Is Underwater Welding Dangerous? Safety Risks Explained

A diver kneels on a sheet-pile wall three metres under the surface, flicks an electrode holder against a steel strut, and a 300-amp arc erupts inside a cloud of bubbles. In that instant he is standing in salt water, holding a live electrode, breathing compressed gas, and working alone with no way to see his own weld. He will not surface until the job is finished — or until something goes wrong. The honest answer to “is underwater welding dangerous” is yes, unambiguously and without drama. But it is dangerous in specific, well-understood, well-controlled ways, and understanding those ways is the difference between a controlled professional operation and a tragedy.

Is Underwater Welding Dangerous? The Honest Short Answer

Yes. Underwater welding is significantly more dangerous than surface welding, and the danger is dominated by four things: drowning, electrical shock, explosive gas, and burn injury, not the arc itself. Most people picture this work as a spectacle: a diver in a helmet, a flash of light, a cloud of bubbles. The spectacle hides the real risk, which has less to do with hot metal and more to do with electricity, gas, and the possibility of not coming back up.

So why is it more dangerous? Because the welder is doing three jobs at once, and those three jobs fight each other.

On a shipyard deck, a welder is an electrician in a dry, non-conductive environment: heavy leather, dry ground, freedom to step back from the arc, and a colleague who can reach the power switch in seconds. Underwater, the same welder is a commercial diver, an electrician inside a highly conductive medium, and a confined-space worker inside a steel structure with no independent exit.

Three Roles, Conflicting Requirements

• Thermal protection versus electrical isolation. A dry suit is standard gear in cold water, because hypothermia will kill a diver long before anything else does. But it adds bulk, reduces mobility, and creates extra surface area a stray current can follow. A wet suit is lighter and more flexible, but works only in warm water or with active heating.
• Hard hat versus helmet. A hard hat is free insurance on a deck. Inside a diving helmet it is impossible, which removes a standard protection against falling objects and changes how you approach work under a suspended load.
• Escape versus task. On land, the welder can walk away from a bad weld, a hot electrode, or an arc flash. At depth, each must be handled by a diver breathing compressed gas, often in zero visibility, with a tender on the surface and a decompression obligation over every minute.

Wet underwater welding typically runs at roughly 300 to 400 amps of direct current, and the diver is in direct contact with a highly conductive salt-water path back to the power source. The current that would arc-weld steel on a deck will, in water and with a compromised insulation barrier, travel through the diver instead of the workpiece.

The Scale of the Occupation

Underwater welding is a small profession inside a small one. NIOSH estimates roughly 3,380 commercial divers in the United States as of 2017, concentrated in coastal states and around the Great Lakes. The Bureau of Labor Statistics recorded 460 nonfatal occupational injuries and illnesses involving days away from work among commercial divers during 2011 to 2017, plus 39 fatal occupational injuries in that same seven-year window.

Those figures cover all commercial diving, not just welding, so the welding subset is smaller still. The population is tiny, the work unforgiving, and the margin for error thin. A job where one mistake can end a life is not like one where a mistake means re-running a weld.

OSHA describes the pattern: commercial divers face drowning, respiratory and circulatory complications, and hypothermia, and those risks worsen with the number and length of dives, the physical demands of the work, and limited visibility. Add welding and you stack a high-energy electrical process on a punishing one.

Dangerous Is Not the Same as Unmanaged

Two ideas people often blur deserve separating. Danger describes the inherent capacity of an activity to cause harm. Risk describes the chance of harm once controls are applied. A high-rise construction site is dangerous, and it is also heavily engineered, inspected, and policed. Underwater welding is dangerous in the first sense and can be professionally managed in the second.

Regulators treat this as a distinct high-hazard activity, and the code shows it. OSHA’s commercial diving standard, 29 CFR 1910.422, carries a dedicated subsection, (g) “Welding and burning”, spelling out rules for isolation switches, grounding, insulation, gloves, and gas venting during a dive. You do not write a subsection for something you consider routine.

What follows is the practical anatomy of that risk. We walk through the five underwater welding methods defined in AWS D3.6M and the four weld classes they produce, then take the four hazard families apart: electrical shock, explosive hydrogen and oxygen, the diver’s environment (cold, depth, visibility, drowning), and the fume and gas exposure nobody expects. We also cover the wet-versus-dry comparison, recurring mistakes in incident reports, and the standards professional teams work to. For the PPE side of the picture, our guide to choosing and handling welding electrodes safely is worth a read.

One last framing: the professionals who do this work well are not braver than you or me. They are better prepared, better supported, and far more conservative about when to stop than a workshop welder would be. That distinction is the difference between a managed risk and a tragedy.

What Underwater Welding Actually Is: Five Methods, Two Fundamental Worlds

The trade uses these terms loosely, which is why “underwater welding” can mean very different things depending on who you ask. The convention, as TWI sets it out, is simple: “hyperbaric welding” is normally used to mean dry welding, while “underwater welding” refers to the wet environment. Both belong to the same family of repair work, and the split between them decides which hazards dominate.

Wet welding means the arc burns in open water with no barrier between the sea and the weld pool. Dry, or hyperbaric, welding means the water is physically displaced, usually by a sealed chamber or a habitat, so the welder works in air at the pressure of the surrounding sea. In the first case the welder is in the water holding a live electrode. In the second, the welder is dry and enclosed, breathing an atmosphere somebody has deliberately engineered. Almost every risk difference in this article follows from that single split.

Steel, by far the most commonly welded material underwater, dominates the workload: ship hulls, sheet piling, offshore platforms, harbour structures and subsea pipelines.

