Plasma Torches 101
Plasma Torches: A Technical Introduction
What Is Plasma?
Plasma is sometimes called the fourth state of matter. Heat a gas sufficiently (or send enough energy its way through other means) and some of the electrons bound to its atoms will start breaking free, leaving the remnant of those atoms as charged particles called ions. The resulting soup of ions and free electrons (the plasma) is electrically conductive and responds to electromagnetic fields.
Pasted image 20260220185704.png
There are two broad ways to categorize plasmas, as "thermal" or "non-thermal". Perhaps not surprisingly, the main difference between them is how hot they are.
In a thermal plasma, the ions and electrons are roughly the same temperature, meaning the bulk gas and ions are quite hot (>10,000°C), as well as the electrons. Thermal plasmas are generally created using an electric arc and a flow of gas. In a non-thermal (or 'cold') plasma, the electrons are hot and mobile but the ions and the bulk gas stay cool (<100°C), so only the electrons are energetic. Non-thermal plasmas are often created in vacuum or inert gas environments using other electric phenomena (dielectric barrier, corona, or glow discharge).
Pasted image 20260220185946.png
Non-thermal plasmas are used in specialized applications such as semiconductor fabrication, medical sterilization, and neon lights, where high heat would be damaging to other aspects of the process. As this essay concerns itself with the use of plasma torches for industrial heating, the rest of this essay will focus on thermal plasma only—the heat generation is the point.
Plasma Torches: What They Are and How They Work
A plasma torch is a device that generates thermal plasma and then directs it through a nozzle. Just as a combustion burner uses a mixture of fuel and air to create and control a steady flow of flame, a plasma torch uses a mixture of electricity and gas to create and control a steady flow of plasma.
The way this is most often accomplished is by creating an electric arc between two electrodes that are positioned across a flowing gas stream (typically air, nitrogen, argon, or hydrogen). The gas absorbs energy from the arc, which ionizes it into an extremely hot, electrically conductive plasma (~5,000 to 25,000°C at the arc, cooling to ~2,000–5,000°C at working distance). This plasma exits the torch in a high-velocity, high-temperature region called the 'plasma jet', which then diffuses into the surrounding environment, creating a larger and more diffuse volume of heated air called the plasma plume.
Two electrodes—typically a cathode (often tungsten or hafnium) and an annular anode (usually copper)—sit inside the torch body. Gas flows between them. An arc is struck, the gas ionizes and exits through a nozzle as a plasma jet.
Plasma torches can be designed in one of two configurations, referring to whether the arc is "transferred" or "non-transferred." A non-transferred-arc torch means the arc passes between two electrodes within the torch body, and the workpiece just sees a jet of plasma and hot gas. This makes the torch operate the same independently of what it's pointed at—metal, ceramic, liquid, whatever. Transferred-arc torches, by contrast, contain one electrode in the torch body and use the workpiece as the other electrode. This is more efficient at delivering heat directly into a workpiece and is useful for melting conductive materials (electric arc furnaces do this), but it limits what you can heat and how you can position the torch.
The dominant configuration for general industrial heating is the non-transferred arc torch, with transferred torches used primarily for cutting (plasma cutting for metal fabrication) or melting (electric arc furnaces).
Scale and Power Ranges
Plasma torches span a remarkable range of scales:
At the small end, handheld or CNC plasma cutters (often used in metal fabrication shops) operate in the 5–25 kW range at 30–100 amps. These are designed for intermittent duty, air-cooled in many cases, and optimized for cutting rather than sustained heating
Pasted image 20260310065749.png
At more moderate power levels, small industrial torches used for thermal spray coating, laboratory research, and specialized chemical processing typically fall in the 10–100 kW range. These are water-cooled & designed for continuous operation, usually for fairly specific & niche processes
Pasted image 20260310065848.png
At the high-power end are industrial process heating torches, used for waste treatment, mineral processing, or heating industrial furnaces. These operate in the range of 100 kW to several megawatts per torch, and large installations often use multiple torches in parallel. A waste-to-energy gasification system might deploy several 1–5 MW torches simultaneously, and plasma arc furnaces for steelmaking or hazardous waste vitrification can operate as high as tens of megawatts of combined arc power.
