Welding Health and Safety by the Numbers
Chapter 2 of 10 · 10 min read
Welding fume: what it is and how much is made
See how welding fume forms from vaporised metal, what it contains and does to the body, compare how much fume each process makes, and calculate why general ventilation alone rarely dilutes it enough.
The plume that rises from a welding arc looks like smoke, but it is something more dangerous. It is mostly metal: tiny particles of oxide formed when metal vaporised by the arc condenses in the air, so small that they float with the plume, pass deep into the lungs, and in some cases into the bloodstream. With them come gases formed by the arc’s heat and light. The fume’s composition depends on the process, the consumable, the base metal and anything on its surface, and its quantity on the process and its settings. In 2017 the International Agency for Research on Cancer classified welding fumes as carcinogenic to humans, its highest category, publishing the evaluation in its Volume 118 in 2018. The early manuals had little to say about fume; modern welding health and safety starts with it.
This chapter explains how fume forms and what it contains, summarises its effects on health, compares how much fume the main processes make, and calculates why general ventilation alone rarely dilutes it to a safe level near the welder.
How fume forms
The arc is hot enough to boil metal. At its centre, metal from the electrode’s tip and the droplets crossing the arc, and to a lesser extent from the pool, evaporates. As the vapour leaves the arc and meets the cooler air, it condenses into extremely small particles and oxidises, forming a cloud of metal oxide particles much smaller than a micrometre across. The particles cluster into chains and clumps, but remain small enough to stay airborne and to be breathed deep into the lungs. Most of the fume comes from the consumable, the electrode or wire and its coating or flux, rather than from the base metal, because the consumable’s tip is the hottest metal in the arc. Coatings and contaminants on the base metal, such as zinc, paint, oil and primers, add their own fume and decomposition products.
What it contains
| Where it comes from | Main health effects | |
|---|---|---|
| Iron oxide | steel consumables and base metal | deposits in the lungs; part of the fume's overall lung effects |
| Manganese | most steel consumables; hardfacing and manganese steel | nervous system damage with prolonged exposure |
| Hexavalent chromium | stainless steel and chromium-bearing consumables | lung cancer; respiratory irritation; skin sensitisation |
| Nickel | stainless and nickel-alloy consumables | lung cancer; sensitisation |
| Zinc oxide | galvanised coatings | metal fume fever |
| Fluorides | electrode coatings and fluxes | respiratory and eye irritation; bone effects with long exposure |
| Ozone | ultraviolet light acting on air, worst with aluminium and gas-shielded processes | lung irritation |
| Nitrogen oxides and carbon monoxide | the arc's heat; carbon dioxide shielding gas | lung irritation; carbon monoxide poisoning in enclosed spaces |
Fume from mild steel welding is mostly iron oxide with smaller amounts of manganese and other elements. Stainless steel fume, by contrast, contains chromium, much of it in the hexavalent form, and nickel, both carcinogens. Fume from flux-cored and manual metal arc welding contains material from the flux, including fluorides. Gases such as ozone and nitrogen oxides form around the arc from the air itself.
What fume does to the body
The effects range from the immediate to the very long term. Heavy exposure to zinc and some other metal fumes causes metal fume fever, a flu-like illness with fever, chills and aching that appears hours after exposure and passes in a day or two, but signals an exposure far too high. Fume and gases irritate the nose, throat and lungs, and welders suffer more bronchitis, asthma and reduced lung function than other workers. Welders are also more susceptible to serious pneumonia. Over years, fume increases the risk of lung cancer, which is why welding fume as a whole is classed as a carcinogen, whatever the metal welded. Manganese, absorbed from the fume, can affect the nervous system with prolonged high exposure. These effects depend on the dose: how much fume is breathed, for how long each day, and over how many years of a working life.
How much fume each process makes
The quantity of fume a process makes is often expressed as a fume formation rate: grams of fume per kilogram of consumable used, or grams per minute of arcing. It varies with the process, the consumable and the settings, especially the current and voltage.
Gas tungsten arc welding makes very little fume, because its electrode does not melt and its filler is added cool to the pool. Gas metal arc welding with solid wire makes moderate amounts, depending on the transfer mode; pulsed transfer makes less than spray at the same deposition. Manual metal arc and flux-cored welding make more, much of it from the flux. Self-shielded flux-cored wires, which generate their own shielding from the flux, make the most. Higher currents and voltages raise fume formation in all of them.
Fume per hour
Because fume is made per kilogram of consumable, the processes that deposit fastest can produce the most fume per hour even if their rate per kilogram is moderate. A flux-cored welder depositing 4 kg of wire an hour at about 12 g/kg produces about 48 g of fume an hour; a manual metal arc welder depositing 1.5 kg an hour at the same rate produces about 18 g. Both are a great deal of fine particulate to release into a workshop, and the welder’s face is a short distance above it.
Why dilution struggles
General ventilation, fans and open doors, dilutes fume by mixing it with clean air. If the fume mixed evenly through the air that passes through the shop, the average concentration would be the fume produced divided by the ventilation flow.
- Fume produced
- 3 × 48 = 144 g/h, about 2,400 mg/min
- Ventilation
- about 100 m³/min, 6,000 m³/h
- average concentration ≈ 2,400 ÷ 100 = 24 mg/m³ throughout the bay
- the welder's breathing zone, in the plume, is typically several times the room average
- to bring the average down tenfold, the airflow would have to rise tenfold, to 1,000 m³/min
The calculation shows why general ventilation is not enough on its own for any significant welding. Even with perfect mixing, which never happens, the average concentration in a bay with a few welders and generous fans is high; and the welder breathes not the average but the plume rising from the arc in front of their face, which is far more concentrated. Diluting that plume enough would need airflows that are impractical to move and, in winter, ruinously expensive to heat. That is why extraction at the source, capturing the fume before it reaches the welder’s breathing zone, is the main engineering control for welding fume, as Chapter 4 explains.