The five methods AWS D3.6M recognises

The AWS D3.6M underwater welding code sorts the entire field into five basic methods:

1. One-atmosphere dry welding. The work happens at surface pressure inside a pressure vessel that is taken down to the job. The welder never meets the surrounding water, and the method is essentially independent of depth.
2. Habitat welding. A large chamber is lowered and sealed around the work area at ambient pressure. The welder works inside the dry habitat, out of diving equipment, much as if standing in a workshop.
3. Dry chamber welding. A small, open-bottomed chamber holds only the diver’s head and shoulders. The diver stays in full diving gear but welds inside a gas-filled bubble.
4. Dry-spot welding. A small transparent enclosure is fixed to the work and filled with gas. The welder remains outside in the water, working through the enclosure.
5. Wet welding. Ambient pressure, no mechanical barrier, the arc burns directly in the surrounding water.

Four weld classes

Within whichever method is used, AWS D3.6M grades the joint by class:

• Class A – comparable in quality and mechanical properties to above-water welding.
• Class B – intended for less critical applications.
• Class C – for joints where load-bearing capability is not the primary consideration.
• Class O – joints made to satisfy some other designated code or specification.

The hybrid: local cavity and dry-spot welding

Dry-spot working sits between the two worlds, and it has quietly become the workhorse of structural subsea repair. A local cavity is sealed to the surface, the water is displaced with gas, and the weld is made inside a small dry envelope while the welder stays outside in the water. The underwater welding research literature shows why it works so well.

Using standard GMAW equipment, the technique produces diffusible hydrogen in the weld metal of roughly 5 to 21 ml per 100 g of deposited iron, depending mainly on shielding gas flow rate. Joints made this way have been accepted for repair work to depths of about 200 m by classification societies, which is well beyond what wet welding can credibly deliver.

Why the method decides the risk

Here is the link back to the central question. The method determines whether the welder is in the water, and that in turn determines which hazards dominate. A welder inside a habitat is essentially a confined-space worker in a controlled atmosphere. A welder doing a wet weld is simultaneously a diver, an electrician working in a conductive medium, and a confined-space worker, and those three roles pull in different directions.

Depth realities follow the same pattern. Wet welding with covered electrodes or flux-cored arc welding is carried out directly in the water down to about 100 m. Dry hyperbaric welding has been performed operationally at roughly 300 m for manual work, with research continuing into deeper applications. An overview of wet and dry welding sets out the trade-offs in more detail.

Method	Welder in the water?	Depth reach
One-atmosphere dry	No	Depth-independent in principle
Habitat	No	200 m and beyond
Dry chamber	Head and shoulders only	Moderate depths
Dry-spot / local cavity	Yes	About 200 m
Wet	Fully	About 100 m

Wet welding stays the default for good reasons: it is quick to mobilise, needs far less equipment and suits awkward geometry. It is also where the choice of process matters most, and our guide to why SMAW still wins in the world of stick welding explains the covered-electrode side that most underwater work depends on.

The Diver’s Environment: Drowning, Cold, Depth and the Limits of Visibility

Stack a welding arc onto a dive and the welder becomes three things at once: operator of a hot process, occupant of a sealed breathing environment, and a person whose survival depends on reaching the surface. The welding hazards are only half the story. The other half is the water, and it keeps punishing the welder even when the arc is busy.

Drowning: the default failure mode

Start with the outcome that keeps every diving supervisor awake. In many incidents the arc is the trigger, but the emergency is not the burn or the shock. It is a diver who cannot resurface. A shock can lock muscles, a helmet can flood, a suit can tear, a gas line can part. Each turns a technical problem into a fight for air, and air is the one thing a submerged diver cannot replace from a spare cylinder on the deck.

A real case: Hong Kong, November 2020

A Work Safety Alert from the Hong Kong Labour Department makes the point exactly. In November 2020 a diver was drowned while carrying out underwater welding on the deck piles of a container terminal, using surface-supplied air diving equipment.

The alert does not treat welding as the root cause. It treats the job as a system, and sets out what that system needs:

• A competent person conducting task-specific risk assessments covering the nature of the work, site and surface conditions, diving mode, plant and equipment, and time underwater, with particular attention to drowning and electrocution.
• A diving supervisor in immediate control of the operation, adequate team size, and team members medically fit for the task.
• A compressor air intake properly located so it cannot draw in engine exhaust or other contaminated air, with hoses kept free of oil and contamination.
• Effective communication with emergency services, plus the capability to deliver first aid.

Cold, fatigue and the body under load

Cold is the second pillar. NIOSH lists drowning, respiratory and circulatory complications, and hypothermia among the marine hazards commercial divers face, and notes that risk grows with the number and length of dives, with the physical demands of the work and with limited visibility.

Hypothermia is the quiet one. A diver in cold water loses heat faster than the body can generate it, so core temperature slides down across a long shift. Judgement narrows, fine motor control degrades, and small errors appear: a poor bead, a rushed electrode change, a garbled call. A technical problem has become a medical one, and the diver is less able to solve either. The Divers Alert Network’s commercial diving safety guidance treats cold stress and fatigue as compounding hazards for exactly this reason.

Depth: pressure, gas and the way back up

Depth changes the job in ways that are invisible from the surface. Rising pressure destabilises the arc, and the deeper the welder goes the harder it becomes to hold a consistent bead. Breathing gas behaves differently under pressure, and nitrogen narcosis degrades judgement and decision-making at depth. Every minute at depth is a debt settled later on decompression.

OSHA notes that although the diving depths used for temporary underwater hull repairs are relatively shallow, rapid surfacing during an emergency can lead to decompression problems. That is the trap: a shallow job creates the impression that the ascent is trivial, and then the emergency forces exactly that fast ascent anyway. Hull and pontoon work on aluminum boat hulls shows how a shallow job can still demand disciplined ascent control.

Working blind in brown water

Visibility is the everyday reality. In silty harbour or river water, suspended sediment turns the water opaque brown within minutes, and there is no natural light at any depth. The welder works from touch, from feel and from a rehearsed procedure, unable to see the puddle or the bead. It is also why the most frequent targets are pipe jackets, sheet piling and structural steel: simple geometry, known location, repeatable procedure.

Lighting is therefore a controlled hazard, not a comfort. So is the physical load. Holding a welding posture, bracing against structure in moving water at depth for hours is heavy work inside a suit that is already heavy, and that load feeds straight back into fatigue, cold stress and error rate.