Pasted image 20260310065837.png
Power level drives almost everything else in the design. Higher power means larger electrodes, more aggressive water cooling, heavier electrical infrastructure, and larger gas flows. The basic physics is the same across scales, but the engineering challenges are quite different—a 10 kW bench torch and a 5 MW industrial unit share operating principles but almost nothing in common mechanically or electrically.
Gas Selection and Its Effects
The choice of plasma gas is one of the most consequential design decisions for a plasma torch system. It affects arc stability, electrode wear, the temperature and enthalpy of the plasma jet, reactivity with the workpiece, and operating cost. The main options are argon, nitrogen, air, hydrogen, and mixtures of these.
Argon has the lowest ionization potential of the common plasma gases (~15.8 eV), which means it ionizes most readily and produces the most stable arc. It is inert, non-reactive with electrodes or workpiece materials, and makes arc starting easy. The tradeoffs are cost (argon is separated from air cryogenically and is not cheap at scale) and relatively low specific enthalpy - because it is monatomic, there is no energy stored in molecular dissociation, so the amount of heat delivered per unit mass flow is modest compared to diatomic gases. Argon is common for research torches, thermal spray, and applications where atmosphere control is critical.
Nitrogen is cheaper than argon and has a higher specific enthalpy because it is diatomic - nitrogen molecules dissociate in the arc and recombine downstream, releasing additional energy at the workpiece surface. Arc stability with nitrogen is somewhat worse than argon but still manageable. The complication is reactivity: at high temperatures, nitrogen can react with metals and some ceramics to form nitrides, which may or may not be desirable depending on the application. Nitrogen also produces NOx emissions when heated in the presence of oxygen, which requires consideration for emissions compliance. For many industrial heating applications, nitrogen is the practical choice—cheap enough to use in volume and reactive enough to be useful in some processes.
Air is essentially free as a plasma gas, which is why it dominates plasma cutting. It is oxidizing, which accelerates the cutting process on ferrous metals. For heating applications, however, oxidizing atmosphere is often a problem—it attacks electrodes (copper anodes oxidize, tungsten cathodes oxidize), degrades refractory linings, and can damage workpiece surfaces. Air plasma can also produce significant NOx. For applications that can tolerate an oxidizing environment, or where the oxidizing chemistry is actually useful (combustion enhancement, for example), air is the obvious cost-minimizing choice.
Hydrogen is unusual in that it has extremely high thermal conductivity and very high specific enthalpy - roughly an order of magnitude higher enthalpy per unit mass than argon. Small additions of hydrogen to an argon or nitrogen plasma dramatically increase the heat transfer rate from the jet to the workpiece, which is valuable when the bottleneck is getting heat into a dense or poorly-conducting material. Hydrogen also creates a reducing atmosphere, which protects metals from oxidation. The downsides are real: hydrogen is flammable and explosive, requiring careful gas handling infrastructure, and pure hydrogen plasma is difficult to sustain stably at high power. As a result, hydrogen is almost always used as a minority additive (5–25% by volume in an argon or nitrogen carrier) rather than as the primary gas.
Mixtures are common in practice. Argon-hydrogen (Ar/H₂) is the standard for high-enthalpy applications like plasma spraying of dense ceramics, where high heat flux to the substrate is needed. Nitrogen-hydrogen (N₂/H₂) provides a reducing atmosphere with good enthalpy at lower cost. The mixture ratio is a tunable parameter that lets designers balance arc stability, enthalpy, and atmosphere chemistry for a specific process.
One further consideration is gas flow rate, which is independent of but coupled to gas type. Higher flow rates increase the mass flux of hot gas delivered to the workpiece but also cool the arc column and may reduce the peak plasma temperature. Finding the right flow rate for a given power level and application is part of torch characterization work.
Power Supply and Electrical Characteristics
An electric arc is not a well-behaved electrical load. Understanding the electrical characteristics of arcs is necessary for understanding why plasma torch power supplies are designed the way they are.
An arc has negative dynamic resistance: as current through it increases, the arc voltage tends to decrease. This is the opposite of a resistor. Left to its own devices, an arc connected to a constant-voltage supply will draw more and more current until something fails. This is why plasma torches are operated with current-controlled power supplies rather than voltage-controlled ones. The supply maintains a set current; the arc finds its own voltage based on arc length, gas composition, and current level. Typical arc voltages are 50-300V depending on torch design and scale; currents range from tens of amps in small systems to thousands of amps in large industrial units.