Reducing fume at the source
Before extraction, the first question is whether less fume can be made. Choices that cut fume include a process with a lower fume formation rate where it will do the job, such as solid wire instead of self-shielded flux-cored wire in the shop, or pulsed instead of spray transfer; consumables with lower fume formation, which some makers offer; the lowest current and voltage that give a good weld; removing coatings, paint and oil from the weld area before welding; and avoiding over-welding, which Welding Automation by the Numbers showed can add half as much again to the weld metal, and therefore to the fume. Each is a substitution or a reduction at the source, which sits higher on the hierarchy of control than any extraction system can.
Shielding gases and carbon monoxide
Carbon dioxide used as a shielding gas, alone or in mixtures, partly breaks down in the arc into carbon monoxide, which is produced in small amounts near the arc. In open, ventilated shops it disperses; in enclosed and confined spaces it can build up, along with the shielding gases themselves, which displace oxygen. Chapter 8 deals with confined spaces.
Ozone
Ozone is formed when the arc’s ultraviolet light acts on the oxygen in the air around it. It is produced most by processes with intense ultraviolet and little fume to absorb it, such as gas metal arc welding of aluminium and gas tungsten arc welding, and it is a lung irritant. Because it forms in the air around the arc rather than in the fume, extraction near the arc captures some of it, and good general ventilation helps disperse it.
Particle size and the lungs
Where a particle lands in the respiratory system depends on its size. Large dust particles are caught in the nose and throat and cleared; smaller ones reach the airways; the smallest reach the deepest part of the lungs, the alveoli, where oxygen passes into the blood and where the lung’s clearing mechanisms are slowest. Welding fume particles are mostly in that smallest range, much finer than the dust from grinding, so a large share of what a welder breathes in reaches the deep lung, where it can stay for a long time. That is part of why fume is so much more harmful, gram for gram, than the coarser dusts of other trades, and why respiratory protection for welding must use filters suited to fine particles, as Chapter 5 explains.
Cutting and gouging fume
Thermal cutting and gouging make fume too, often more than welding. Air carbon arc gouging and plasma cutting vaporise and blow out a great deal of metal, and oxy-fuel cutting produces iron oxide fume and gases; coatings on the work add their own. Cutting and Gouging by the Numbers deals with fume capture on cutting tables and in gouging bays. A fume survey of a welding shop that ignores its cutting and gouging stations misses some of its largest sources.
Who else breathes it
The welder is closest to the fume, but not the only person exposed. Helpers, fitters tacking and grinding nearby, supervisors and anyone working downwind in the same bay breathe the fume that escapes the welder’s plume and spreads through the building. In shops without good extraction, a haze of fume hangs under the roof by mid-afternoon and slowly settles back to the breathing zone of everyone on the floor. Exposure measurements for non-welders in the running case showed levels well below the welders’ but not negligible, which was one more argument for extraction at the source over respirators, which protect only the wearer.
Fume and smoking
Smoking multiplies the lung damage from many workplace dusts and fumes, and welders who smoke have a higher risk of lung disease than those who do not. Smoking in the shop also carries fume particles from hands and clothing into the mouth. The shop’s health programme offered support to welders who wanted to stop smoking, and provided washing facilities so that welders could wash hands and face before eating, drinking or smoking, which reduces the metal they swallow as well as breathe.
Fume on clothing and skin
Fume settles on clothing, skin, hair and every surface in the shop. Welders who wear their work clothes home carry metal dust with them, and in the case of stainless and nickel work, sensitising and carcinogenic metals. The shop provided overalls that stayed at work, laundered by a service, and lockers that separated clean from work clothing, practices recommended for work with hazardous metals.
The running case: the fume survey
This is an illustration. The coordinator listed every welding task in the shop with its process, consumable, base metal and coating, and the hours spent on it each week. Self-shielded flux-cored welding on the trailer chassis, about 60 hours a week in an enclosed bay, and manual metal arc welding on galvanised fittings, about 10 hours a week, stood out as the largest fume sources; stainless welding, about 15 hours a week by gas tungsten arc and gas metal arc, stood out for its hexavalent chromium. The coordinator commissioned an occupational hygienist to measure exposures on those tasks first, as Chapter 3 describes, and asked the consumable supplier about lower-fume alternatives for the chassis wire.
A failure: metal fume fever
This is an illustration. Before the coordinator’s appointment, a welder spent a long afternoon welding galvanised brackets inside a small enclosure, without extraction, to keep the wind off. That night the welder developed a fever, chills, a metallic taste and aching muscles, and was off work the next day.
The symptoms were those of metal fume fever from the zinc fume. The episode passed, but it showed an exposure far above what is acceptable. The shop now removes galvanising from the weld area before welding wherever the design allows, and welds galvanised parts only at stations with working extraction.
Exercise: fume from a day’s welding
A welder deposits about 3 kg of self-shielded flux-cored wire an hour at a fume formation rate of about 18 g/kg, for 4 hours of arc time a day. How much fume is produced in a day? If a bay with two such welders has general ventilation of about 80 m³ a minute, what would the average concentration be during welding, assuming perfect mixing? What does this suggest?
Each welder produces about 3 × 18 = 54 g of fume an hour of arcing, and about 54 × 4 = 216 g a day.
Two welders produce about 108 g an hour, about 1,800 mg a minute, so the average concentration with 80 m³ a minute of ventilation would be about 1,800 ÷ 80 ≈ 22.5 mg/m³, before the much higher concentration in each welder’s plume. General ventilation alone is nowhere near enough; the fume must be captured at the source, and the consumable and process reviewed for lower-fume options.