The gear conflict nobody can design away

Finally, note the PPE contradiction. A dry suit gives excellent thermal protection, which matters in cold water, but it is heavy, bulky and reduces mobility. A wet suit is lighter and easier to move in, which matters when a diver is fighting current and holding a stable stance, but it works only in warmer water or with active heating. Neither option is simply better. The right answer depends on water temperature, depth, task duration and the diver’s own tolerance of heat stress. Marine-side versions of the same trade-offs are covered in our guide to AJ Marine welding processes, standards and safety.

Electrical Shock: The Hazard Everyone Underestimates

When people think of the dangers of underwater welding, their minds go straight to pressure, heat, or the arc itself. In reality, electricity is the most unforgiving threat of the job. On land, air acts as an insulator. Underwater, conductive saltwater surrounds the diver, turning their entire body into part of a potential circuit for any stray current.

This isn’t speculation. The National Research Council’s Marine Board made this clear in its report Underwater Electrical Safety Practices, stating that the principal electrical hazard for divers is non-lethal electric shock when in close proximity to a high-powered source without adequate protective dress.

AC vs DC: A Matter of Survival

The type of current is the deciding factor for how dangerous a shock can be. Alternating current (AC) causes the heart to seize in a sustained contraction, while direct current (DC) allows it to recover between impulses. Because of this, far less AC is needed to reach a dangerous level for a diver.

Current type	Safe upper limit applied to a diver’s body	Practical consequence
Alternating current (AC)	About 10 milliamperes	Extremely dangerous at low levels; never used for underwater welding
Direct current (DC)	Roughly 60 milliamperes	Standard supply for underwater welding due to its higher tolerance threshold

One Failure Was Enough

The historical record backs this up. The only reported case of electrocution during commercial underwater cutting or welding was caused by the failure of a rectifier within an AC-DC rectifier-type power supply. This single component failure reverted the supply to AC while the diver was in the water.

It’s a humbling lesson. Much like why SMAW remains reliable on the surface, simplicity and robust equipment matter greatly. Underwater, however, there is little room for even the smallest equipment failure to be forgiven.

The Regulations That Prevent It

To combat these risks, the US commercial diving standard 29 CFR 1910.422 lays out strict requirements for welding and burning underwater. These aren’t suggestions, they’re accident prevention measures:

• Tended current control: A switch to interrupt power to the electrode must be under the control of a dive team member in voice communication with the diver, and kept open unless actively welding or burning.
• Proper grounding: The welding machine frame must be grounded to prevent it from becoming live in conductive water.
• Rated insulation: Cables, electrode holders, and connections must be properly insulated and rated for the maximum current.
• Diver protection: Insulated gloves must be provided to the diver.

The most critical of these is the tended switch. A diver changing electrodes with live current is one of the easiest ways to complete a circuit through their body, and a present tender is the only thing stopping it in that moment.

Pragmatic Safeguards For The Job

Regulations are only half of the equation. In practice, a few hard equipment rules are what keep a diver safe:

• Constant-current power: A constant-current machine is required, as it can handle the erratic resistance of a wet electrode far better than a constant-voltage supply.
• Immediate topside disconnect: A positive operating disconnect switch should be placed on the welding lead at the topside station for instant power cutoffs.
• Water-cooled holders: Standard above-water electrode holders will overheat and fail underwater, creating unnecessary risks in an already volatile environment.
• Vigilant positioning: The diver must never touch the electrode, a live electrical part, and the workpiece all at the same time. This is the single most preventable way to become the path to ground.
• Rigorous inspections: Every connection must be tight, and all cables inspected before entry. A failed ground or damaged cable is not a minor issue.

Shock Doesn’t End At The Surface

The dangers aren’t always immediate. The National Research Council noted that repeated shocks can deteriorate metallic dental fillings, warranting continued medical and dental surveillance for any diver that has taken one. Additionally, as AMRON points out, a shock that was harmless before is not guaranteed to be harmless again. Depth, salinity, suit integrity, and body position can all change from dive to dive.

With that in mind, agencies such as NIOSH treat electrical work underwater as its own distinct high-hazard activity. It’s not an extension of surface welding. It’s an entirely different electrical environment, and it demands to be treated as such.

Hydrogen, Oxygen and Explosions: The Hidden Gas Danger

The gas risk is easy to miss when you read the hazard list for an underwater welding job. The arc is a 3,000 degree ignition source, but the flame is small and neatly contained in a bubble. What is less visible is what that bubble does, and what collects above the diver in the steel. Three gas stories stack up underwater: hydrogen generated by electrolysis, oxygen deliberately supplied for cutting, and flammable hydrocarbons already trapped inside the structure. Any one of them, at the right concentration, turns a welding job into a detonation.

Hydrogen From Electrolysis

The arc breaks water molecules apart. That reaction produces hydrogen and oxygen in roughly the proportions of the water they came from, which is a violently flammable mixture. It shows up in three characteristic places: in the bubble envelope around the arc, in the overhead of compartments where rising water pushes gas into the highest corner, and in the substructure of the job, where voids, drains and undercuts act as chimneys for it.

A flooded space is not automatically a safe space. Large-diameter pipes and hull structures with poor ventilation can hold a combustible gas pocket even when they look full of water, simply because the gas is trapped above the waterline. Divers Alert Network puts it plainly: underwater welding and burning can produce hydrogen and oxygen mixtures that are highly explosive, and closed compartments, structures or pipelines subjected to the heat of underwater burning or welding need to be flooded or purged with water and vented.

Oxygen From The Cutting Hose

Oxy-fuel cutting makes the problem worse, because oxygen is not a by-product here. It is piped to the torch on purpose, and in an enclosed space it raises the oxygen concentration of the surrounding atmosphere. The more oxygen in the mix, the less fuel is needed to ignite. A mixture that sits safely on a shipyard deck can burn explosively in a flooded void.

