since it uses shop compressed air. Oxygen plasma gives cleaner, more weld-ready edges and better speed on carbon steel, while nitrogen or H35 (argon-hydrogen) plasma produces superior edges on stainless and aluminum with less oxidation. Match the gas to the metal and the edge quality you are selling; many shops run air for structural work and switch to N2 for stainless.
When does buying a fiber laser make financial sense versus outsourcing?
Because a fiber laser cell runs from roughly 250,000 dollars into the millions, it generally needs sustained utilization (often one-and-a-half to two shifts) to amortize. If your laser-suitable work is intermittent or seasonal, outsourcing to a service center gives you laser tolerances without the fixed cost or the idle-machine risk. Run the annual laser-hours you can realistically fill; if the machine would sit idle much of the week, outsource and revisit the decision as volume grows.
What gases and purities should I stock for each process?
For oxy-fuel, use 99.5 percent or better cutting oxygen plus acetylene (best pierce), propane, or propylene as the fuel, and never run acetylene above 15 PSIG. For plasma, clean, dry shop air works for carbon steel, with nitrogen or H35 for stainless and aluminum and oxygen for premium mild-steel edges. For fiber laser, use oxygen assist on carbon steel and high-pressure nitrogen (often 99.999 percent) for oxide-free stainless and aluminum; high-volume shops frequently justify a nitrogen generator or bulk liquid tank over cylinders.
Which process gives the tightest tolerance and squarest edge?
Laser is the clear winner, holding tolerances around plus or minus 0.002 to 0.010 inch with a kerf as narrow as 0.004 to 0.020 inch and a heat-affected zone often under 0.010 inch. Nitrogen-assist laser edges on stainless come off the table oxide-free and weld-ready with no secondary finishing. Plasma runs a slight bevel and wider kerf, and oxy-fuel is square but coarse, so for precision parts and clean edges laser eliminates the most downstream labor.
What are the key safety and code requirements for these cutting methods?
OSHA 1910.252 and 1910.253 govern cutting operations and oxygen-fuel gas equipment, and 1910.1000 sets fume exposure limits that make extraction essential when cutting galvanized, coated, or stainless material (zinc and hexavalent chromium fumes). Store oxygen and fuel-gas cylinders separated by 20 feet or a rated barrier per NFPA 51, secure all cylinders upright, and use correct CGA connections. Cut edges that feed structural welds must meet AWS D1.1 edge-quality provisions, so cutting quality is part of your overall code compliance, not just cosmetics.
Find a Supplier Near You
Browse thousands of verified welding and industrial gas distributors across all 50 states.
Browse the Directory →