DC power supplies are the standard for industrial plasma heating. The supply typically rectifies three-phase AC mains power to a DC bus, then conditions it to a stable, current-controlled DC output. Older designs used thyristor (SCR) phase-controlled rectifiers, which are robust and relatively simple but have poor power factor at partial load and limited dynamic response. Modern installations increasingly use IGBT-based inverter power supplies, which offer better control bandwidth, higher efficiency, and lower harmonic distortion on the supply side. The tradeoff is greater complexity and sensitivity to fault conditions.
Arc starting requires special attention because the gap between electrodes must first be ionized. Contact starting (touching the electrodes together and then withdrawing) works but causes electrode wear. High-frequency (HF) starting uses a brief burst of high-voltage, high-frequency signal to ionize the gap without contact - this is the characteristic "snap" when a plasma cutter or TIG welder strikes an arc. For industrial torches, a pilot arc (a low-power auxiliary arc between the cathode and the nozzle/anode) is often used to pre-ionize the gas path before the main arc is established at full power.
Arc voltage is not independently adjustable. This is a point of confusion for people coming from other electrical systems. Given a fixed current, arc voltage is determined by arc length (electrode spacing), gas type and flow rate, and plasma temperature. The power supply sets current; the system geometry and operating conditions set voltage; power is the product of the two. Changing power requires changing current, changing electrode geometry, or changing gas conditions - usually current is the control variable.
Power factor and power quality become significant at larger scales. A large plasma installation draws substantial current and can introduce harmonics and reactive power onto the facility electrical system. Power factor correction equipment and harmonic filters are often part of the balance-of-plant for multi-megawatt installations. This is not a concern for a 20 kW research torch but becomes a real engineering and utility compliance issue above a few hundred kilowatts.
AC power supplies are used in some torch configurations, particularly multi-electrode AC torches where three-phase power is fed directly to three electrodes in a roughly symmetric configuration. The arc reverses direction each half-cycle, distributing electrode wear more evenly across all three electrodes. The power supply can in principle be simpler (a three-phase transformer rather than a full rectifier-inverter system) but arc stability is inherently more challenging with AC because the arc extinguishes and re-ignites each half-cycle. AC torches tend to require more sophisticated arc stabilization approaches (magnetic field stabilization, gas swirl) to prevent the arc from wandering or extinguishing at the zero-crossing.
Torch Design: Electrodes, Arc Configuration, and Cooling
Electrode Configuration
The cathode is where electrons leave the electrode metal and enter the arc. In DC torches, the cathode is typically a rod of tungsten (often with a small percentage of lanthanum oxide or thorium oxide to improve electron emission and arc stability) or hafnium. Tungsten has a high melting point (~3,400°C) and good electron emissivity, making it suitable as a hot cathode. Hafnium is preferred for air and oxygen plasma applications because it forms a self-limiting oxide layer that protects the base metal.
The anode is where the arc attaches on the downstream end. In non-transferred arc torches, the anode is typically a copper cylinder or ring that forms part of the nozzle assembly. Copper is chosen for its high thermal conductivity - it must transfer heat away from the arc attachment point rapidly enough to avoid melting. This is accomplished through aggressive water cooling channels immediately behind the arc-facing surface. Despite the cooling, the copper surface at the arc root is locally at extremely high temperature; arc attachment is a dynamic phenomenon and the arc root "moves" across the anode surface to distribute the heat load.
DC vs. AC
DC torches have asymmetric electrode wear because the cathode and anode operate at different conditions—the cathode runs hot and emits electrons, while the anode absorbs the electron current and sees a different thermal and erosion pattern. This means the two electrodes have different service lives and replacement schedules, which adds maintenance complexity. On the other hand, DC arcs are inherently stable and well-characterized, which makes system design and control more tractable.
AC torches distribute wear more symmetrically because each electrode alternates between cathode and anode functions each half-cycle. In three-phase AC configurations with three electrodes, the wear is distributed across all three. This can extend electrode life and simplify consumable management. The price is arc instability—the zero-crossing extinction and re-ignition transient puts higher demands on torch geometry and power supply design to maintain arc continuity.
Cooling
Water cooling is not optional—it is structurally essential. The arc generates local temperatures far above the melting point of every material used to make the torch body. The torch survives only because water cooling removes heat from the electrode and nozzle surfaces fast enough to maintain a thin solid layer between the plasma and the bulk metal.