The OSHA enforcement directive CPL 02-00-151 spells this out and is worth reading in full. A flooded compartment is not necessarily safe for cutting and welding: oxygen from the diver’s welding and cutting O2 supply hose, hydrogen generated by electrolysis, and other gases may collect in the overhead of a compartment if it is not properly vented, and if the diver cuts or welds into that area, a serious explosion can occur. The same guidance offers a simple gas-management trick: make vertical cuts and welds starting high, in shallow water depth, and finish low, in deeper water depth, so gases released early have somewhere to go instead of pooling where the arc is working.

Flammable Vapours Already In The Structure

The third source is the oldest and the sneakiest. The structure being repaired may already contain fuel: empty tanks with vapour residue, oil in ballast spaces, fuel and hydraulic oil in bilges, paint and coatings that off-gas, and hydrocarbon residue in pipes. Arc heat can ignite material the diver cannot see or smell, because the helmet’s own gas stream often scrubs those smells away.

This is also where topside familiarity misleads people. our guide to air acetylene welding covers how oxy-fuel torches behave in ordinary workshop conditions. Underwater, the same torch is working inside a sealed, poorly vented steel box filled with water. The flame behaviour and the back-pressure of the gases around it are entirely different.

What The Code Actually Requires

Federal law is unusually specific. 29 CFR 1910.422(g)(5) requires that, prior to welding or burning on closed compartments, structures or pipes that contain a flammable vapour, or in which a flammable vapour may be generated by the work, those spaces be vented, flooded, or purged with a mixture of gases that will not support combustion.

Read that last clause carefully, because it is where jobs go wrong. Venting alone is not enough unless it actually removes the flammable gas. Opening a valve at the top of a compartment does not help if the pocket sits in a different overhead void or behind a bulkhead. Purging and flooding work better because they change the whole atmosphere rather than nudge it.

Why does this matter more than a shock or a burn? An explosion inside a steel structure kills instantly and is essentially unsurvivable. Electric shock gives a diver a moment of pain and a chance to release the electrode. A burn gives seconds to get clear. A flash fire in a sealed compartment gives nothing, and it usually arrives the instant the arc strikes. That is why gas control on underwater welding is not paperwork to be closed out on the surface. It is the control that decides whether the dive ends with a weld.

The Arc, the Heat and the Blind Weld: Why Underwater Welding Is Technically Harder

So far we have covered the hazards that attack the welder directly: drowning, electric shock, explosive gas and heat. But a second family of risk sits quietly behind all of them, and it only shows up later. A poor weld made on a quay wall, a pier leg or a hull plate may look acceptable on the day and pass a cursory visual check. Then it fails months later, sometimes without warning, in a storm or under a vessel’s load, and the failure lands in public space above or beside it.

That is why underwater welding is not simply “welding, but wet.” The physics of the medium change the metallurgy of the joint, and the metallurgist’s problem is a safety problem. If the process is technically harder, every hour of diver time in the water carries more risk than the equivalent hour on a shop floor. The most effective risk reduction in this trade is often not telling the diver to be more careful. It is choosing a process that leaves the diver less to do.

Cooling rates: the number that changes everything

In dry air, a weld cools at a rate the filler metal and procedure are designed around. Surround the same joint with water and the heat is ripped out of it many times faster. Published technical literature on wet underwater welding reports cooling rates in the 800 to 500 °C range of roughly 415 down to 56 °C/s depending on depth and process, figures dramatically higher than in comparable dry welds.

Those numbers sound abstract until you follow them through the metallurgical chain. Rapid cooling means the weld metal and heat-affected zone (HAZ) have less time to develop toughness, so ductility drops. The microstructure can also end up hard and brittle, and a brittle microstructure combined with diffusible hydrogen is the classic recipe for hydrogen-induced cold cracking. Diffusible hydrogen, which permeates the hot weld metal during solidification, can sit in susceptible microstructures and crack the joint after cooling, sometimes while the diver is still in the water. The literature is blunt about this: diffusible hydrogen combined with brittle microstructures is a principal cause of crack formation in underwater welds, and underwater welding reviews in the technical literature describe diffusible hydrogen pickup as one of the defining challenges of the wet process.

Cracking: cold in the carbon steel, hot in the stainless

Different metals fail in different ways, and the underwater welder has to know which one they are dealing with.

• Cold cracking is the main problem when welding high-strength low-alloy steels and in dissimilar joints, where a hard HAZ meets a different thermal expansion behaviour and the combination of high restraint, high diffusible hydrogen and fast cooling tips the balance towards cracking.
• Hot cracking shows up in fully austenitic stainless steel weld metal, where the austenitic structure has low thermal conductivity and limited tolerance for the strain that occurs as the weld shrinks while still hot.

On a deck, a welder can grind a suspect area, re-run a bead and inspect. Underwater, repair means going back down, cutting out in the dark, and managing the same hydrogen and gas problems again, this time with less certainty about what lies beneath the previous weld.

Porosity: bubbles frozen into the metal

Porosity is present to some extent in all wet welds. It forms from molecular hydrogen, carbon monoxide or water vapour trapped as the metal rapidly freezes. The main factors are water depth, electrode covering and arc stability, which is why underwater consumables are formulated and classified differently from dry-land electrodes rather than simply being the same rod in a wet bag.

Each pore is a tiny stress concentrator and a tiny crack initiator. Multiply that by the hydrogen and the fast cooling, and the joint is well on its way to being the thing that fails later.

Arc stability and the effect of pressure

Increasing pressure makes the welding arc unstable. As depth increases, the arc voltage rises dramatically and the arc root contracts, so the arc becomes harder to hold and harder to direct. Operators compensate with special underwater electrodes, tightly controlled amperage and voltage, and short, deliberate travel.

This is a technical difficulty with a safety edge. An unstable arc means spatter, erratic fusion and more heat input in the wrong places, on a joint that is already cooling too fast and picking up hydrogen.

The blind weld: the bubble that protects you also hides the work

One of the strangest features of wet welding is that the same gas envelope that stabilises and shields the arc also completely obscures the weld. The welder cannot see the puddle, the bead or the toe of the joint. Work proceeds from feel, from a rehearsed procedure, and from the diver’s experience of the job in front of them.

That is a large part of why wet welding produces lower-quality joints than dry hyperbaric welding, and why acceptance of a wet weld rests heavily on procedure qualification and diver workmanship rather than on inspection of the finished bead. Add the fact that defects beneath the weld surface are difficult to detect and you get the inspection problem: non-destructive testing and inspection planning are more demanding, which means more diver hours, more time in the water and more exposure to everything discussed in the previous five sections.

It also explains why mechanical aids are attractive. Laser welding systems and other remote or automated approaches, including robotic work inside a local dry cavity, aim to remove the diver from the arc and from the blind-work problem entirely. Even simple mechanisation, like a purpose-built camera and lighting rig feeding a surface operator, shifts the judgment call from “can the diver see it?” to “can the team verify it?”, which is a far better place to be standing.

The practical takeaway

There is a direct line from metallurgy to safety. A faster-cooled, hydrogen-laden, porous weld that nobody can see properly is more likely to fail, and failure of a structural joint can put people who are not divers at risk years later. That is why the choice of process, the choice of consumable, the qualification of the procedure and the inspection plan are all safety controls in their own right.

It also reframes the question at the heart of this article. When a client asks whether underwater welding is dangerous, the honest answer is that the risk lives in the details: in the cooling rate, in the hydrogen, in the blind bead, in the depth. Fix those by choosing the right method and the job becomes manageable. Leave them to chance and no amount of caution on the surface will help, because the weld itself is the problem.

What the Welder Actually Breathes: Fumes, Gases and Confined-Space Exposure

There is a visual trick that fools anyone watching underwater welding for the first time. The arc sits inside a bubbling blanket of gas, and the fume looks neatly contained right there at the joint. It is not. The bubble does not remove the fume. It collects it and delivers it straight into the one sealed space on the job: the diver’s helmet.

That makes respiratory exposure the most quietly under-rated hazard in underwater welding. It rarely causes a dramatic accident on its own. It produces long-term damage in a workforce that is already small, and it is why experienced commercial divers need ongoing medical surveillance. OSHA’s commercial diving framework treats the diver’s breathing environment as a controlled parameter, not a detail.

What welding fume actually is

Welding fume is not campfire smoke. It is fine particulate made of metal oxides, generated when the arc or hot weld pool vaporises base metal, filler metal and coating or flux, then condenses the vapour into particles small enough to reach the lower lung. Mild steel, aluminium and iron fall under one broad category, “total welding fume.” Stainless steel, cadmium-bearing materials, lead-coated steel, copper, nickel and chrome alloys produce far more toxic fractions that must be controlled against their own limits.

The numbers that matter are measured in the welder’s own breathing zone, not in a lab:

Contaminant	OSHA limit	NIOSH recommended limit
Total welding fume (iron, mild steel, aluminium)	5 mg/m³, 8-hour TWA	See 29 CFR 1910.1026
Hexavalent chromium, Cr(VI)	5 µg/m³, 8-hour TWA (action level 2.5 µg/m³)	0.2 µg/m³
Manganese fume	5 mg/m³ ceiling	1 mg/m³ TWA / 3 mg/m³ STEL

Hexavalent chromium: the one to take seriously

Welding stainless steel and chrome alloys converts metallic chromium into the hexavalent form, and the resulting fume is genuinely nasty. Cr(VI) irritates and burns the eyes, skin, nose, throat and lungs, and it is a proven carcinogen. OSHA’s permissible exposure limit is 5 µg/m³ as an 8-hour TWA with an action level of 2.5 µg/m³, while NIOSH recommends a far more protective exposure limit of 0.2 µg/m³ for all Cr(VI) compounds. The scale is not trivial: NIOSH estimates roughly 558,000 U.S. workers are exposed to airborne Cr(VI) compounds in the workplace, with welding among the most affected industries.

Manganese follows a different pattern. It comes off ordinary steel welding and it targets the nervous system, which is why chronic exposure is associated with movement disorders resembling Parkinson’s disease. NIOSH notes that confined space welding can significantly increase exposure to manganese fumes, and OSHA lists respiratory irritation and systemic poisoning from toxic fumes and particles among the associated hazards of shipbuilding and repair welding and cutting operations. OSHA’s welding fact sheet lays out the whole exposure picture in plain language.

The underwater twist: the helmet becomes the delivery system

On a shop floor, fume rises and drifts away. Underwater there is nowhere for it to go. Wet welding surrounds the arc with a gas envelope of its own, and in a bell or helmet the diver breathes inside that envelope through a closed circuit. Fume that would have dispersed harmlessly over a grating is now concentrated in the volume of the helmet and inhaled with every breath.

This is why surface-supplied air, delivered from a compressor with verified intake placement, filters and continuous gas quality monitoring, is a genuine engineering control rather than an administrative nicety. Dry and hyperbaric welding is not automatically fume-free either: move the same arc into a chamber and you have replaced an open plume with a confined atmosphere that needs the same discipline, only with better monitoring available. And because wet welding is dominated by stick and flux-cored processes, the fume profile of the consumable matters as much as the base metal, which is worth understanding alongside any guide to why SMAW still dominates commercial work.

The exposure waiting topside

The diver is not the only one breathing this. Deck crews face fume carried out through the water and up the umbilical, plus hazards unrelated to welding chemistry: radiation from x-ray and radioisotope weld inspection, noise from grinding and hydrographic survey equipment, hot slag and sparks falling from deck level, and crane and vessel traffic working right beside the dive site. Monitoring has to cover the vessel, not just the helmet.