The anode cooling circuit is typically the most critical. Water flows through channels machined or formed directly behind the anode surface, maintaining very high flow velocities to ensure nucleate boiling doesn't transition to film boiling (which would catastrophically reduce heat transfer). Cathode cooling is also necessary but typically less aggressive because the tungsten cathode transfers heat primarily through the solid electrode rather than through arc radiation.
Cooling water flow, inlet temperature, and pressure are monitored as safety and process variables. Loss of cooling water flow is a trip condition that should shut down the torch immediately—without cooling, the anode will melt within seconds at full power.
The cooling circuit also represents a direct energy loss, which is addressed in the next section.
Energy Balance and Thermal Efficiency
Not all electrical power input to a plasma torch becomes useful heat in the process. Understanding where the energy goes is important for comparing plasma heating honestly against alternatives and for optimizing system design.
The total electrical input power to the torch splits into four main flows. The first is the enthalpy of the plasma jet—the thermal energy carried by the hot gas as it exits the nozzle. This is the useful output and the number that goes in the numerator when calculating thermal efficiency. The second is electrode cooling losses—heat conducted into the electrode bodies and removed by the cooling water circuit. The third is torch body and nozzle cooling losses—heat conducted into the surrounding structure, also removed by cooling water. The fourth is radiation from the arc column, which is generally a small fraction at the power levels and pressures relevant to industrial torches (unlike, say, a welding arc where radiation is more visible and significant).
For a well-designed non-transferred arc torch, thermal efficiency (defined as jet enthalpy out divided by electrical power in) is typically in the range of 50–75%. The specific number depends heavily on torch design, operating point, and gas type. Higher current operation tends to improve efficiency because the useful arc power scales faster than the cooling losses. Larger torches generally achieve better efficiency than smaller ones for the same reason. The cooling losses are roughly proportional to electrode surface area, while arc power scales with current and voltage.
Diatomic gases (nitrogen, hydrogen, air) complicate the efficiency picture in an interesting way. Energy goes into dissociating diatomic molecules in the arc - splitting N₂ into two N atoms, for example - which doesn't contribute to gas temperature in the arc column. When the dissociated gas recombines downstream (at the workpiece surface or in the plume), that dissociation energy is released as heat. If recombination happens at the workpiece, this represents useful energy delivery that doesn't show up in a simple temperature measurement of the jet. The actual heat delivery to the process can therefore be higher than the jet temperature alone would suggest, particularly for hydrogen-bearing mixtures.
Comparison to combustion requires care. A well-designed combustion system converts 85–95% of fuel chemical energy to process heat. Plasma thermal efficiency of 50–75% sounds worse, but the comparison is not straightforward. Plasma delivers heat at temperatures that combustion cannot reach. And the electrical energy input itself was generated somewhere upstream - at a fossil fuel power plant (perhaps 35–45% efficient) or from renewables. The relevant comparison for climate and cost purposes is the full energy chain, not just the torch efficiency in isolation. When powered by cheap, low-carbon electricity, plasma heating can be both economically and environmentally competitive despite lower point-of-use efficiency.
The cooling water losses also represent a potential opportunity. A large plasma installation rejects substantial heat through its cooling circuit - a 1 MW torch at 65% efficiency rejects roughly 350 kW through cooling water. Depending on the facility, this might be recoverable for lower-temperature process heat or building heating rather than simply dumped to a cooling tower.
What Plasma Torches Are and Are Not Good For
Plasma torches are not going to be a universal replacement for existing industrial heating processes. Low-temperature process heat (<500°C) is better served by heat pumps or resistance heating - simpler, cheaper, and more efficient for that temperature range. Induction heating is superior for many metal heating and melting applications where the geometry and material are suitable. Resistance heating works well for applications with uniform temperature requirements and appropriate geometries. Combustion is robust and cheap.
Plasma's sweet spots are: high temperatures beyond combustion's reach, applications requiring controlled or carbon-free atmospheres, situations demanding high power density or rapid response, and increasingly, applications where the economics of cheap electricity make direct electric heating competitive with fuel, particularly where the facility wants flexibility between electric and fuel-based operation.
That's all for this initial primer, I will be following up with a more in-depth look at some technical aspects and a market analysis of the pasma torch space.