Controls that actually work

Certified local fume extraction where the work allows it. Supplied-air respiratory protection, or correctly fitted P100/HEPA filtering respirators, when extraction cannot keep up. Personal gas monitoring in the helmet and in the compartment. And surface-supplied breathing gas of verified quality, which is where choosing a reputable supplier and equipment source matters: a guide to welding gases, services and safety covers the quality checks that decide whether your breathing air is actually safe. *(The site may earn a commission on some product links.)* None of this changes the bigger point: the best fume control is still choosing a method that puts less metal through an arc in a sealed helmet in the first place.

Wet vs Dry Welding: How the Method Choice Changes the Risk Profile

Much of the danger in underwater welding comes from the environment, not the arc. So why not take the water away? Commercial diving crews do, and that choice is the biggest fork in the road for “is underwater welding dangerous.”

Wet Welding: Fast, Flexible and Forgiving of the Job Site

Wet welding is what most people picture: the diver works in full gear in the water and the arc burns in the surrounding sea.

Most wet underwater welding uses shielded metal arc welding (SMAW) or flux-cored arc welding (FCAW) with underwater-developed consumables. The flux covering or core decomposes into a protective gas bubble

Wet vs Dry Welding: How the Method Choice Changes the Risk Profile

Much of the danger in underwater welding comes from the environment, not the arc. So why not take the water away? Commercial diving crews do, and that choice is the biggest fork in the road for “is underwater welding dangerous.”

Wet Welding: Fast, Flexible and Forgiving of the Job Site

Wet welding is what most people picture: the diver works in full gear in the water and the arc burns in the surrounding sea.

Most wet underwater welding uses shielded metal arc welding (SMAW) or flux-cored arc welding (FCAW) with underwater-developed consumables. The flux covering or core decomposes into a protective gas bubble around the arc, doing double duty: it shields the weld from the water and gives the welder a rough view of the puddle.

Speed and flexibility win: a dive team, welding machine, cables and electrodes is a fraction of the equipment a dry chamber needs, and the diver can reach geometry no habitat could seal around. Hence its standard use on structural steel, sheet piling, quay walls and tubular piles.

Estimated cost and time for a dry welded repair is roughly double that of a wet welded repair, because of the support equipment that must be mobilised, rigged, sealed and controlled.

The price is metallurgical: water pulls heat from the joint far faster than air, driving rapid cooling, hydrogen pickup and porosity. The achievable AWS D3.6M weld class is generally lower than for critical work, and the welder works blind inside the bubble.

Dry Hyperbaric Welding: The Chamber That Removes the Water

Dry hyperbaric welding flips the problem on its head. A sealed steel chamber is placed around the workpiece or pipeline and the water inside displaced, leaving the welder in a dry, controlled gas atmosphere at the surrounding depth’s pressure. It is a pressure vessel, not an air-filled room: the welder breathes helium or helium-oxygen mixtures at the job’s depth.

Because atmosphere, preheat, heat input and parameters can be managed precisely, weld quality comparable to conventional surface welding is achievable. That is why dry hyperbaric welding became the established method for high-quality joints, pipeline tie-ins and subsea pipeline repairs, in North Sea use since the 1980s. Dry hyperbaric systems have been deployed successfully at depths of approximately 250 metres, with deeper work handled by closed-bell and saturation diving systems.

The downside is everything the chamber brings: the habitat must be fabricated, positioned, sealed against the workpiece and pressure-tested, and gas management, lighting, fire protection and gas detection must be engineered into a small industrial site on the seabed. The risk profile shifts rather than disappears: inside, the diver is no longer an electrician standing in a conductive medium, but a confined-space worker breathing a mixture they cannot smell, where a gas leak or fire has nowhere to go. Dry work removes the electrocution and electrolysis problems and substitutes a more manageable set of hazards.

Local Cavity and Dry-Spot Welding: The Middle Ground

The local cavity or dry-spot technique sits between the two. A transparent or open-bottomed enclosure is fixed over the joint and gas pumped in to displace the water from a small working volume, so the welder stays outside in the water but works into a dry pocket. Diffusible hydrogen in the weld metal from this technique runs roughly 5 to 21 ml per 100 g of iron depending on gas flow rate, and joints made this way can meet classification society requirements down to around 200 m.

How the Choice Actually Gets Made

Quality, depth and specification decide it, in that order.

• Critical, pressure-containing or fatigue-loaded joints get dry work. Tie-ins and pressure-boundary repairs go into a habitat.
• Lower-risk structural work gets wet. Piling, sheet piling and non-critical repairs, faster and cheaper.
• Depth can force it. Beyond dry system limits, wet is the only option unless saturation diving is mobilised.
• Specification can override everything. Classification society approvals and owner specifications often name the method first.

Why Inspection Is Harder in the Water

Assuring the integrity of underwater welds is difficult because defects are difficult to detect beneath the surface of the weld, and worst for wet welds, where the pool was never fully visible. Quality assurance therefore moves upstream: procedures are qualified in advance (ISO 15614-9 wet, ISO 15614-10 dry hyperbaric) and welders separately (EN ISO 15618-1 wet, EN ISO 15618-2 dry), under the OSHA commercial diving framework and the AWS D3.6M code. A qualified procedure and welder means the weld need not be perfect the first time.

Wet welding buys speed and flexibility at the price of a lower weld class and harder inspection; dry hyperbaric buys surface-comparable quality at the price of a seabed pressure vessel. Both are dangerous, just differently. Our guides to why SMAW still wins and all-position electrode selection cover wet work’s backbone processes.

The Mistakes That Cause Most Underwater Welding Incidents

Underwater welding incidents are rarely exotic. Nearly every serious event traces back to a short, repeating list of human and equipment errors, and each one has a control that would have prevented it. This is the troubleshooting list commercial diving supervisors run through before a job starts.

1. The isolation switch left closed or untended

This is the classic electrocution pathway. 29 CFR 1910.422(g)(1) requires the current supply switch to interrupt current flow to the electrode, to stay in the open position except while the diver is actually welding or burning, and to be tended by a dive team member in voice communication with the diver. A diver changing electrodes with live current in the water has no second chance.

Control: name the switch tender explicitly, verify voice contact before the first arc, and treat any switch movement as a spoken call, not a silent one.

2. The diver bridging the circuit

Touching the electrode or a live electrical part and the workpiece at the same time makes the diver the path to ground through the surrounding water. Equipment guidance from underwater cutting and welding suppliers is blunt about this: the diver must never allow themselves to become part of the electrical circuit.

Control: inspect electrode holders and cables before entry, keep gloves and sleeves dry, and never rest a live holder against a wet suit or a steel member for leverage.

3. Using AC power supplies

DC is the standard for underwater work because AC is more harmful to the body at lower current levels. The one recorded electrocution in commercial underwater cutting or welding came from the failure of a rectifier inside an AC-DC welding power supply, which is why modern guidance says AC power supplies are not recommended. Current selection and polarity checks belong on the pre-job checklist, not to the diver’s discretion once in the water.

Control: confirm a DC source, verify output settings at the machine, and keep a suitable stick welding power supply on the job list for wet work.

4. Damaged dive dress

A torn or abraded wet suit, or a dry suit with compromised insulation, removes the only barrier between the diver and the water they are standing in. A National Research Council panel on underwater electrical safety found that wet suits worn while cutting and welding underwater should have no worn spots, tears, holes or other indications of loss of integrity of the closed-cell structure.

Control: inspect suit, gloves, hood and wrist seals under light before the diver enters, and quarantine any suit of questionable condition. It matters for the electrode work too, since underwater electrode selection and handling assume a sound insulating layer.

5. Poor gas management

Cutting or welding into a pocket of collected hydrogen in an overhead space or inside a pipe is the second great killer after drowning. The fix is not more caution, it is engineering: flood, purge and vent the specific structure, monitor the gas, and plan where the displaced gas goes. Oxygen-fed oxy-fuel cutting makes this worse, not better, because it enriches the space it is pointed at.

Control: a written gas management and purging plan for that specific structure, with a monitor assigned and abort criteria agreed.

6. Loss of communication

Every control in the system assumes the tender can hear the diver, so a failed comms set is an emergency trigger rather than an inconvenience. The IMCA code of practice for the safe use of electricity under water addresses shock prevention and risks such as toxic emissions from insulation, but no control replaces a working voice link.

Control: two-way voice check before entry, a backup comms method on the surface, and a hard rule that loss of comms ends the dive.

7. Equipment misuse

Electrode holders designed for water cooling will overheat if used out of the water. Undersized or uninsulated cables, and connections that are hand-tight rather than wrench tight, are equally common. Carbon arc gouging tools carry the same arc gouging hazards plus extra heat load in the water.

Control: a documented tool-box check before the job, spares on the surface, and connections torque-checked by the tender.

8. Fatigue, cold and depth

Cold water drains body heat faster than a welding diver can generate it, and fatigue plus cold is where technical problems become emergencies. Depth adds narcosis, which degrades judgement exactly when judgement matters. Divers Alert Network’s commercial diving safety guidance treats team size, medical fitness and task-specific planning as core controls rather than paperwork.

Control: task-specific risk assessment, adequate team size for the mode, scheduled warm breaks, and a diver stood down on early signs of cold stress.

That list is short because the physics is not complicated. Water conducts, gas collects overhead, steel stays cold, and people tire. Professionals who work this way safely are not braver than anyone else. They are the ones who check the switch, the suit, the gas plan and the comms before the diver goes in, and who treat any deviation as a reason to stop.

Standards, Rules of Engagement and How Professionals Actually Control the Risk

Underwater welding sits where two heavily regulated activities overlap: commercial diving and welding. The framework that governs it is more concrete than anything most shop welders ever deal with, and it is decided by codes rather than opinion.

The Regulatory and Standards Stack

In the United States the base is OSHA’s commercial diving standard, 29 CFR 1910 Subpart T. Two parts matter most. 1910.422 contains paragraph (g) on welding and burning during diving operations: the tended current-supply switch, and the requirement to vent, flood or purge closed compartments before any cutting or burning begins. The same section demands two-way voice communication between each surface-supplied air or mixed-gas diver and a dive team member at the dive location, plus decompression, repetitive and no-decompression tables and depth-time dive profiles kept at the dive location, not in the office. Then 1910.410 on training and qualification requires personnel designated to operate electric cutting and welding equipment in diving operations to have experience or training in its safe use, and that is where OSHA cites violations when the gear is run by someone who does not know what they are doing.

On the welding side, AWS D3.6M, the Underwater Welding Code, is the reference document. It defines the variables that affect underwater welding and sets out welding and inspection procedures so work of a known quality level can be conveniently specified. That last clause is the point: D3.6M is how an owner says “this joint must meet Class A” and means something specific.

Then there is a layer people rarely consider. On top of welder and procedure qualification, additional project or class approval may be required by the operator, client or a classification society such as DNV, Lloyd’s Register, ABS or Bureau Veritas. For offshore platforms, subsea pipelines and pressure-containing structures, the class society is effectively the referee. A welder can be fully qualified under ISO and still be barred from that joint without sign-off on the procedure, the consumables and the diver’s ticket.

Divers Alert Network has published assessments of underwater welding hazards, and labour authorities issue case alerts when a fatality exposes a gap. One example: the Hong Kong Labour Department’s alert on the drowning of a diver during underwater welding work required a suitable diving supervisor to be appointed in writing and to be in immediate control of the diving operation, and specified that the supervisor should not dive while acting as diving supervisor. That single sentence captures a core principle: the person running the job cannot also be in the water.

A Working Checklist

Codes are abstract until you turn them into a pre-job list. This is the compressed version professionals actually run against.

1. Task-specific risk assessment by a competent person, not a generic procedure.
2. A written Diving Rules document, signed before anyone enters the water.
3. A named diving supervisor who is not diving and controls the operation.
4. Team size and medical fitness matched to the diving mode and task.
5. A gas management and purging plan for the structure, with a route for gas to leave.
6. Verified compressor air quality and intake location before entry.
7. Electrode holder, cable and insulation inspection, plus a ground check of the machine frame.
8. A knife switch on the lead side, open by default, closed only on the diver’s instruction.
9. Two-way voice communication checked before entry and after any change of task.
10. A rescue plan with emergency services notified, naming who does what in the first five minutes.
11. PPE including insulated gloves and eye and face protection, suit condition inspected.
12. Abort criteria everyone has seen: loss of comms, current in the water, gas alarm, cold stress.

Those abort criteria deserve emphasis. A job anyone on the team can stop cleanly is a controlled job. Writing them down matters because adrenaline makes stopping feel like failure, and the diver who has been underwater for three hours is the least likely person to call it off.

The Honest Conclusion

So yes, underwater welding is dangerous, and more dangerous than the same weld in a shop, but not mainly because of arc flash or arc eye. The real mechanisms are a short, documented list: electricity finding a path through a body in salt water, hydrogen collecting in an overhead, a diver who cannot see the puddle and cannot surface when something goes wrong, and cold eroding judgement. What makes the work survivable is not bravery. Each mechanism has an engineering control, an administrative control and a code reference behind it: isolating the diver from the circuit, purging and flooding the structure, a supervisor on the surface who is not diving, a qualified procedure and welder, abort criteria agreed before entry. The professionals who work this safely are not braver, only better prepared and quicker to stop. That is a system, not a personality trait.

For context on what a regulated qualification framework looks like from the inside, our guides to why SMAW still wins and to welding standards across industries are useful reading. On the money side, compare welding equipment costs and gas supply pricing against local labour rates and you will see why this stays a specialist trade rather than a general fabrication service.

Frequently Asked Questions

Is underwater welding more dangerous than regular welding?

Yes — significantly, and mostly for reasons that have nothing to do with the arc itself. The welder is simultaneously a diver, an electrician working in a highly conductive medium, and a confined-space worker, and the controls for each role can conflict. Add the dive hazards (drowning, cold, decompression, poor visibility) and you get a job that a competent, fully equipped commercial dive-welding team does routinely, but that is never routine work.

What is the single biggest risk in underwater welding?

Drowning is the outcome that kills most often, and it is usually the end of a chain rather than a single cause — a shock, a gas explosion, a loss of visibility or a suit problem that leaves the diver unable to surface. The control is not a single device but a system: a tended isolation switch, two-way voice communication with a tender, a written abort plan and a rescue capability that is rehearsed before the job starts.

Is it safe to weld underwater with a normal welding machine and electrode?

No. Underwater welding requires purpose-built equipment: a constant-current DC power source, electrode holders designed for water cooling with far heavier insulation, properly insulated and correctly sized cables, a positive operating disconnect switch in the welding lead side, and electrodes formulated for wet welding. A ground-fault in a standard machine, a damaged suit or a hand contacting both the electrode and the workpiece can put the diver directly in the circuit.

Why is direct current used instead of alternating current underwater?

Because AC is physiologically more dangerous at lower current levels — the National Research Council put the safe upper limit at about 10 mA for AC applied to a diver versus roughly 60 mA for DC. That is a six-fold difference in the current that reaches a dangerous threshold, which is why DC is standard and AC power supplies are specifically not recommended for underwater welding or cutting.

Can you really get an explosion underwater?

Yes. The welding arc breaks down water and generates hydrogen by electrolysis, and oxygen is supplied deliberately for oxy-fuel cutting. In an enclosed structure these gases can collect in an overhead or inside a pipe, and a flooded compartment is not automatically safe. OSHA requires such compartments to be vented, flooded or purged with a gas mixture that will not support combustion — and venting alone is not sufficient unless it actually removes the flammable gas.

Is dry (hyperbaric) welding safer than wet welding?

It is safer for the welder in several ways — a dry, gas-filled atmosphere means less fume exposure, better visibility, controlled cooling and weld quality comparable to surface welding — but the job as a whole carries more logistical risk: sealing a habitat, managing pressure and gas, and typically working at greater depth. Dry hyperbaric welding is also roughly twice the cost and time of a wet repair, which is why wet welding remains the practical default for structural work that does not need habitat-level control.

How deep can underwater welding be performed?

Wet welding with covered electrodes or flux-cored arc welding is performed directly in the water down to roughly 100 m, while local cavity (dry-spot) welding can meet classification society requirements down to about 200 m. Dry hyperbaric welding has been carried out operationally to around 250–300 m using saturation and closed-bell divers, with laboratory research continuing well beyond that; the practical limit is diver physiology and the logistics of building and sealing a chamber at depth.

What happens if a diver is shocked underwater?

A shock can cause respiratory or cardiac arrest, and burns are possible at higher current levels. Even a shock that causes no immediate symptoms should be treated seriously, because the effect of a given contact changes with depth, current, water salinity and suit condition. The diver should be recovered, the current switched off immediately at the isolation switch, and the diver given oxygen and medical assessment, with the incident documented so the cause can be removed.

Do I need a commercial diving qualification to do underwater welding?

Effectively, yes. OSHA’s commercial diving standard treats underwater welding and burning as a regulated activity within commercial diving operations, and 29 CFR 1910.410 requires that personnel operating electric cutting and welding equipment in diving operations have experience or training in its safe use. Professional practice also requires welder qualification to AWS D3.6M or EN ISO 15618-1/-2 and a qualified welding procedure (AWS D3.6M, ISO 15614-9 or ISO 15614-10).

How can underwater weld quality be checked if the defects are hidden?

It is genuinely harder, especially for wet welds, because defects beneath the surface of the weld are difficult to detect. That is precisely why underwater welding leans on procedure qualification rather than relying on inspection afterwards: the welding procedure specification and the welder’s qualification are qualified under the actual underwater conditions, and project specifications may also require classification society approval. Where a critical, pressure-containing joint must be verified, dry hyperbaric welding with planned inspection access is normally chosen.